My wife is the least audiophile person in the world. As long as she likes the song, she really doesn't care whether it's played on a kitchen radio or on a high-end system. I recorded a mix tape for her for Christmas. All songs came from my streaming service ... except for one that wasn't available online in the desired version. I recorded it from CD. My surprise was great when, while listening to the tape, she said: "That song sounds kind of different than the others. Somehow… better". What can I say?! If even my wife hears it, then there is definitely a difference in quality.
I agree that humans, which don't overthink anything still hear everything effortlessly :-) However, only one song is a very small sample. It could have been the song itself.
Very interesting! Yes, it certainly DOES say a lot that she still noticed a difference as well. For me, it's become virtually impossible to notice a difference b/t streaming (not sure if was 320 mbps mp3 or FLAC now) and downsampled from FLAC AAC downloaded files in terms of the quality of the track itself. HOWEVER, I still DO notice a difference in that the streamed version has a noticeable "hum" or buzz with it, where the same track (yes, same hi-res version of the same album) DID NOT have this issue. So that by itself still makes the downloaded version superior to the streaming version in my opinion.
My wife is the least audiophile person I know. She despises my hifi system. I bought her a little DAB radio a couple of years ago and it is her preferred listening choice. She is equally happy watching a movie on our 32 inch LCD tv in the spare room as she is watching it on our 77 inch OLED in the lounge!
I have always enjoyed and learned much from your videos. I am 70 years old with Tinnitus and VOR…so as a person I can not be that discerning audiophile has long gone, but I still am a serious and discerning listener. . I started listening to music as a love and passion since I was five, my dad was really into it and thus I never lost that love. I am not going to come across as any expert here but the gentleman’s quandary can come from so many sources or combination of sources. I spent 40 years in IT and communications. Yes, a network, shoot your 110v power source, neighbors and your wall chargers, phones, washing machines, ceiling fans, LED lights etc can add significant noise to any signal. But to think todays high speed internet is 'the' culprit to adding a significant or even perceived degradation in quality is a stretch. Yes, in a way it can but it would not be in such a way that the fidelity has been compromised in any meaningful way. There are exceptions but modern technology has so many error corrections the actual loss of the bits/data along the way would be rare or make that difference be rare. Again, most interference after it has arrived to your streamer will most likely be local. As a long time HAM radio communicator I watch this stuff on my equipment RF signaling equipment. Today in the electronic world we are surrounded by unimaginable man made RF noise pollution. A $7,000 radio and $7,000 amplifier Ham radio quality is measured on its ability to detect and remove this garbage. I know when my neighbor three houses down turns on her cheap ceiling fan. I can tell when the relays in the washing machine turn on to squirt water in the washer. I have to unplug my wireless desk side phone charger. Just touching the surface here with the RF garage that can and if you have Amazon cables or noisy AC for power, it can truly add to the loss in fidelity of your sound system, much more so than the ‘Internet’. But just saying it’s the music being streamed over the Internet causing the problem, if you have a good high speed service, would be the last place to look. Someone commented on this vid about the source of the master or remastered. I personally feel that is one of the largest issues or maybe just differences between what is streamed compared to CDs or remastered LPs etc. I find and hear this difference (having CDs, LP and Quboz). Sometimes Quboz sounds better than an excellent condition original LP and vice versa. Comparing my Dark Side of the Moon album that I got in I think 1971 is superior to the the new remastered I got off Amazon last year. That is literally crap!!!! But Quboz Dark Side of the Moon is fine. But other original release LPs sound better than Quboz and again, some do not. So I think where the master or remaster came from plays a big roll here. Second, DACs… The things that turn our digital into analog audio. I know you have far more experience than I do for sure but I have three and what a massive difference they make on streaming from so called high fidelity digital streaming sources. I use a BluOS as the streamer, but I do NOT use their DAC. I bypass the BluOS DAC. I have a really nice DAC built into my Anthem preamp , I bypass that too. I go straight from the streamer to the REM DAC and analog from there to the preamp. This is like night and day compared to using the DAC on BluOS, I mean very large difference in fidelity. The Anthem DAC is superior to the BluOS DAC…actually propably good enough for the majority of us. But it was not until I listened and configured the REM for my ears. It is not massive but it surely made a difference. So , as you know DACS play a big roll here. Great topic but there is so much going on here from where did the master come from, what DAC are you using, how good is your cartridge, man made RF interference everywhere, hearing loss. So in conclusion for me, I can not say any one source is superior or less than the other. As long as you stick with a NON-WIRELESS streamer (yes that can add substantial noise to the fidelity) and your source is hardwired to your router, you stick to, pick your flavor (I use Quboz), a good streaming source. Sometimes my LPs are better, sometimes Quboz is better ( I honestly gave up on CDs though I own them).. I will say this for sure at 70 years old and listening to my $30,000 of stereo and custom acoustics installed (wow acoustics that made a Hugh difference more than anything) is that in my 20s with far less superior audio equipmnet , it just sounded better back then haha. Our most important ability to comment and make comparisons in fidelity comes down to preference and how good or bad your ears have/will become.
Thanks for that post. We have some remarkable similarities. I too am 70, I also got my love of music from my dad, I too had a 40 year career in IT, and I also was a HAM radio operator (having built my first HeathKit with my dad for my 10th birthday). Very cool.
Thanks, I do agree in many points you mention, specialy about hearing loss. But my conclusion is that the problem with audio is created in 90% not by loss of hearing which affects top band, but it is in basic band (which in my not only opinion but also experience) is not exceeding 9kHz. That band is still easily heard in my 70 ties - I can still hear up to 12 kHz. With "additional" 12- 20 Khz we experience tones which are "emerging" as supplement of not good quality in basic band. It is because top band has very mixed phases and are easy to produce for tweeters . If only ears can do it even they are hardly linear (on reason of perfect reflections and games of phases) they are usualy easy percepted by young ears . Hearing in that high band makes false confidence that everything below is presented and creates impression of very detailed presentation. Even hughe ovation from gig sounds very elegant and distant but I say is supposed to sound as annoying shout. - If that top tones are not in recording (due to kind of music or mastering), band between 1 - 9 kHz is in parts damped due to different reasons then listeners have no such "supplement " in top and doubt about quality of whole recording or they think about their hearing loss. Beacuse It sounds simply dull. Reasons of bad quality I account mostly to hidden resonances in reproduction channel which are source of phase no coherence resulting in compensation by themself or their harmonics to zero as well as overlaying by louder resonant harmonisc. My experience is that with use of widerange capable in 30- 18 kHz speaker and simple SE mono amplifier I hear stronger and more exciting sopranos than with 2 way HI FI supplemented by electrostats and exceptional class B tube amplifier. I try many ways to make them equal but so far it is only enough near for not being worry about my hearing loss. Now I decided to build another better equipped stereo 2W SE tube amp to make better research . I could buy one but for me it will be more fun
I have an arcam avr20 with a couple of matching power amps running a 5.2.4 configuration and I would just like to say... coming from A 2-channel to A multi-channel setup, I couldn't believe how much my noise floor had risen. Anyway the more cables I shielded the quieter and my room became! One's got to remember unshielded cables act as antennas for attracting RF and the longer the cable the better the antenna! Probably explains why when I used to disconnect my Heights and Surrounds the noise floor noticeably came down. The last group of cables I shielded with my HDMIs and low & behold the room became slightly quieter yet again.
The streaming sevices are all about LUF's ( loudness units full scale), so the music is measured and altered for volume. There is no standard dB max or min across the services. In my view this is what people are perceiving as inferior sound quality as compared to their own CD because they are messing with the dynamic range of recorded music.
From what I understood understood about lufs, and loudness normalization, is they're not changing the dynamic range. They're increasing the peak volume levels so that they're hitting 0 DB, or saturation, for the highest maximum volume. But the overall dynamic range is the same
Followed Paul’s advice, I added a new CD transport to my high end DAC. The quality of my CD collection played on this is WAY better than any of my streaming services (TIDAL and Apple Music). The detail within the huge dynamic range is extraordinary.
@@josefbuckland i wonder why audiophiles accept such meters when they otherwise all day long cry "everything which isn't strictly needed has negative impact to the sound quality - even stuff which is completly bypassed just because it's there" 🙂
You speculated that the 'significant reduction in quality' (the letter you read) may due to streaming vendor's compression of their files. Well, why don't you produce evidence? I did a test, and found a 70 min ultraHD album downloaded from Amazon music unlimited takes up 1.35 GB of space. That is a massive file size! I just hope you have data/fact to support your claims.
I remember those albums of Nirvana on Qobuz. I own the original CDs and my GF the remastered versions. Some years ago Qobuz only had the remastered versions online. I compared them with my GFs ripped CDs and every single bit was identical. Later Qobuz added the original masters. I compared them with my ripped CDs and every single bit was exactly the same. If there's a difference, it's either - a different master (like mine and my GFs CDs) - played in non exclusive mode (on Windows the internal mixer resampling) - played over a different DAC (CD has to be played digitally to the same dac !) - played over a lossy transport (bluetooth, ...), or a noise transport (jitter, ...) I never experienced lower quality with Qobuz compared to my own collection and I consider my hifi system to be quite hi-end (focal sopra 2s here) Tidal (mqa) is something else. A 24 bit mqa is 15 bit audio (+1 bit mqa signal + 8 bit lossy compresse hi-res estimates). A 24/48k mqa decompressed to a lossy 15/96 max. The rest is upsampling to the maximum the dac can handle and adding nothing. The number on the dac display is meaningless too, as it shows what the master was BEFORE it was reduced to mqa. So if you hear mqa sounds worse, that's why. (But I don't have to tell you Paul, I saw your meaning about mqa 🙂)
As a mastering engineer, I'll add that CD and Tape are truly inferior mediums. Tape being the absolute worst. DSD being the absolute best but hard to process.
I agree with what you are saying here Paul, and I don’t have anything to add on that issue, but also I’d like to bring up the fact that you could have or anyone actually could have a particular music CD and depending on which version they have, an original, a remaster, or even which remaster, and when and where/how, they’re not always all the same in terms of sound quality and, in my experience, a lot of times they didn’t make it better .. that’s another annoying thing to think about.. it’s just sad that a lot of the people don’t have the equipment that allows them to hear the difference, but for the people that can, it’s a nightmare!
I think it does come down to the source material rather than streaming per se. I have some CDs that sound marginally better but with other titles the streaming sounds better. I always noticed a reduction in quality (usually soundstage and dynamics) if streaming without the Audirvāna program which talks straight to the external DAC without any computer OS port handling to get in the way. I find Audirvāna an essential piece of the streaming jigsaw.
@Tech-geeky that's just not possible. Tape is the least dynamic medium. When you master for Tape you need to pretty much cut everything under 45 Hz and a shelf under 200Hz. Low energy information would make the needle jump. The same process is used for the highs at a lesser extent. The transients are also more compressed. You just prefer the sound of Tape. It's not better.
The largest incremental improvement in what I hear in my music from all my individual equipment investments came from taking my favorite streaming music on Qobuz and buying the files from Qobuz and putting them on my server and playing them wired directly end-to-end (without WiFi, Bluetooth, IP, LAN, WAN, Internet, etc.). Further, as a former top exec at a global publically traded data company, I share your view that if it saves storage or bandwidth costs, it happens, no matter how well hidden or adamant the claims are to the contrary.
It's quite true. A cd player, based upon many many factors in respect to its build and components, will almost ALWAYS sound better than any streaming. It's proven and it's a fact.
My system: Auralic Aries G1.1 streamer ; Chord M-Scaler - Chord TT2; Music First Audio V2 Classic preamp; March Audio P421 Purify monoblock power amps; Martin Logan ESL-X speakers. I had a Cambridge Audio CXC CD transport for a short while but sold it as it sounded worse than locally stored flac and wav files stored on a usb flash drive connected to the streamer. I just can't hear any difference between the locally stored wac and flac and streamed 44.1/16 CDs from Qobuz; streamed hires files (in general) have an uplift in SQ compared to the flac and wav files. Interestingly CDs ripped via EAC sound better compared to the Auralic inbuilt CD ripper. One other (very small) improvement resulted from the inclusion of a iFi iPurifier 3 connected from streamer to M-Scaler. I may consider trying a better CD transport on a home trial basis if there is a prospect of wringing out a bit more SQ from a digital source; however, until I can hear an appreciable improvement I'll eschew the claimed superiority of the silver discs for now.
I often find albums on Qobuz that have been remastered and sounds much better than my original CD, a good example would L.A. Woman or Monty Alexander live in Montreux (1976) very well recorded maybe I should purchase the download to compare but my current set-up (no hard drive server does not allow this test). I understand this is mostly due to earlier CD not so great quality rather than answering the question yet I’m really happy with the overall Qobuz sound quality (Wilson speakers, Moon Audio streamer dac)
I knew it!! Glad you're exploring this and great question posed by the viewer - I've exactly the same experience. All things being equal, playing via Tidal doesn't sound as good as playing my Flac's. I'm not disagreeing with anyone else's suggestions (I've tried download and normalisation off etc etc) - my experience is I can't get near my Flacs. My system isn't even particularly high end for the community. This isn't a question getting a better hardware, cables or room treatment - it's a question of 'What am I paying Tidal for' (library aside).
A couple of controlled tests: 1) The Tidal app for Android has a download feature, where one may download the particular music for later offline playing. One can then do listening tests streaming it vesus not streaming of the exact music track. 2) I have also done direct A to B comparisons of TIDAL streaming vs CD files that had been copied to my local laptop. I could not tell the difference for this particular music, but I will try additional tests.
Website try and do the same... "hearing tests" : They take a track in different bit-rates and/or HQ in some cases if you have the equipment to hear it on, and lets the user judge for themselves. Its getting crazy... No one knows how to use tech, our to protect from malware, Now they are shoved into this audio-phonic space ? *walks off*. If CD's were cheaper to offshoot the ever increasing streaming services catalog, you bet i'd be getting my music on compact disc.. At least you can say you OWN it.... Streaming services, you never own it... People still think in 2023, they own it, but all you gotta do it look at the arguments when it happens.. What does that tell you? No one really understands do this day. Its a never ending cycle which will never understood. And we call that better ?? All were really riding on in 'convince' and 'price' That's it, because it defiantly is not about quality...
FLAC (8) level lossless compression should be fine. I did hear of one musical artist who used Deezer. He uploaded his songs and then later downloaded them and found them to be perfectly identical. What other services are using is up for question but many think Qobuz sounds best. Tidal and Apple Music are ranked about the same. This is one reason many "audiophiles" say you have to spend $$$ on your digital front end to get the same sound quality you can get from a $500 (and up) CD player. Some even go as far as saying you need a better router too, use wired Ethernet if you can, and add better, quieter power supplies to all your "digital front end" devices. Jitter should be taken care of by any quality DAC with a buffer and its own internal reclocking, so that isn't (or shouldn't be) the cause for any sonic differences you hear. It is an interesting observation though, that many people make. Streaming sounds "good", but a local CD played on your own player going through the same DAC can sound a touch better. Hopefully all that will be better ironed out in the next few years as streaming becomes the de facto method of music listening, except for dyed-in-the-wool audiophiles who insist on CD and vinyl albums. (I like both just fine - and the great music discovery you can get from streaming). Heck, I still play many cassettes and love them.
_Why streaming music is lower quality_ it is only when a different and lower quality conversion is used. If you stream HiRes then the quality is as high as it can be
Me too. It's a big difference for me as an audiophile, when i play dvd format and swapping it with streaming serv. (both use rca) there's something missing, the dynamics, the lowest and the highest frequency. The bass in dvd format is more ample, and in the high frequencies you can really the ping in hi hats and it's more life like
I have very little interest in streaming. As someone that has accumulated a vast library that I keep on my hard drives, I would, in a minute, go for an integrated with DAC that takes input from a dedicated SSD to feed my speakers. Two things have kept me from doing that. 1) It seems the industry, for some reason, does not want us to, so finding the components is ridiculously difficult (Lumin M1?, Innuos Pulse?). 2) There doesn't seem to be an interface solution, except to use a spare laptop or your phone and a ridiculously expensive service like Roon. In my case, I think it would be nice to navigate my library on my TV.
Dear Paul, streaming from the internet means that the file has to travel from its storage location through copper wires and optical fibers for thousands of miles to arrive at your receiving device. Many storages and servers are in Asia and Europe. The Eternet quality cables on internet boxes, modem, router, and signal speed are extremely important. The WI-FI is mostly used among several devices, which compromise the download speed and stability quality. I recommend checking the internet installation and all the items I mentioned above to ensure a clean and fast stream. Try to use new CAT 8 Eternet cables of the Internet setup and connect the DAC or streaming device directly on an Eternet port. Taking this measurements will help to receive the most pure signal possible. Paul, thanks for sharing your years of experience with us!!!🙏
Hey Paul I tried Qobuz streaming on a 7 day trial period after my son was raving about it and compared it to my CD copies and I was extremely disappointed to the point where I was questioning my own hearing. The streaming sounded flat and less dynamic than my CDs. It was obvious to the point of not even attempting to listen for subtle differences. I won't be subscribing to any streaming services anytime soon, I love my CD collection and I'm sticking with it.
I just stream UA-cam Music (the paid / not free service) from my iPhone / iPad into one of my nice vintage systems and I'm a very happy guy. I have zero interest in chasing a better sound. To me, it's all about the emotional connection to the music. To each their own.
Thanks for the download your songs to get better quality suggestions below. I immediately tried that and compared the Tidal download vs the CD rip of the same CD I have. Previously the CD rip was audibly better. Now with the Tidal downloaded version they sound identical after I AB compared then. All this through Zenstream via Tidal connect. Quite a big improvement so thanks.
I know your post is a year old, but I’m curious as to how you replayed the downloaded Tidal music through your system. Tidal downloads are only possible within the Tidal apps. I could be wrong.
Nowadays you can stream CD quality or better with lossless FLAC. Technically a 192kHz 24 bits track via modern streaming (such as Amazon Music HD) is much superior to CD or DSD quality. If it sounds inferior it’s because the source track is inferior. MQA, MP3 or other lossy formats of course will be inferior. Perhaps the truth is a combination of streaming providers still haven’t cleaned up some of their music repository to CD files and some files intentionally being optimized to mobile devices with more compression for better per device revenue (more clicks). The notion that there are data errors, jitter or processing noise involved was valid 2 decades ago but not nowadays where DACs can manage the flow asynchronously and ‘modern processors crunch FLAC files as a tiny effort. We gotta talk about today.
There is more than that… when we supposed to be lossless, powerful download capacity, same system , same DAC… note the same results? Streaming company found a way for cheating… compression for storage and uncompressed for used? Don’t know… but there is something else!
@@Skye_the_toller Possibly some tracks are still saved in a lossy format and streamed in FLAC, which is sort of cheating. From what I heard this is just a transition issue as services go lossless.
"Technically a 192kHz 24 bits track via modern streaming (such as Amazon Music HD) is much superior to CD or DSD quality" : NO, nothing beats the NATIVE DSD since it a has sampling frequency of 2.8224 MHz and that it has not been remixed (because remixing a DSD recording requires a conversion to PCM, and the original resolution is then lost). Try it, and you will be blown away by the quality, dynamics, resolution, and realism! ;) Be careful: there are plenty of SACDs on the market that incorporate DSD files that are just conversions from PCM recordings. Choose a brand like Pentatone (or others) that makes native DSD recordings.
@ I’ve listened to DSD and when it sounds awesome it also sounds awesome in PCM version. It sounds great because audiophile care has been put into the production process. All new DSD distributed music is an inferior transcoded lossy version of PCM (call it DXD or whatever) and still better than those who went through a noisy DSD DAC/analog mixer/ADC “PCM free” mastering process. Only when older music existing on analog tape is to be digitally distributed, DSD can make some sense, but since almost nobody is using DSD, best is to distribute it in better than CD PCM. The sampling frequency of DSD is only with a 1 bit sigma delta value that yields a lot of ultrasonic noise also. A PCM stream of 192kHz 24 bits will have excellent precision in every measurement and with such sample rate you can get the phase and level right for any audible frequency. Some people say DSD-64 is equivalent to PCM 96kHz 24 bits but I would prefer 192kHz 24 bits to leave any doubt out. Already at 44.1kHz CD quality sample rate and the average audiophile not hearing much above 12kHz, the treble precision issues and low pass filtering artifacts are not as important as you imagine.
I use a Naim Uniti Atom to stream Tidal and my local FLACS. and can't say I notice any real drop off between the two or the CD played through a transport. Although I don't get too hung up about ultimate SQ, I'm more of a music first listener. When I want the best SQ I play Vinyl.
Different masters/editions aside, you have higher potential for SQ with your digital files as they have a higher capacity than vinyl, and have less distortion added from the format.
I'd love to hear what others are doing. Here's my current set up. I started streaming and purchasing hi-res and CD quality music through Qobuz. I stored all of my CD's and purchases on a Bluesound Vault. I then purchased an NAD M10v2 to integrate with the Vault. I have nmy NAD CD player connected to the NAD M10 with RCA's and the Vault could be connected but it isn't because it's sending the music to the NAD through my home network. So far I can't hear a difference listening to a given album via the wireless Vault connection or the CD player. The biggest difference is how good/bad the recording is and again it sounds good/bad using either method I described.
I started off about 2 years ago to make my listening more convenient using the Vault 2i. I have over 12K songs in FLAC and at first played through my HT system all hard wired, then got the little Pulse Flex 2i's for the kitchen (sold all my albums and 35 yr old high end equipment), just after Xmas added to my unused 30yr old sound room new speakers, sub and integrated amp hard wiring the Vault to it. I also did all the treatments for the walls and at 70 I like listening to music more than ever. I also spent a bunch on cables as I made a pair of interconnects for my old system which blew me away on the difference between the class A power amp and pre-amp. Paul's video makes me wonder if getting a subscription on Tidal is worth it, Qobuz isn't available in Canada due to government rules.
Interconnect quality might also contribute to the problem. When I first got a streamer I connected it to my DAC using a generic USB cable I happened to have on hand. The result was not good. Suspecting the source of the problem, I replaced the generic USB cable with a fairly high-end one and the improvement was dramatic, almost but not quite as good as playing a CD or SACD.
Can anyone explain how this could technically make a difference with the audio quality? A USB cable exists only in the digital domain, meaning the “quality“ of what is transmitted, will be identical on the cheapest of cheap USB cables, and one made of solid gold, the only difference should be the speed at which the transfer of the ones and zeros occurs. Unless I am missing something?
I'm fairly certain that services like Tidal and Qobuz don't rip their own content from CD, at least not for their entire libraries; I suspect they get digital files from the music labels then process and encode several copies at various bitrates and formats. Assuming I'm correct, It's obvious that the digital files aren't what's on a CD because both services have 24/96 streams of many of their songs. It wouldn't surprise me at all if one or more steps in the processing and conversion pipeline are impacting the sound, maybe even deliberately.
I would tend to agree that it's the labels that are supplying the digital files, and thus the overuse of the "remastered" moniker. Putting on my tinfoil hat, I'd go so far as to say that those supplied files likely have a sonic "watermark" to tip the labels off on which streaming platform's copyright/encoding-decoding process has been broken (and the digital files end up *somewhere* else for free).
There has to be compression somewhere down the line. The data goes through so many servers with so many variables along the way. On top of that expense, and connection quality. Best just straight downloading the file, or buy physical media.
Media streaming uses UDP which doesn't do error correction. This is due to the continuous and time sensitive nature of streaming. So if packets get dropped or bits get flipped, your DAC will have to fill in the gaps. Think of it like quantization noise. It's the same with streaming TV. Sometimes the picture gets blurry when your WiFi is acting up or your ISP slows down.
"Media streaming uses UDP which doesn't do error correction. " Wrong - most streaming provides are using TCP/IP. But even when using UDP that does not mean that there is no error correction. It only means that no error correction happens in the network stack. Still applications using UDP can implement their own methods for error correction which might be optimized for streaming. For example the application using UDP may decide that it only request a resending of an defect package in case the buffer is filled enough to have enough time for that package to be resend. If the buffer is nearly empty it might be better to play the defect package as there is typically only 1 bit wrong in that package and such a small error is normally not hearable while an interruption due to a buffer underrun is always hearable. "This is due to the continuous and time sensitive nature of streaming. So if packets get dropped or bits get flipped, your DAC will have to fill in the gaps." You have no idea how streaming works. Internet speed is varying all the time - normally data is transmitted much faster than it is "consumed" by the DAC but there might be also short periods there no packet is received at all. Packets might even arrive in wrong order. Therefore there is ALWAYS a buffer before the DAC and as long as that buffer does not run empty (which never happens if your average Internet speed is much higher than needed to transport the amount of audio data) where are no gaps which have to be filled. "Think of it like quantization noise." Quantization noise has NOTHING to do with streaming. Quantization noise is caused by the Analog to Digital conversion there analog values are converted to "samples" and this samples have always a limitted resolution. Audio CDs are using 16 bit samples - so they can "only" represent 2^16=65536 different values but the analog source has an infinite number of possible values. There are infinite different voltages between 0V and 1V but with 16 bit samples you can only encode 65536 different values - so sampling always causes a (small) deviation form the original value which is called "quantization noise". But even using 16 bit samples this "noise" is so small that it is not hearable. Using a technic called dithering you can even "move" that noise to frequencies there the human ear is less sensible. "It's the same with streaming TV. Sometimes the picture gets blurry when your WiFi is acting up or your ISP slows down." That happens when there is to much traffic on you WiFi (or to many other WiFis on the same frequency band in your neighboorhood) and/or if you Internet connection is not fast enough or its speed is varyiing to much (e.g. due to a shared medium which might happen if your Internet access is based on a former cable TV connection their many users are sharing the same cable).
@@thomaswalder4808 I appreciate your insights but not your condescending tone. This is a friendly forum. People who enjoy music are usually cool. If you feel the need to bash people with your superior intellect, seek out a political forum.
@@sly_perkins It is not my "superior intellect" - it is just about facts. If you can show that anything I wrote is wrong I will apologize. And by the way - this is not a HiFi forum - it is UA-cam
One of the things that was touched on was the hostile environment an stream has to traverse in order to reach your system. There are many outside factors that can affect it through all that copper/fiber it has to travel through. That being said, I listen to rips (FLAC) off a music server, CD, and stream. To be honest to my ear the sound quality seem to be on equal footing. The only time I hear something that has a noticeable improvement is contrasting a well engineered SACD versus a High Res stream from Tidal and the like. When listening to the same album from the SACD. The sound is cleaner, faster, and having more detail than the high bit Stream version of that album. As mentioned at the beginning, my theory is that the signal has a far shorter path and there is less chance of grunge from things like routers and switches getting introduced.
It's physically impossible for routers and switches or any network or computer device to make any kind of sonic difference in the way your thinking. That's not how computers work, at all. In all instances, every single time, data moves from one device to another, it is copied. Even within a single device, for example from the storage to memory, or memory to the CPU or CPU to network card; that data is copied bit perfectly to the next stage. Noise is ignored and not transmitted to the next step, the signal is regenerated at each copy. Every single time. Noise CANT be transmitted beyond a single step it's just not possible. It's not a continuous chain like analogue. A digital chain is discrete steps and is completely immune to noise below a certain threshold. (Which is a very high threshold and extremally noticeable when it happens. Buzzes, pops, pauses etc) This is happening in side you computer, streamer, CD player and DAC all the time. Each step may introduce it's own new noise, sure, but it doesn't propagate. This is kind of the entire point of digital signals, noise imunity. The ONLY time noise and/or jitter can make a sonic difference in the digital realm is the final step, inside the DAC, hitting the converter circuited itself. Any high end DAC should have galvanic isolation and jitter compensation of some kind before it gets there, though not all do. If your DAC is not doing this, then the problem is entirely, 100%, the fault of the device feeding it, not something up stream. If your DAC is fed over USB it can only be the DAC's fault. The only other possible explanation is that the set of bits you are receiving is different from the digital master. Which can only happen if your streaming service is lying to you.
@@G3rain1 Agree, sadly some companies want to make snake oil business out of the data path to the DAC. For anyone understanding how IP networks work, this is quite laughable. Even any cheap modern DAC can handle buffering packets and run the clock with zero audible jitter and talks about “FLAC decoding noise from processors” and other such topics are also in snake oil territory. Rather modern DACs can output with SNR near 120dB with proper internal power rails filtering. With bad engineering, all things can go wrong, but that doesn’t mean we need to use them as real concerns. Modern online audiophile streaming is already here and companies need to adjust to this reality with real products. I’m thinking about putting a pair of LS60 in my gym room. KEF is getting it.
Pure nonsense - ethernet by definition is bit perfect - there are data way larder and much more sensitive to a single bit flip than laughable small audio
@@CraigArnolduk and that's the problems: audiophiles don't understand digital technology nor ethernet nor physics - hence you can sell them anything and as long it's expensive enough you can call it "Highend"
I'm currently investigating this. I had problems earlier with my stream and I started tinkering around. I ran direct connection from my modem/router as a bench. Then I tinkered around with an external USB post antenna replacing the wifi antenna with a bigger one I had laying around and that maxed out the signal. I had a solid 54 Mbps (former stream was 3 bars). From then on my stream from Apple and Amazon unlimited sound fantastic. I tested CD vs stream adjusting my volume with a DB meter to specific marked points in a song. After doing that output match I could not tell the difference in the stream and CD on two direct comparisons. More testing to do, but I could not really notice a difference running through my SMSL DAC.
When I was rebuilding a new system from scratch cheaply, I didn't have a streamer, straight from computer to DAC. The difference in sound quality between streaming and local files was big. Streaming sounded awful. I added a streamer= computer-streamer-DAC and it made a huge difference, now the difference in audio between streaming and local files is, small enough that I often can't tell the difference to be quite honest. Your mileage may vary, but I learnt a while ago that if your source is a computer, you really should think about a streamer, sound quality difference is not small and there's some great streamers for affordable amounts of money.
@@cliz305 signal chain, so laptop, streamer, dac. Having a streamer unit in between DAC and laptop to isolate electrical noise. I have an iFi Zen Stream, other streaming units are available, including I've heard- some very very good raspberry pi type units, very affordable. I think Cheapaudiophileman has recommended some Wiim units, something like that- best to check on his channel.
The stramer is more important than many people belive, I just upgraded my pi hifiberry digi+ pro to a Ian Canada Pi build, and the difference is night and day. I spoke with a friend about it and we made an anlogy to turntable, where the streamer is the pickup, and dac is the deck/pickup amp. The streamer I build is battery and supercapacitor based in powersupply, reclocker and transport and it is not a small defference, it's actually quite huge both with streaming and local files.
For those that are qurious about more info: ua-cam.com/video/Tc_QfRTbb6U/v-deo.html And here is what I wrote in the thread after building it and first couple of listens: So I buld it. I chose the PurePi FifoPiMa 1.5 TransportPi AES and Accusilicon 45+49mHz clocks Had an Audiofonics Powersypply and a Raspberry 4 from an earlier build based on HiFiberry digi+ pro It was very easy to assemble the boards, and didn't touch anything else. Connected to an Lyngdorf 2170 with a Supra COAX cable and Raidho XT-1 speakers with Some Dali Cables, no power conditioning. What a difference against the HiFi berry but also a friend that has a Aries G1 was blown away with some, maybe most aspects of this build, he promised to bring Aries over next time he's in town as his stereo is much different to mine... What is immediately apparent is the way the sound was flowing with ease, like I only have heard from much more expensive stuff, the effortlessness comes from a total black background. and the imaging much more 3D and holographic, what also hit me was how the bass on older recordings was fuller and on newer recordings much more nuanced, everything sounded much more coherent especially well recorded/produced music sounds much, much better and somehow more sound, but the one thing that stands out is the ease, how the music flows, and that aspect surprised me the most in how much impact it has on the whole picture... This upgrade is sonically the biggest one I have experienced in quite some time, I just went from a Digi+ Pro, Hegel H90 and a RME dac, I'm sure the Lyngdorf also has a hand in this, but it went from different but same with the Digi+ Pro, H90 and RME to something very, very special with the Ian Canada streamer... I'm for one am a very happy camper. I'd just whish I had tried this streamer with the RME I was so very fond of.
I bought a cd from yatao and a wav download from the same album of them. I burn a cd with the downloaded songs and then when I received the original CD compare both.... The original CD was better sounding than the downloaded versión. Streaming and download music are practical. But with time I became aware the sound quality is relative. And most of the time depend of so many factors in the audio chain. From the recording to the final system that will play the song.
I think his audio quality is being automatically adjusted when his Internet connection has a poor bandwidth. That option is usually turned on by default but you can turn it off but then you'd have your music buffering in between if your connection drops.
@@ThinkingBetter A lot of people say that, but it seems that no matter where you live, there always seems to be connection problems of some sort, whether it be local RF interference or weak signal problems from the ISP or whatever, there’s no flawless system. Obviously better to use a wired connection to avoid some of the wireless pitfalls, but unfortunately .. I live in a heavily populated area and nothing but bandwidth and connection problems.
I am connecting my laptop as a streamer using WI-FI. That might have something to do with the quality level. I am not sure yet if the recordings themselves are tampered with or the delivery process is compromised. When I play yessongs on vinyl it sounds great. The HD version on Amazon sounds good but loses depth and width on the Soundstage. I am still new to this so it might be me.
You didn’t mention the software used to play it… the differences are often the apps or software used. If u use something like roon it will play local or tidal or qobuz with the “same” software removing this. The main reason I see for differences is their original copies of albums are different to yours. You may have a cd from one country that’s different from another and the one on streaming was from a diff master etc. Storage is cheap these days. That’s really not an issue it’s just a cost of doing business like anything. 5 bucks per terabyte per month isn’t expensive when revenue is in the millions per month Think about it. 50 million albums if on average 1 gigabyte per album is 50,000 terabytes which is 250k a month storage costs. If they have 5 million subscribers at 20 bucks a pop like qobuz pricing that’s 100 million a month revenue
1) Similar to a vinyl rig, IME it takes care to get streaming to be on par with local file playback. Your modem, router, switch, streamer and all their power supplies' quality (or lack thereof) matter. Cable and unit isolation also matters. 2) Can local file playback equal CD playback? Yes, but it depends on the quality of your DAC versus your CD/SACD player. This is my experience with a very resolving system and much experimentation. 3) Most people don't perform equal caliber CD player to streamer comparisons, with many folks assuming their CD player's USB input (using the internal DAC) is as good as the CD player's primary (more sonically optimized) playback stacking the odds against streaming.
Personally, Tidal's MQA leaves a lot to be desired. The only two streaming services even close to what I would consider as hi-fi would be Amazon Music and Qbuz. I feel like Qbuz has a little bit better dynamic range and sound stage, but Amazon's catalog is much larger. So I mostly listen to Amazon Music's Hi-Fi service. It's crazy how different music can sound from artist to album to producer to playback method to the equipment. It's a very deep rabbit hole.
Yes, Amazon Music HD streams lossless CD or better quality already and it works. Someone should do some A/B data comparison with the real CD data to see if anything is compressed or done on the Amazon side on particular tracks. If so, it’s human actions doing it and shame on that.
@@bkkersey93 No need to debate too much. Fact is that MQA is a proprietary LOSSY format. FLAC is an open LOSSLESS format. MQA was created for those who wanted more “quality” than MP3/AAC, but didn't want the file sizes of FLAC. MQA is alright if your internet speed is very limited but there is no sense of it if you have a modern adequate connection that can also do much more demanding tasks such as 4K video streaming.
In my experience the streaming service is the main factor, Tidal is for me the lowest quaility, Qobuz was slightly better, I have not tried Apple music, however for me the best sounding streaming service is Amazon Music HD, I have music on CD and when compared to the same track from Amazon Music HD, I cannot really hear a difference, and I do have a hi resolution system. However, the quaility of the recording /mastering seems to be the main factor, a poorly mastered piece of music is poor from CD or Streamed, a 192kHz piece of music from Amazon HD vs a 44kHz CD recording is noticably better. How I see it is that there is good and bad in all the different mediums, but for me Tidal is by far the worst......
Pre digital everything was simpler. Not simple, but simpler. I can see no other logical explanation than compression here. Im not an audiophiliac Luddite. Dolby took the obvious and standard approach with their product back in the day. If it is possible to manipulate wavelengths electronically by computer then its theoretically possible to do the same with noise/THD. I think a new approach is needed. We want natural 1st generation copies of the original performance with nothing ever added. We want no THD or noise. "Compression is suppression and causes audiophile depression"
Hi Paul - wish you a happy new year and by the way: I am currently building a system that outputs all digital sources via just one DAC. I need this "standard" music source for a small development around a ClassD amplifier. Luckily, so far I have found that the sources, transport routes and compression methods have no influence on the sound if the file then got to me without any bit errors. To avoid jitter, etc., I use an old Intel server with 2 XEON CPUs and 138GB RAM, in which the music file has to be complete before I play it. The fact that CD players, streamers and whatever else sound different is due to the fact that different DACs and different electronics are used for playback. So it will be impossible to hear the same music with no audible differences between the CD and the sound card. 🙂 Andreas
@@williammurray9055 Hi William Yes, my program loads the file before releasing it. More importantly, I can send a constant amount of data to the DAC at a guaranteed consistent rate. I use Linux for this, since Windows is not real-time capable. For my project, I chose the CD-quality WAV file format. The bitrate is then 2 channels x 16 bits x 44.1 kHz = 1411.2 kbps which are constantly sent to the DAC. Of course, this doesn't work directly with compressed files, which is why I convert the data to WAV format in my memory before playing it. This is not rocket science either, as there are already very good software components for the individual tasks. 🙂 Andreas
The differences being heard are most likely nothing to do with the streaming services. It is most likely that the music on the CDs listened to are older masterings which tend to be more dynamic. Streaming services tend to only get the most recent masterings which are almost always more dynamically compressed than the original mastering. That it is the most likely answer not bandwidth or file compression.
What you said is what I've heard for years about streamed music. It makes sense as portability dealt music fidelity a serious blow years ago and settling for less became the norm.....till the young folk brought vintage equipment back.
I have a substantial vinyl collection and an even more substantial -CD collection, vut, these days I stream the vast majority of my music, it’s just so damned convenient. I have a highly resolving system incorporating a premium, state of the art independent DAC. I subscribe to multiple streaming services but, Apple seems to be the favourite when my streamer is searching for tracks, albums or play lists. Yes, I have had friends and family round to help me do a blind test between a track played from a CD and that same track streamed from Apple. If I listen intensely, I occasionally hear minuscule differences. That doesn’t make one format better than the other, it just means they are different. I do however find that if my streamer is connected via my home broadband network, there is a notable deterioration in sound quality compared with using a mobile router. There could be 101 explanations for that phenomena. By the way Paul, I can’t speak for other streaming services, but I do know that Apple has a library in excess of 76 million tracks. Just as an aside, I’ve also blind tested friends and family, playing the same track on vinyl and CD. The majority preferred vinyl. That doesn’t make vinyl better, it just means that on my system, the majority of friends and family would choose vinyl.
The thing that drives me bonkers is relying on what sites like Tidal and Qobuz say about the bit-rate and quality without being able to easily check it. To take an audiophile classic, Michael Rabin's The Magic Bow, as released by EMI on CD (whether as Strings by Starlight or in their box sets of his recordings) has a sample rate that is not full CD quality. I can rip my CD and look at the spectrum analysis and readily see that. With a streaming service, it's nowhere near that easy.
I dont see how seeing the Spectral Analysis of a tune could give you any indication of the quality of that file, with test files with tones and such sure, but not music.
" The Magic Bow, as released by EMI on CD (whether as Strings by Starlight or in their box sets of his recordings) has a sample rate that is not full CD quality" If it was released as an audio CD it must have a sample rate of 44,1 KHz. In theory they could do the analog to digital conversionen with a lesser sample rate and then afterwards convert that to 44,1 KHz by interpolation - but why should anybody do such wierd things?
I stream tidal a lot on high res. I got a Fiio dm13 (cd player) and started listening to some of my cds. Wow. It was apparent how much more dynamic, open, overall dramatically better. I love streaming for the variety and convenience, but the sound from the cd is way better.
Audiophile switches definitely matter. I was sceptical about it at first (bits are bits) but then I got a Sotm switch and was pleased by sonic improvements even with a stock ethernet cable.
Only when you are dumb and don't understand ethernet - your holy switch by definition can't distinct if a packet belongs to an audio stream or a word document
@@Harald_Reindl Only when you are more dumb not to realize there is electrical noise passing into your streamer together with the data stream. But the dumbest people think they know everything by not trying out by themselves and not hearing what others share
@@larazss3254 Muhaha electrical noise on the ethernet layer? How do you imagine that could change the decoded data? You morons know nothing about digital data and layers
It depends on the songs....some songs sound great...some sound noisy....it all depends on the original studio recording quality I guess....and adding a DAC in between the phone and your output will bring the audio quality the to the next level.
Happy New Year to everyone at PS Audio. This was an interesting question you answered today. I recently switched from using my IFI Go Blue from exclusively in Bluetooth mode ( Apple iPhone using Amazon Music HD) ..and it’s nice . I recently switched from Bluetooth to directly wired from the iPhone ( with the abomination of Apple connectors) and WOW…is it a night and day difference..nope ..but there’s more space , more dynamics and a sense of it just bringing more to my ears . Take care and may ‘23 be a good one for all.
Let's assume for a minute that you're using a computer, the stream is bit-for-bit the same as a CD rip, and it still sounds different. We have to ask what else is different? Is there two different pieces of software you are using to play the FLAC vs the stream? If so, are they using the same driver stack to communicate with the DAC? (WASAPI exclusive, ASIO, kernel streaming, DirectSound) Are they resampling to the DAC's sample rate using different methods? When I used Tidal, there was definitely a difference between the CD rip and the stream. And not just the mqa files, as those we know are not bit-for-bit. I have since switched to Qobuz, (I will hold my tongue about Tidal for now) and, after making sure Qobuz and Foobar were both using ASIO, I could no longer hear a difference. I suppose there should be should have been some difference as Foobar was set up to do its own resampling, but it wasn't big enough to notice right away so I did not do any further testing.
yeah, it is about exclusive and non-exclusive access of the player application to the audio hardware. if exclusive mode is used, streaming sounds exactly like the CD. if non-exclusive mode is used, streaming sounds muddy and lacks dynamic. it is a night and day difference.
The problem is always the same. Jitter and electrical noise. I used filters, audiophile switches with clk and in the most cases they clean the sound, better silences, better dynamic but always seamed to be lost some naturally, musicality, until i hookup my server directly by optical fiber the sound it started to have a much greater cleanliness with some silences like black holes. With this experience, I came to the conclusion that I was doing bandages before and now I have treated the disease. The optical fiber, being immune to RFI and EMI, is not affected, taking the signal directly to the server always clean. Today is very difficult to me distinguish the streaming from flacs or wav.
I really think it all comes down to system synergy to get a good streaming setup to sound great. It took me a bit but I found what I wanted by trading out a couple DACs and Streamers to find the right combination. Now Qobuz sounds better than CD in most cases on my current system. Even the correct USB cable makes a pretty big difference. All about system synergy and what sounds good to your ear 😊
@@davidgoffin1537 I started off with a Bluesound and then tried it with a R2R DAC. That DAC failed and I wanted a better streamer so ended up with Gustard R26 into a Zen Stream with external power supply. This combo is perfect for the type of sound I like.
I am pretty sure there is ZERO compression in transfer or storage on Qobuz, just the lossless file provided by the artist. At least that's what I gathered from David Solomon from Qobuz (who also used to work for tidal).... I know MQA is compressed and Lossy. My thoughts. ALSO - Ensure you are listening to the SAME Mastering.
There are 2 factors to consider. 1. The fidelity of the data. Is it bit perfect from cd transport, a network device or a commercial streaming service. The first two are identical. The streamer may be using a lossy compression or volume leveling 2. The quality of the 'transport' which determines timing jitter and noise levels that accompany the data. The 'AirLens' or other devices will supply the data with less noise and a given phase noise. Given a bit perfect data stream, the data stream should be better than from a conventional CD Transport given your DAC doesn't have the same function. I rip cds to a network hdd and play using a streamer with very good galvanic isolation and a SOTA clock. Playback is far better than from a CD player or transport. If I access a streaming service and find inferior sound quality, then it is the fidelity of the data that has been impacted. AirLens as Paul describes it would be similar. But it cannot correct a data stream that is not bit perfect as supplied from a streaming service.
Paul, as I was listening to your explanation of storage and transmission, could this be the real reason behind MQA? I'm a Qobuz user, so I don't have a dog in this hunt, but it does raise an interesting idea that as time goes on and music services are pushed to store even more recordings.
I'm not Paul but yes storage costs. Its not just an x number of petabytes in 1 server suite somewhere in a datacenter. That data needs to be redundant so there is a backup cluster. Making it 1.5x the cost. And of couse, its not only 1 dc but multiple around the globe (bc. uptime/failover). So again 10x the cost. And of course every file has different quality versions. So again x-times the costs. And then comes data transferring between the datacenters and eventually the transferring towards the customers. Anyway, every MB you can save on file size is big money over a longer time period like 1 yr.
@@the_dude182 Hi, yes that all makes sense. Originally, I thought MQA was an interesting idea for Europe where unlimited data plans weren't as widespread as in the US. Occasionally, I see albums deleted from Qobuz, and when I factor storage that makes sense. As a user, I really don't want to be bothered with MQA. I have an iFi Signature DAC (v1) that sounds much better when I change the firmware to the non-MQA version. I listen to much more music because of Qobuz than I ever have in the past.
This may be subjective, but I don't think streaming service matters. It may have more to do with implementation. For example, SACD/DSD on a locally implemented hardware based system should be the choice for all. To test streaming,I directly compared Qubuz to Tidal and found MQA was more to my taste. I got a similar level of the thing audiophiles would look for: soundstage and imaging and details never heard before, but my implementation is specific to me. To lean in more on this topic, I would like to take Paul's suggestion on using a hardware cd player and compare that to local file storage and retrieval methods using network and computers (which is what I suspect maybe the limitting factor) to see if there are differences. Update a/b direct CD transport to streamed local file via Roon. No noticeable SQ difference (good for me as I can be happy with my method of storage). However, a/b/c test cd player vs 44.1/16 bit hifi vs tidal Master 44.1/16 bit. Master had the best SQ, 2nd best was cd player, and the tidal hifi quality track sounds the worst. As always, I'm never really done, but there seem to be many dynamics when comparing.
I dont THINK anymore Paul says buy a Mcintosh so thats what I will do. Thats the great thing he is Honest why push his own stuff i mean just look at the speakers in his house.
In the early days of computer audio, there was a holy grail of Pure Music version. It only ran on an older version of Mac OS and if you upgraded the Mac OS, it broke that particular release of Pure Music. That version sounded amazing. I stuck with that version of MacOS (not upgrading/updating) for as long as I could. The developer of Pure Music was asked about what he did - and from memory, he said he worked on power optimisation - reducing power spikes. It was a monumental effort and I don't think he ever tried doing that again. Coincidentally, when the CTO of ESS was talking about how his company transitioned from computer audio (with great SNR) to audiophile audio, he found that in the blind tests he conducted with various engineering samples, the audiophile designers always chose the versions where the various states in the digital-audio conversion had minimal power spikes as it transitioned from different states - and not how quickly it reached the final state. On my own testing, many years ago, I had a Wadia iTransport that you could dock an iPhone (then a 3GS or 4?) to, and it would output the digital audio files through a SPDIF output that you could send to any DAC. At that time, my DAC of choice was the Ayon CD-5S which coincidentally had a built in CD transport (I2S). I played the same file (ripped from a CD) through an iPhone on the iTransport and also directly through the built in CD player. The CD player won. But on the advice of a friend, he asked me to rip it in WAV. I had previously used Apple Lossless. This time, the WAV version got me a lot closer to the CD. Everyone told me that converting from FLAC to WAV in real time took so little CPU cycles that it couldn't possibly affect the sound - but maybe there was just those transient power spikes, however short, that made a difference. I remember the original Digital Lens on the old PerfectWave Bridge used to implement something similar, didn't it Decode to WAV ahead of time and put it in a buffer to be sent out. Is something like that in the new Air Lens?
One thing I notice is that any song I 'stream' has a particular quality but when I download that same exact song and play it from the download it sounds noticeably a lot better. All my settings are set to stream and download at the highest quality available and I have very fast internet. Not sure why there is such a noticeable difference but it seems to have to be in the 'streaming' process itself.
I noticed this with spotify, the down loads sound way more dynamic and punchy compared to streaming the same song, even tho i have it set to stream on highest quality and download in hi quality also.
"Streaming" implies a continual stream of data. It means that "perfect data" is not the only goal, also it has to meet the non interruption goal. When the transmission is less than perfect, it may drop data or lower the quality (ie bitrate). All of this is to ensure instant and not interrupted playback. Otherwise you would have to wait that the entire file is downloaded to be reliably played. Also there is the problem of what CoDec is used, what source was used, what format sound is stored, etcetera. I don't really know if some of this applies to high quality streams offeredby some. DSD streaming is another story.
@@Harald_Reindl You're focusing on bandwidth. You should be focusing on checksum, or lack there-of. Streaming has no checksum so, even if the bandwidth is monumental, OP is correct that some of the packets of data can be lost.
@@stevens1041 are you drunken? I earn my money with IT and if the ethernet layer would randomly drop packets music would be the smallest problem! I can safely transfer gigabyte large virtual machines where a single bit-flip would be fatal and you clown pretend your music will be damaged? The problem with you audiofools is that you lack any education and technical knowledge
@@stevens1041 Spotify instead uses TCP. Firstly, having a reliable transport protocol simplifies protocol design and implementation. - Oh and when you are using UDP a trained monkey can make sure data integrity and will do so
" All of this is to ensure instant and not interrupted playback. Otherwise you would have to wait that the entire file is downloaded to be reliably played." If you have an Internet connection which deliveres data between 200 and 300 KBytes per second and for playback you need 150 KBytes per second you not need to wait for the entire file being downloaded. If your Internet only delivers 160 KBytes in average but transmission rate is varying betwen 140 KBytes and 200 KBytes the software should wait until the buffer is filled with some amount of data to make the chances high that there are no interruptions in the playback.
Playing back CD is like unpacking something that was delivered from across the street VS. streaming is like unpacking something that was delivered across the country via going thru different carriers and repackaging processes in between. Even both packages come from the same source, you can still likely to tell a (big) difference if you have sharp senses (ie resolving systems).
Nonsense - if that would be true files which don't survive a bit flip would randomly arrive corrupted and in that case every compressed and encrypted file transfer would be impossible
Hello, I master records for a living. Files digitally captured during playback from Tidal of masters that I've done have nulled perfectly against my house archives of the same masters. So, AFAIK, the data is the same as my source. Any differences in sound would have to be dependent on the end-user's system or network.
What if we rip the CDs? And we “stream” them from our NAS or from an attached hard drive? Does it still have noise? Is so, how do we play music files for the best sound? Maybe using something like the Eversolo dump a8. It can have an internal M2 NVMe SSD… would that be better than any thing that is using Ethernet and/or usb connections?
many commenters here show their ignorance on digital processing. All digital networks transfer data absolutely perfectly, no matter the route or distance. This is essential or so many of todays technologies would be impossible. If the CD and streamed versions sound different it is probably due to different Masters. Is your CDs sound consistently better than CD quality streaming then you have a setup or equipment issue.
Part of the issue with streaming is the way the packets of data are transmitted across the various hops (switches, routers, etc.) they traverse on their way to your computer, and how your computer handles those packets when they arrive. When you download a complete file, the packets are normally transmitted using TCP, which dictates that your computer check to make sure the packets are in order and none are missing. If there is something missing or out of order, the computer will rearrange the packets and/or request the server resend missing packets. Because of time restraints, streaming and audio/visual communication packets are sent over UDP, which means your computer receives a blast of packets that is often many fewer than the server sent (packets are frequently dropped at each hop the data traverses), and sometimes they arrive out of order, and your computer does what it can to rebuild the data (as you are listening) with what it receives. There is less of an issue the closer you are to the server (fewer hops to lose packets on), and the better your internet connection is (more bandwidth and faster data retrieval). Hopefully soon, routers, switches, servers, and networks will be more efficient, internet speeds will continue increasing, and dropped or out-of-order packets will no longer be a concern for music streaming. In the meantime, simply downloading the .flac files to local storage should make a huge difference.
Oof my friend you are only partially informed. The switching has nothing to do with the audio quality as you should know. Also streaming is buffered so generally this is not an issue
"Because of time restraints, streaming and audio/visual communication packets are sent over UDP," That is not correct. Time restraints are a problem if you want to transmitte audio or video close to real time - for example the live transmission of a sport event or telephony over Internet. Streaming is not real time - the data is always buffered before it is played and as long as your Internet access is faster than needed for the playback there is enough time to request a resend of defect packets or to reorder the packets in the correct sequence. Many streaming provides use TCP/IP - and even if the use UDP that does not mean that there is no error handling. The difference to TCP is that the application have to decide on its own how to handle errors.
I think there's multiple detrimental effects that streaming has. One is the timing in music is extremely important. Networks aren't set up to have it occur at 44khz. Never any reason to. Speeds vary and interfere with it. The other is error correction that used an algorithm to 'guess' at what sound to make when it senses a corrupt bit. You don't hear it skip but that doesn't mean there's no error. To bad we can't shut off error correction on cd players etc. To then realise how often it goes into error correction. I use a NAS and don't stream from the net but for you tube on occasion. Even local flags don't sound as good as a cd.
"One is the timing in music is extremely important. Networks aren't set up to have it occur at 44khz." The Internet does not care in anyway about sampling rates of digital music. The Internet transmit packets containg binary data. It is like downloading a e-Book with 1000 pages (which takes only a couple of seconds). The timining of that download has nothing to do with the speed you are later reading this book. Streaming differs from file downloads only in the way that you are able to start reading the "book" already even if the download is not completed yet. After the first pages are transmitted you can start reading - and while your are reading more and more following pages are transmitted in the background. "The other is error correction that used an algorithm to 'guess' at what sound to make when it senses a corrupt bit." There are different methods to handle errors. TCP/IP requests a resend of a packet if it detects that the received packet has errors (which is done using a CRC checksum). Audio CDs have instead additional bits which are allowing to correct such errors "on the fly" - only in the case there are too many errors which could not be fixed any more some "guessing" happens.
@@thomaswalder4808 so why does local always sound better? If it gets a bad packet error correction kicks in. Streaming is asynchronous. Doesn't ask for it back. Can't. The packets work on the data being computer correct. Not audio correct when it's recived. It's a digital signal passed on an analog square wave form. That's subject to jitter and timing delays. The end unit resolves the errors. What you are saying is there's no need for error correction at the dac. Huh?
@@steveaustin7306 "so why does local always sound better?" You believe it sounds better because you expect problems with streaming. That is known as expectation bias. Think about you have two Big Macs in front of you which were made at the same time - but somebody tell you that the second one was made 3 days ago and stored without cooling and heaten up again in a microwave. Which one would taste better? "If it gets a bad packet error correction kicks in." Correct - but your understanding of "error correction" is wrong. In case a bad packet is received the streamer detect that it is bad because its CRC-checksum is wrong. In this case the streamer requests the packet again from the streaming server. Where is no "guessing". "Streaming is asynchronous. " Correct - therefore streaming always needs a buffer on the receiver side. This buffer is filled with packets arriving from the Internet and readen synchronous by the DAC for playback. A normal Internet connection could deliver the packets much faster than needed for playback - so this buffer is normaly always full. In case a bad packet is received there is plenty of time to request this packet before the buffer would be empty and the playback must be interrupted. In the early days of the Internet where most data was send over analog telephon lines such bad packets were common. Today it is extremly rare. My Router keeps statistics of bad packets - I not had a single bad packet in over one year.... But as described - even if that occur it does not impact the playback at all. "Doesn't ask for it back. Can't." Why "can't"? In TCP/IP it is unavoidable that bad packets are requested again. "The packets work on the data being computer correct. Not audio correct when it's recived." Receives packets are either correct or not. The CRC checksum used to determine if a packet is correct or not does not distinct between audio data or other data. It even does not know what type of data is in the packet. "It's a digital signal passed on an analog square wave form." There are many ways to encode digital data in an analog wave form. In the very early days sine waves with two different frequencies were used. A sine wave with one frequency was used to encode a "0" - a sine wave with the other frequency was used to encode a "1" Ethernet uses different methods like "Pulse-amplitude modulation" or "Quadrature amplitude modulation" - these do not look like square waves... But that is not the point anyway as long as the digital information could be extracted from that analog wave (however it looks) in most cases. If it can't then packets have to be retransmitted... "That's subject to jitter and timing delays." Yes - and therefore a buffer is needed before the DAC which get rid of all that jitter and timing delays - and even packets arriving in the wrong order. And every streamer has such a buffer "The end unit resolves the errors." Correct - so what is the problem? "What you are saying is there's no need for error correction at the dac. Huh?" That depends on what you understand by the term "dac". Normally a dac is just a digital analog converter which has some digital input lines (for example 16 ) and converts the value represented by this digital input lines to an analog voltage in a defined range. In the HiFi world DAC often means not only the Digital to Analog Converter itself but also some additional processor to convert different digital audio encoding standards and protocolls (like USB or SPDIF) But yes - there is no need for error correction of "bad packets" as this is already done in the streamer which also fixes "timing delays" and "Internet jitter" using a buffer as described above.
I use Logitech Media Server + Squeezelite (running on PiCorePlayer) as my music streamer. I find that the settings of streaming buffer makes a hell of a difference to the resulting sound quality. I use 64000Kb stream and 192000Kb output and I find that the difference between locally stored flac files and tidal (non MQA) virtually vanished
Qobuz to me sound very good with jazz which I tend to listen to 95 percent of the time. Can't tell any difference frm CDs. But once you go popular music things begin to fall apart. Some music sounds better on CDs others on Qobuz.
Streaming can sound as good and better than CD. It requires a fair amount of investment, but it can be done. Given the number of steps involved in getting a streamed file to a DAC, I'm surprised by that. But in comparing 16/44 streamed files to the same CD's, my ears tell me that the streamed files sound just as good. PS. That's via a high bandwidth wired (very important) ethernet connection and a high quality bridge.
Huge conflict of interest for those depending on 'music on plastic' to make a living to defend 'streaming' and vice-versa. Thus, abstract opinions matter not, only the audio experience of an individual. I dont make my living of music, and like both formats. Music on plastic because of its physical/nostalgic/ritual component, streaming (via a Node) because of convenience. The TIDAL masters sound extremely good. As good as my best Music on plastic stuff, which includes some UHQR stuff. Voilà my truth.
Qobuz lets you download and store their files offline (Tidal as well, but in the mobile device app as far as I know) - when I playback these offline files, my DAC claims to be reading the same bit rate etc. off the file as advertised by Qobuz. I do this wired (on a short path) - of course, the files originally came to me from the Qobuz servers and data stream. Is my DAC being fooled when reading the resolution of the offline files? One thing I did notice, when I downloaded all my identical Qobuz and Tidal playlists as offline files (in the highest res offered) to my phone, the Qobuz library amounted to many, many more GB of storage. I don't know, maybe this is because Tidal hi-res is MQA and that packet is smaller than Qobuz FLAC? Either way, I thought it a bit suspicious as fair number of my playlists were at CD resolution only as hi-res was not available for some of the artists I like on either platform (or the original recordings were early digital at 16/44) .
That makes sense, which would explain why few services actually offer FLAC and HD Audio. However, whole streaming wars, is the same as video wars... It also really depends which equipment your listening on as well.. No one will notice if your listening on a pair of Apple Earpods, but on a high end systems, you would..' It's no different than a playing an mp4 video,, with 1.6 Gig compressed file size.... No one would notice on a laptop or a 30' inch display, but will be pixelated on a 50' inch.. It depends on the user. that's all it is... This is why i don't let anyone tell me which is better, because its more simple than that.. Personal choice only.
Hi Paul, you just hitting the nail, let saying there is B3 CD player and Ripping, WiFi, etc. This B3 engineering and manufactured in UK by Martin Brennan. Should you be able to give more information about this B3 CD player. Are they compression file as well? Compression is the files we compressed to smaller size which we lost the original size- as well sound quality lost.
I use my PlayStation 4. Extremely cheap I know, but it can read FLACs, you can plug your favorite USB DAC onto it... Plus you can use Spotify and UA-cam if you want to. It's not the most silent thing in the world, but even with open back headphones I can't hear the fans spinning
if you dont hear adifference, its time to get your ears checked. I never stream music. downloading in 320kb mp3 or FLAC/WAV yes. but the most of all i prefer even my Vinyl albums (with good dynamics, not every label has, like "cough" CBS) above my cd's. they also good tho, but if you can get an album in ufcd MSFL ultradisc, i recommend those to buy if you like good dynamics
I'm wondering if the loss in quality is during streaming from a service (qobuz), or is it in the files they storage itself? in the first case the sound quality would be good if you download/buy the musicfile from the service! anyone knows more about this?
Equipment differences, streamers vs cd players, internal power supplies, caps etc, then DAC’s internal vs external are among the first layer of controllable differences, so it is hard to compare apples and oranges. I find even if I stream and upsample to 192Khz using the dac inside my cd player - an ARC Ref CD9 - the cd sound’s obviously better in every way. I’ve enjoyed streaming well recorded tracks and they sound good but cd’s sound so much more dynamic.
Interesting video! not sure it would be ethical if streaming services that claim to offer cd quality cut costs in that manner. I lately put much of my music in lossless ALAC format on my iPhone, which I airplay to an old apple tv which feeds it into my amp's DAC. Don't have a CD player so this is the best digital experience I can get. Any thoughts on this method? Would going over wifi and through all these devices somehow deteriorate the quality of lossless audio and would a cd player improve my listening experience?
The biggest influence on the quality of music, if the master copy. Someone once told me, that when a track is released multiple master copies are mixed and distributed, with each mix specifically catering for it’s intended recipient, be that cd, vinyl, radio or 1 or more of the many streaming services. So, when you hear differences between the same track played from cd or a streaming service, it is likely the difference in the master recording you are noticing!
While I don't really consider myself a true audiophile (my systems are resolving, but not to the point where I can discern huge differences), I can definitely tell a difference between listening to my original Brothers In Arms CD on my 90's Denon versus a 320kb version of that album streamed. Granted, I don't have the best D/A, but you can tell that there is a lack of dynamics with the streamed version versus the CD, no doubt about it. This example is only one of many. As I said, there is not a HUGE difference, but if I am in a "listening" mood, it is CD every time without question.
Speaking of streaming's short comings...on paper flac and the original wav file look the same, and on null tests (which I've done) they null, but flac lacks some high end. There's something that the null test cannot measure, but it can still be heard and felt. Not what many people want to hear, but it's the truth. People who suffer from hearing loss cannot hear the difference. Too bad for them, they're missing out. Many prefer meta data over sound quality.
If you have RCA inputs in your amp ( main in) which is direct amp signal , you can put outputs 6.3 mm banana from sound card which already have DAC preamp into that rca main in amp inputs and you WILL get better sound quality from spotify or youtube .
A little fun with the KEF driver in the background since I watched a video from the Norwegian hifi magazine "Stereo+" youtube channel wisiting the KEF factory😝
A lot of the differences people hear is in their equipment or how they have things connected up.Your streamer DAC and cd player DAC are not going to sound the same.Try running your cd player through the same DAC using it as as a transport.You may be surprised.
So many variables introduced when using the streaming option, from server storage to bandwidth differences. The vast majority of listeners will be happy with convenience, at the expense of ultimate quality.
My wife is the least audiophile person in the world. As long as she likes the song, she really doesn't care whether it's played on a kitchen radio or on a high-end system. I recorded a mix tape for her for Christmas. All songs came from my streaming service ... except for one that wasn't available online in the desired version. I recorded it from CD. My surprise was great when, while listening to the tape, she said: "That song sounds kind of different than the others. Somehow… better". What can I say?! If even my wife hears it, then there is definitely a difference in quality.
Ive been married for 35 years and my friend, I hate to break it to you because someday you'll find out...
Your wife hears EVERYTHING.
I agree that humans, which don't overthink anything still hear everything effortlessly :-) However, only one song is a very small sample. It could have been the song itself.
Very interesting! Yes, it certainly DOES say a lot that she still noticed a difference as well.
For me, it's become virtually impossible to notice a difference b/t streaming (not sure if was 320 mbps mp3 or FLAC now) and downsampled from FLAC AAC downloaded files in terms of the quality of the track itself. HOWEVER, I still DO notice a difference in that the streamed version has a noticeable "hum" or buzz with it, where the same track (yes, same hi-res version of the same album) DID NOT have this issue. So that by itself still makes the downloaded version superior to the streaming version in my opinion.
@@barney6888 🤣🤣😂🤣
My wife is the least audiophile person I know. She despises my hifi system. I bought her a little DAB radio a couple of years ago and it is her preferred listening choice. She is equally happy watching a movie on our 32 inch LCD tv in the spare room as she is watching it on our 77 inch OLED in the lounge!
Only speaking about Tidal, I turned off "normalization" and noticed significant improvement in overall quality of sound.
Yes....and turn off the battery saving features of the phone and also WiFi. Download all your Tidal songs and listen without streaming....there you go
The same on Spotify, thou it is compressed. Normalization should be off.
@@KentAndersson-ig9qd what is normalization? Is it like replay gain? I've never used streaming music unless absolutely without a paddle
I have always enjoyed and learned much from your videos. I am 70 years old with Tinnitus and VOR…so as a person I can not be that discerning audiophile has long gone, but I still am a serious and discerning listener. . I started listening to music as a love and passion since I was five, my dad was really into it and thus I never lost that love. I am not going to come across as any expert here but the gentleman’s quandary can come from so many sources or combination of sources. I spent 40 years in IT and communications. Yes, a network, shoot your 110v power source, neighbors and your wall chargers, phones, washing machines, ceiling fans, LED lights etc can add significant noise to any signal. But to think todays high speed internet is 'the' culprit to adding a significant or even perceived degradation in quality is a stretch. Yes, in a way it can but it would not be in such a way that the fidelity has been compromised in any meaningful way. There are exceptions but modern technology has so many error corrections the actual loss of the bits/data along the way would be rare or make that difference be rare. Again, most interference after it has arrived to your streamer will most likely be local. As a long time HAM radio communicator I watch this stuff on my equipment RF signaling equipment. Today in the electronic world we are surrounded by unimaginable man made RF noise pollution. A $7,000 radio and $7,000 amplifier Ham radio quality is measured on its ability to detect and remove this garbage. I know when my neighbor three houses down turns on her cheap ceiling fan. I can tell when the relays in the washing machine turn on to squirt water in the washer. I have to unplug my wireless desk side phone charger. Just touching the surface here with the RF garage that can and if you have Amazon cables or noisy AC for power, it can truly add to the loss in fidelity of your sound system, much more so than the ‘Internet’. But just saying it’s the music being streamed over the Internet causing the problem, if you have a good high speed service, would be the last place to look.
Someone commented on this vid about the source of the master or remastered. I personally feel that is one of the largest issues or maybe just differences between what is streamed compared to CDs or remastered LPs etc. I find and hear this difference (having CDs, LP and Quboz). Sometimes Quboz sounds better than an excellent condition original LP and vice versa. Comparing my Dark Side of the Moon album that I got in I think 1971 is superior to the the new remastered I got off Amazon last year. That is literally crap!!!! But Quboz Dark Side of the Moon is fine. But other original release LPs sound better than Quboz and again, some do not. So I think where the master or remaster came from plays a big roll here.
Second, DACs… The things that turn our digital into analog audio. I know you have far more experience than I do for sure but I have three and what a massive difference they make on streaming from so called high fidelity digital streaming sources. I use a BluOS as the streamer, but I do NOT use their DAC. I bypass the BluOS DAC. I have a really nice DAC built into my Anthem preamp , I bypass that too. I go straight from the streamer to the REM DAC and analog from there to the preamp. This is like night and day compared to using the DAC on BluOS, I mean very large difference in fidelity. The Anthem DAC is superior to the BluOS DAC…actually propably good enough for the majority of us. But it was not until I listened and configured the REM for my ears. It is not massive but it surely made a difference. So , as you know DACS play a big roll here.
Great topic but there is so much going on here from where did the master come from, what DAC are you using, how good is your cartridge, man made RF interference everywhere, hearing loss.
So in conclusion for me, I can not say any one source is superior or less than the other. As long as you stick with a NON-WIRELESS streamer (yes that can add substantial noise to the fidelity) and your source is hardwired to your router, you stick to, pick your flavor (I use Quboz), a good streaming source. Sometimes my LPs are better, sometimes Quboz is better ( I honestly gave up on CDs though I own them)..
I will say this for sure at 70 years old and listening to my $30,000 of stereo and custom acoustics installed (wow acoustics that made a Hugh difference more than anything) is that in my 20s with far less superior audio equipmnet , it just sounded better back then haha. Our most important ability to comment and make comparisons in fidelity comes down to preference and how good or bad your ears have/will become.
Thanks for that post. We have some remarkable similarities. I too am 70, I also got my love of music from my dad, I too had a 40 year career in IT, and I also was a HAM radio operator (having built my first HeathKit with my dad for my 10th birthday). Very cool.
Thanks, I do agree in many points you mention, specialy about hearing loss. But my conclusion is that the problem with audio is created in 90% not by loss of hearing which affects top band, but it is in basic band (which in my not only opinion but also experience) is not exceeding 9kHz. That band is still easily heard in my 70 ties - I can still hear up to 12 kHz. With "additional" 12- 20 Khz we experience tones which are "emerging" as supplement of not good quality in basic band. It is because top band has very mixed phases and are easy to produce for tweeters . If only ears can do it even they are hardly linear (on reason of perfect reflections and games of phases) they are usualy easy percepted by young ears
. Hearing in that high band makes false confidence that everything below is presented and creates impression of very detailed presentation. Even hughe ovation from gig sounds very elegant and distant but I say is supposed to sound as annoying shout. -
If that top tones are not in recording (due to kind of music or mastering), band between 1 - 9 kHz is in parts damped due to different reasons then listeners have no such "supplement " in top and doubt about quality of whole recording or they think about their hearing loss. Beacuse It sounds simply dull.
Reasons of bad quality I account mostly to hidden resonances in reproduction channel which are source of phase no coherence resulting in compensation by themself or their harmonics to zero as well as overlaying by louder resonant harmonisc.
My experience is that with use of widerange capable in 30- 18 kHz speaker and simple SE mono amplifier I hear stronger and more exciting sopranos than with 2 way HI FI supplemented by electrostats and exceptional class B tube amplifier. I try many ways to make them equal but so far it is only enough near for not being worry about my hearing loss. Now I decided to build another better equipped stereo 2W SE tube amp to make better research . I could buy one but for me it will be more fun
I have an arcam avr20 with a couple of matching power amps running a 5.2.4 configuration and I would just like to say... coming from A 2-channel to A multi-channel setup, I couldn't believe how much my noise floor had risen. Anyway the more cables I shielded the quieter and my room became!
One's got to remember unshielded cables act as antennas for attracting RF and the longer the cable the better the antenna! Probably explains why when I used to disconnect my Heights and Surrounds the noise floor noticeably came down. The last group of cables I shielded with my HDMIs and low & behold the room became slightly quieter yet again.
👍
👍
The streaming sevices are all about LUF's ( loudness units full scale), so the music is measured and altered for volume. There is no standard dB max or min across the services. In my view this is what people are perceiving as inferior sound quality as compared to their own CD because they are messing with the dynamic range of recorded music.
According to the literature, all the LUF thing changes is overall loudness, not DR
I listen to a song on a cd then the same song on apple stream To me the sound is the same just the volume is different.
From what I understood understood about lufs, and loudness normalization, is they're not changing the dynamic range. They're increasing the peak volume levels so that they're hitting 0 DB, or saturation, for the highest maximum volume. But the overall dynamic range is the same
Followed Paul’s advice, I added a new CD transport to my high end DAC. The quality of my CD collection played on this is WAY better than any of my streaming services (TIDAL and Apple Music). The detail within the huge dynamic range is extraordinary.
May I ask which one you got?
Lol
@@johnroady9495 It is the new Audiolab CD transport 9000.
Followed Pauls advice went for Mcintosh and your right the lure of Blue VU meters is simply the drug we all need why deny it LOL
@@josefbuckland i wonder why audiophiles accept such meters when they otherwise all day long cry "everything which isn't strictly needed has negative impact to the sound quality - even stuff which is completly bypassed just because it's there" 🙂
You speculated that the 'significant reduction in quality' (the letter you read) may due to streaming vendor's compression of their files. Well, why don't you produce evidence? I did a test, and found a 70 min ultraHD album downloaded from Amazon music unlimited takes up 1.35 GB of space. That is a massive file size! I just hope you have data/fact to support your claims.
Hitchens's razor: "What can be asserted without evidence can also be dismissed without evidence" .
I remember those albums of Nirvana on Qobuz. I own the original CDs and my GF the remastered versions.
Some years ago Qobuz only had the remastered versions online. I compared them with my GFs ripped CDs and every single bit was identical.
Later Qobuz added the original masters. I compared them with my ripped CDs and every single bit was exactly the same.
If there's a difference, it's either
- a different master (like mine and my GFs CDs)
- played in non exclusive mode (on Windows the internal mixer resampling)
- played over a different DAC (CD has to be played digitally to the same dac !)
- played over a lossy transport (bluetooth, ...), or a noise transport (jitter, ...)
I never experienced lower quality with Qobuz compared to my own collection and I consider my hifi system to be quite hi-end (focal sopra 2s here)
Tidal (mqa) is something else. A 24 bit mqa is 15 bit audio (+1 bit mqa signal + 8 bit lossy compresse hi-res estimates).
A 24/48k mqa decompressed to a lossy 15/96 max. The rest is upsampling to the maximum the dac can handle and adding nothing.
The number on the dac display is meaningless too, as it shows what the master was BEFORE it was reduced to mqa.
So if you hear mqa sounds worse, that's why.
(But I don't have to tell you Paul, I saw your meaning about mqa 🙂)
As a mastering engineer, I'll add that CD and Tape are truly inferior mediums. Tape being the absolute worst. DSD being the absolute best but hard to process.
I agree with what you are saying here Paul, and I don’t have anything to add on that issue, but also I’d like to bring up the fact that you could have or anyone actually could have a particular music CD and depending on which version they have, an original, a remaster, or even which remaster, and when and where/how, they’re not always all the same in terms of sound quality and, in my experience, a lot of times they didn’t make it better .. that’s another annoying thing to think about.. it’s just sad that a lot of the people don’t have the equipment that allows them to hear the difference, but for the people that can, it’s a nightmare!
I truly enjoy your vids. Thanks for taking the time and energy to share. Here’s hoping ‘23 is a banner year for the McGowans.
I think it does come down to the source material rather than streaming per se. I have some CDs that sound marginally better but with other titles the streaming sounds better. I always noticed a reduction in quality (usually soundstage and dynamics) if streaming without the Audirvāna program which talks straight to the external DAC without any computer OS port handling to get in the way. I find Audirvāna an essential piece of the streaming jigsaw.
...and Vinyl sounds better still.. It's a cascading-effect. I think i can understand why these two are returning.
@Tech-geeky that's just not possible. Tape is the least dynamic medium. When you master for Tape you need to pretty much cut everything under 45 Hz and a shelf under 200Hz.
Low energy information would make the needle jump. The same process is used for the highs at a lesser extent. The transients are also more compressed. You just prefer the sound of Tape. It's not better.
Bullshit
Paul, just thank you so much. For all you do, your stories keep me alive.
And they are making you dumb
@@Harald_Reindl No, they are literally keeping me alive. If you knew my life story, you would understand.
@@thomaswachter7782 well, you can be happy after someone made you dumb - 10% here is quality content and the rest idiotic nonsense
@@Harald_Reindl I never said I agree with everything on the channel. I said his content keeps me alive. That is all I have said.
The largest incremental improvement in what I hear in my music from all my individual equipment investments came from taking my favorite streaming music on Qobuz and buying the files from Qobuz and putting them on my server and playing them wired directly end-to-end (without WiFi, Bluetooth, IP, LAN, WAN, Internet, etc.). Further, as a former top exec at a global publically traded data company, I share your view that if it saves storage or bandwidth costs, it happens, no matter how well hidden or adamant the claims are to the contrary.
If you don't mind, could you expand on that? are you playing the downloaded files from a laptop wired to a receiver?
@@andrehendrikyou can do this on a desktop as well.
I have always wondered why even a basic cd player sounds better than my Tidal. Thanks for bringing this up
Its placebo. There is no difference. Paul even says that cables make a difference witch is proofen by many people is wrong.
It's quite true. A cd player, based upon many many factors in respect to its build and components, will almost ALWAYS sound better than any streaming. It's proven and it's a fact.
My system: Auralic Aries G1.1 streamer ; Chord M-Scaler - Chord TT2; Music First Audio V2 Classic preamp; March Audio P421 Purify monoblock power amps; Martin Logan ESL-X speakers.
I had a Cambridge Audio CXC CD transport for a short while but sold it as it sounded worse than locally stored flac and wav files stored on a usb flash drive connected to the streamer. I just can't hear any difference between the locally stored wac and flac and streamed 44.1/16 CDs from Qobuz; streamed hires files (in general) have an uplift in SQ compared to the flac and wav files. Interestingly CDs ripped via EAC sound better compared to the Auralic inbuilt CD ripper. One other (very small) improvement resulted from the inclusion of a iFi iPurifier 3 connected from streamer to M-Scaler.
I may consider trying a better CD transport on a home trial basis if there is a prospect of wringing out a bit more SQ from a digital source; however, until I can hear an appreciable improvement I'll eschew the claimed superiority of the silver discs for now.
First UA-cam channel I'm watching in the New Year. Always worth it! :)
Dear Paul, I am new to audiofile world; your videos provides a source of learning and valuable insight. Many thanks!
I often find albums on Qobuz that have been remastered and sounds much better than my original CD, a good example would L.A. Woman or Monty Alexander live in Montreux (1976) very well recorded
maybe I should purchase the download to compare but my current set-up (no hard drive server does not allow this test). I understand this is mostly due to earlier CD not so great quality rather than answering the question yet I’m really happy with the overall Qobuz sound quality (Wilson speakers, Moon Audio streamer dac)
Analog to digital and convert them back from Digital to Analog. They will has some lost/ distortion on the dynamics.
I knew it!! Glad you're exploring this and great question posed by the viewer - I've exactly the same experience. All things being equal, playing via Tidal doesn't sound as good as playing my Flac's. I'm not disagreeing with anyone else's suggestions (I've tried download and normalisation off etc etc) - my experience is I can't get near my Flacs. My system isn't even particularly high end for the community. This isn't a question getting a better hardware, cables or room treatment - it's a question of 'What am I paying Tidal for' (library aside).
A couple of controlled tests: 1) The Tidal app for Android has a download feature, where one may download the particular music for later offline playing. One can then do listening tests streaming it vesus not streaming of the exact music track. 2) I have also done direct A to B comparisons of TIDAL streaming vs CD files that had been copied to my local laptop. I could not tell the difference for this particular music, but I will try additional tests.
Website try and do the same... "hearing tests" : They take a track in different bit-rates and/or HQ in some cases if you have the equipment to hear it on, and lets the user judge for themselves.
Its getting crazy... No one knows how to use tech, our to protect from malware, Now they are shoved into this audio-phonic space ? *walks off*. If CD's were cheaper to offshoot the ever increasing streaming services catalog, you bet i'd be getting my music on compact disc..
At least you can say you OWN it.... Streaming services, you never own it... People still think in 2023, they own it, but all you gotta do it look at the arguments when it happens.. What does that tell you? No one really understands do this day. Its a never ending cycle which will never understood.
And we call that better ?? All were really riding on in 'convince' and 'price' That's it, because it defiantly is not about quality...
@@Tech-geeky what on earth is your point
Always use wired connection for streaming. Never ever use wireless for streaming.
Happy New Year to you Paul and everybody ! That's why I keep collecting CD, SACDs, DVD-Audio, Blu-ray Audio ! Cheers !
FLAC (8) level lossless compression should be fine. I did hear of one musical artist who used Deezer. He uploaded his songs and then later downloaded them and found them to be perfectly identical. What other services are using is up for question but many think Qobuz sounds best. Tidal and Apple Music are ranked about the same.
This is one reason many "audiophiles" say you have to spend $$$ on your digital front end to get the same sound quality you can get from a $500 (and up) CD player.
Some even go as far as saying you need a better router too, use wired Ethernet if you can, and add better, quieter power supplies to all your "digital front end" devices.
Jitter should be taken care of by any quality DAC with a buffer and its own internal reclocking, so that isn't (or shouldn't be) the cause for any sonic differences you hear.
It is an interesting observation though, that many people make. Streaming sounds "good", but a local CD played on your own player going through the same DAC can sound a touch better. Hopefully all that will be better ironed out in the next few years as streaming becomes the de facto method of music listening, except for dyed-in-the-wool audiophiles who insist on CD and vinyl albums. (I like both just fine - and the great music discovery you can get from streaming). Heck, I still play many cassettes and love them.
_Why streaming music is lower quality_ it is only when a different and lower quality conversion is used. If you stream HiRes then the quality is as high as it can be
Me too. It's a big difference for me as an audiophile, when i play dvd format and swapping it with streaming serv. (both use rca) there's something missing, the dynamics, the lowest and the highest frequency. The bass in dvd format is more ample, and in the high frequencies you can really the ping in hi hats and it's more life like
I have very little interest in streaming. As someone that has accumulated a vast library that I keep on my hard drives, I would, in a minute, go for an integrated with DAC that takes input from a dedicated SSD to feed my speakers. Two things have kept me from doing that. 1) It seems the industry, for some reason, does not want us to, so finding the components is ridiculously difficult (Lumin M1?, Innuos Pulse?). 2) There doesn't seem to be an interface solution, except to use a spare laptop or your phone and a ridiculously expensive service like Roon. In my case, I think it would be nice to navigate my library on my TV.
Auralic Aries G1.1 can be bought with a 2TB internal SSD drive.
Dear Paul, streaming from the internet means that the file has to travel from its storage location through copper wires and optical fibers for thousands of miles to arrive at your receiving device. Many storages and servers are in Asia and Europe.
The Eternet quality cables on internet boxes, modem, router, and signal speed are extremely important. The WI-FI is mostly used among several devices, which compromise the download speed and stability quality. I recommend checking the internet installation and all the items I mentioned above to ensure a clean and fast stream.
Try to use new CAT 8 Eternet cables of the Internet setup and connect the DAC or streaming device directly on an Eternet port. Taking this measurements will help to receive the most pure signal possible.
Paul, thanks for sharing your years of experience with us!!!🙏
Hey Paul I tried Qobuz streaming on a 7 day trial period after my son was raving about it and compared it to my CD copies and I was extremely disappointed to the point where I was questioning my own hearing. The streaming sounded flat and less dynamic than my CDs. It was obvious to the point of not even attempting to listen for subtle differences. I won't be subscribing to any streaming services anytime soon, I love my CD collection and I'm sticking with it.
Same here! The same applies to my Blu ray movies and TV shows. 1080p 2k. And 4k. No comparison
You need to plug your modem/router into the wall with an LPSU and use high quality ethernet cables to make it at least a fair battle.
@@connorduke4619 you dont need high end network gear. networking would not work if not bit perfect.
With the advancement in the codecs BT is getting better... I can actually live with it now for casual listening
I just stream UA-cam Music (the paid / not free service) from my iPhone / iPad into one of my nice vintage systems and I'm a very happy guy. I have zero interest in chasing a better sound. To me, it's all about the emotional connection to the music. To each their own.
Happy New Year to Paul and the PSA team 🎉
Thanks for the download your songs to get better quality suggestions below. I immediately tried that and compared the Tidal download vs the CD rip of the same CD I have. Previously the CD rip was audibly better. Now with the Tidal downloaded version they sound identical after I AB compared then. All this through Zenstream via Tidal connect. Quite a big improvement so thanks.
I know your post is a year old, but I’m curious as to how you replayed the downloaded Tidal music through your system. Tidal downloads are only possible within the Tidal apps. I could be wrong.
Nowadays you can stream CD quality or better with lossless FLAC. Technically a 192kHz 24 bits track via modern streaming (such as Amazon Music HD) is much superior to CD or DSD quality. If it sounds inferior it’s because the source track is inferior. MQA, MP3 or other lossy formats of course will be inferior. Perhaps the truth is a combination of streaming providers still haven’t cleaned up some of their music repository to CD files and some files intentionally being optimized to mobile devices with more compression for better per device revenue (more clicks). The notion that there are data errors, jitter or processing noise involved was valid 2 decades ago but not nowadays where DACs can manage the flow asynchronously and ‘modern processors crunch FLAC files as a tiny effort. We gotta talk about today.
Thanks. I stated what i'm doing in the comments and my experience so far sounds similar to yours.
There is more than that… when we supposed to be lossless, powerful download capacity, same system , same DAC… note the same results? Streaming company found a way for cheating… compression for storage and uncompressed for used? Don’t know… but there is something else!
@@Skye_the_toller Possibly some tracks are still saved in a lossy format and streamed in FLAC, which is sort of cheating. From what I heard this is just a transition issue as services go lossless.
"Technically a 192kHz 24 bits track via modern streaming (such as Amazon Music HD) is much superior to CD or DSD quality" : NO, nothing beats the NATIVE DSD since it a has sampling frequency of 2.8224 MHz and that it has not been remixed (because remixing a DSD recording requires a conversion to PCM, and the original resolution is then lost).
Try it, and you will be blown away by the quality, dynamics, resolution, and realism! ;)
Be careful: there are plenty of SACDs on the market that incorporate DSD files that are just conversions from PCM recordings. Choose a brand like Pentatone (or others) that makes native DSD recordings.
@ I’ve listened to DSD and when it sounds awesome it also sounds awesome in PCM version. It sounds great because audiophile care has been put into the production process. All new DSD distributed music is an inferior transcoded lossy version of PCM (call it DXD or whatever) and still better than those who went through a noisy DSD DAC/analog mixer/ADC “PCM free” mastering process. Only when older music existing on analog tape is to be digitally distributed, DSD can make some sense, but since almost nobody is using DSD, best is to distribute it in better than CD PCM. The sampling frequency of DSD is only with a 1 bit sigma delta value that yields a lot of ultrasonic noise also. A PCM stream of 192kHz 24 bits will have excellent precision in every measurement and with such sample rate you can get the phase and level right for any audible frequency. Some people say DSD-64 is equivalent to PCM 96kHz 24 bits but I would prefer 192kHz 24 bits to leave any doubt out. Already at 44.1kHz CD quality sample rate and the average audiophile not hearing much above 12kHz, the treble precision issues and low pass filtering artifacts are not as important as you imagine.
I use a Naim Uniti Atom to stream Tidal and my local FLACS. and can't say I notice any real drop off between the two or the CD played through a transport. Although I don't get too hung up about ultimate SQ, I'm more of a music first listener. When I want the best SQ I play Vinyl.
Different masters/editions aside, you have higher potential for SQ with your digital files as they have a higher capacity than vinyl, and have less distortion added from the format.
I'd love to hear what others are doing. Here's my current set up. I started streaming and purchasing hi-res and CD quality music through Qobuz. I stored all of my CD's and purchases on a Bluesound Vault. I then purchased an NAD M10v2 to integrate with the Vault. I have nmy NAD CD player connected to the NAD M10 with RCA's and the Vault could be connected but it isn't because it's sending the music to the NAD through my home network. So far I can't hear a difference listening to a given album via the wireless Vault connection or the CD player. The biggest difference is how good/bad the recording is and again it sounds good/bad using either method I described.
I started off about 2 years ago to make my listening more convenient using the Vault 2i. I have over 12K songs in FLAC and at first played through my HT system all hard wired, then got the little Pulse Flex 2i's for the kitchen (sold all my albums and 35 yr old high end equipment), just after Xmas added to my unused 30yr old sound room new speakers, sub and integrated amp hard wiring the Vault to it. I also did all the treatments for the walls and at 70 I like listening to music more than ever. I also spent a bunch on cables as I made a pair of interconnects for my old system which blew me away on the difference between the class A power amp and pre-amp. Paul's video makes me wonder if getting a subscription on Tidal is worth it, Qobuz isn't available in Canada due to government rules.
It’s simple. On Apple Music turn off sound check. On other streaming services you are out of luck. I believe they do sound check on the server side
Interconnect quality might also contribute to the problem. When I first got a streamer I connected it to my DAC using a generic USB cable I happened to have on hand. The result was not good. Suspecting the source of the problem, I replaced the generic USB cable with a fairly high-end one and the improvement was dramatic, almost but not quite as good as playing a CD or SACD.
Can anyone explain how this could technically make a difference with the audio quality? A USB cable exists only in the digital domain, meaning the “quality“ of what is transmitted, will be identical on the cheapest of cheap USB cables, and one made of solid gold, the only difference should be the speed at which the transfer of the ones and zeros occurs.
Unless I am missing something?
Bullshit streaming sucks because of wifi polution
I'm fairly certain that services like Tidal and Qobuz don't rip their own content from CD, at least not for their entire libraries; I suspect they get digital files from the music labels then process and encode several copies at various bitrates and formats. Assuming I'm correct, It's obvious that the digital files aren't what's on a CD because both services have 24/96 streams of many of their songs.
It wouldn't surprise me at all if one or more steps in the processing and conversion pipeline are impacting the sound, maybe even deliberately.
I would tend to agree that it's the labels that are supplying the digital files, and thus the overuse of the "remastered" moniker. Putting on my tinfoil hat, I'd go so far as to say that those supplied files likely have a sonic "watermark" to tip the labels off on which streaming platform's copyright/encoding-decoding process has been broken (and the digital files end up *somewhere* else for free).
There has to be compression somewhere down the line. The data goes through so many servers with so many variables along the way. On top of that expense, and connection quality. Best just straight downloading the file, or buy physical media.
Media streaming uses UDP which doesn't do error correction. This is due to the continuous and time sensitive nature of streaming. So if packets get dropped or bits get flipped, your DAC will have to fill in the gaps. Think of it like quantization noise. It's the same with streaming TV. Sometimes the picture gets blurry when your WiFi is acting up or your ISP slows down.
Nailed it. Was looking for this comment.
"Media streaming uses UDP which doesn't do error correction. "
Wrong - most streaming provides are using TCP/IP. But even when using UDP that does not mean that there is no error correction. It only means that no error correction happens in the network stack. Still applications using UDP can implement their own methods for error correction which might be optimized for streaming.
For example the application using UDP may decide that it only request a resending of an defect package in case the buffer is filled enough to have enough time for that package to be resend. If the buffer is nearly empty it might be better to play the defect package as there is typically only 1 bit wrong in that package and such a small error is normally not hearable while an interruption due to a buffer underrun is always hearable.
"This is due to the continuous and time sensitive nature of streaming. So if packets get dropped or bits get flipped, your DAC will have to fill in the gaps."
You have no idea how streaming works. Internet speed is varying all the time - normally data is transmitted much faster than it is "consumed" by the DAC but there might be also short periods there no packet is received at all. Packets might even arrive in wrong order.
Therefore there is ALWAYS a buffer before the DAC and as long as that buffer does not run empty (which never happens if your average Internet speed is much higher than needed to transport the amount of audio data) where are no gaps which have to be filled.
"Think of it like quantization noise."
Quantization noise has NOTHING to do with streaming. Quantization noise is caused by the Analog to Digital conversion there analog values are converted to "samples" and this samples have always a limitted resolution. Audio CDs are using 16 bit samples - so they can "only" represent 2^16=65536 different values but the analog source has an infinite number of possible values. There are infinite different voltages between 0V and 1V but with 16 bit samples you can only encode 65536 different values - so sampling always causes a (small) deviation form the original value which is called "quantization noise".
But even using 16 bit samples this "noise" is so small that it is not hearable. Using a technic called dithering you can even "move" that noise to frequencies there the human ear is less sensible.
"It's the same with streaming TV. Sometimes the picture gets blurry when your WiFi is acting up or your ISP slows down."
That happens when there is to much traffic on you WiFi (or to many other WiFis on the same frequency band in your neighboorhood) and/or if you Internet connection is not fast enough or its speed is varyiing to much (e.g. due to a shared medium which might happen if your Internet access is based on a former cable TV connection their many users are sharing the same cable).
@@thomaswalder4808 I appreciate your insights but not your condescending tone. This is a friendly forum. People who enjoy music are usually cool. If you feel the need to bash people with your superior intellect, seek out a political forum.
@@sly_perkins It is not my "superior intellect" - it is just about facts. If you can show that anything I wrote is wrong I will apologize.
And by the way - this is not a HiFi forum - it is UA-cam
One of the things that was touched on was the hostile environment an stream has to traverse in order to reach your system. There are many outside factors that can affect it through all that copper/fiber it has to travel through. That being said, I listen to rips (FLAC) off a music server, CD, and stream. To be honest to my ear the sound quality seem to be on equal footing. The only time I hear something that has a noticeable improvement is contrasting a well engineered SACD versus a High Res stream from Tidal and the like. When listening to the same album from the SACD. The sound is cleaner, faster, and having more detail than the high bit Stream version of that album. As mentioned at the beginning, my theory is that the signal has a far shorter path and there is less chance of grunge from things like routers and switches getting introduced.
It's physically impossible for routers and switches or any network or computer device to make any kind of sonic difference in the way your thinking. That's not how computers work, at all. In all instances, every single time, data moves from one device to another, it is copied. Even within a single device, for example from the storage to memory, or memory to the CPU or CPU to network card; that data is copied bit perfectly to the next stage. Noise is ignored and not transmitted to the next step, the signal is regenerated at each copy. Every single time. Noise CANT be transmitted beyond a single step it's just not possible. It's not a continuous chain like analogue. A digital chain is discrete steps and is completely immune to noise below a certain threshold. (Which is a very high threshold and extremally noticeable when it happens. Buzzes, pops, pauses etc) This is happening in side you computer, streamer, CD player and DAC all the time. Each step may introduce it's own new noise, sure, but it doesn't propagate. This is kind of the entire point of digital signals, noise imunity. The ONLY time noise and/or jitter can make a sonic difference in the digital realm is the final step, inside the DAC, hitting the converter circuited itself. Any high end DAC should have galvanic isolation and jitter compensation of some kind before it gets there, though not all do. If your DAC is not doing this, then the problem is entirely, 100%, the fault of the device feeding it, not something up stream. If your DAC is fed over USB it can only be the DAC's fault.
The only other possible explanation is that the set of bits you are receiving is different from the digital master. Which can only happen if your streaming service is lying to you.
@@G3rain1 Agree, sadly some companies want to make snake oil business out of the data path to the DAC. For anyone understanding how IP networks work, this is quite laughable. Even any cheap modern DAC can handle buffering packets and run the clock with zero audible jitter and talks about “FLAC decoding noise from processors” and other such topics are also in snake oil territory. Rather modern DACs can output with SNR near 120dB with proper internal power rails filtering. With bad engineering, all things can go wrong, but that doesn’t mean we need to use them as real concerns.
Modern online audiophile streaming is already here and companies need to adjust to this reality with real products. I’m thinking about putting a pair of LS60 in my gym room. KEF is getting it.
Pure nonsense - ethernet by definition is bit perfect - there are data way larder and much more sensitive to a single bit flip than laughable small audio
That sounds vaguely plausible if you don't understand how Ethernet works.
@@CraigArnolduk and that's the problems: audiophiles don't understand digital technology nor ethernet nor physics - hence you can sell them anything and as long it's expensive enough you can call it "Highend"
I'm currently investigating this. I had problems earlier with my stream and I started tinkering around. I ran direct connection from my modem/router as a bench. Then I tinkered around with an external USB post antenna replacing the wifi antenna with a bigger one I had laying around and that maxed out the signal. I had a solid 54 Mbps (former stream was 3 bars). From then on my stream from Apple and Amazon unlimited sound fantastic. I tested CD vs stream adjusting my volume with a DB meter to specific marked points in a song. After doing that output match I could not tell the difference in the stream and CD on two direct comparisons. More testing to do, but I could not really notice a difference running through my SMSL DAC.
When I was rebuilding a new system from scratch cheaply, I didn't have a streamer, straight from computer to DAC. The difference in sound quality between streaming and local files was big. Streaming sounded awful. I added a streamer= computer-streamer-DAC and it made a huge difference, now the difference in audio between streaming and local files is, small enough that I often can't tell the difference to be quite honest. Your mileage may vary, but I learnt a while ago that if your source is a computer, you really should think about a streamer, sound quality difference is not small and there's some great streamers for affordable amounts of money.
What do you mean by streamer=computer-streamer-dac? Could you elaborate?
@@cliz305 signal chain, so laptop, streamer, dac. Having a streamer unit in between DAC and laptop to isolate electrical noise. I have an iFi Zen Stream, other streaming units are available, including I've heard- some very very good raspberry pi type units, very affordable. I think Cheapaudiophileman has recommended some Wiim units, something like that- best to check on his channel.
The stramer is more important than many people belive, I just upgraded my pi hifiberry digi+ pro to a Ian Canada Pi build, and the difference is night and day. I spoke with a friend about it and we made an anlogy to turntable, where the streamer is the pickup, and dac is the deck/pickup amp.
The streamer I build is battery and supercapacitor based in powersupply, reclocker and transport and it is not a small defference, it's actually quite huge both with streaming and local files.
For those that are qurious about more info:
ua-cam.com/video/Tc_QfRTbb6U/v-deo.html
And here is what I wrote in the thread after building it and first couple of listens:
So I buld it.
I chose the
PurePi
FifoPiMa 1.5
TransportPi AES
and Accusilicon 45+49mHz clocks
Had an Audiofonics Powersypply and a Raspberry 4 from an earlier build based on HiFiberry digi+ pro
It was very easy to assemble the boards, and didn't touch anything else.
Connected to an Lyngdorf 2170 with a Supra COAX cable and Raidho XT-1 speakers with Some Dali Cables, no power conditioning.
What a difference against the HiFi berry but also a friend that has a Aries G1 was blown away with some, maybe most aspects of this build, he promised to bring Aries over next time he's in town as his stereo is much different to mine...
What is immediately apparent is the way the sound was flowing with ease, like I only have heard from much more expensive stuff, the effortlessness comes from a total black background. and the imaging much more 3D and holographic, what also hit me was how the bass on older recordings was fuller and on newer recordings much more nuanced, everything sounded much more coherent especially well recorded/produced music sounds much, much better and somehow more sound, but the one thing that stands out is the ease, how the music flows, and that aspect surprised me the most in how much impact it has on the whole picture...
This upgrade is sonically the biggest one I have experienced in quite some time, I just went from a Digi+ Pro, Hegel H90 and a RME dac, I'm sure the Lyngdorf also has a hand in this, but it went from different but same with the Digi+ Pro, H90 and RME to something very, very special with the Ian Canada streamer... I'm for one am a very happy camper. I'd just whish I had tried this streamer with the RME I was so very fond of.
I bought a cd from yatao and a wav download from the same album of them. I burn a cd with the downloaded songs and then when I received the original CD compare both.... The original CD was better sounding than the downloaded versión. Streaming and download music are practical. But with time I became aware the sound quality is relative. And most of the time depend of so many factors in the audio chain. From the recording to the final system that will play the song.
I think his audio quality is being automatically adjusted when his Internet connection has a poor bandwidth. That option is usually turned on by default but you can turn it off but then you'd have your music buffering in between if your connection drops.
Nowadays that isn’t a problem when you have a good connection.
You can fix that in settings. It will just buffer more but still have highest quality
@@ThinkingBetter A lot of people say that, but it seems that no matter where you live, there always seems to be connection problems of some sort, whether it be local RF interference or weak signal problems from the ISP or whatever, there’s no flawless system. Obviously better to use a wired connection to avoid some of the wireless pitfalls, but unfortunately ..
I live in a heavily populated area and nothing but bandwidth and connection problems.
@@ThinkingBetterDepends where he lives…
I am connecting my laptop as a streamer using WI-FI. That might have something to do with the quality level. I am not sure yet if the recordings themselves are tampered with or the delivery process is compromised. When I play yessongs on vinyl it sounds great. The HD version on Amazon sounds good but loses depth and width on the Soundstage. I am still new to this so it might be me.
You didn’t mention the software used to play it… the differences are often the apps or software used. If u use something like roon it will play local or tidal or qobuz with the “same” software removing this.
The main reason I see for differences is their original copies of albums are different to yours. You may have a cd from one country that’s different from another and the one on streaming was from a diff master etc.
Storage is cheap these days. That’s really not an issue it’s just a cost of doing business like anything. 5 bucks per terabyte per month isn’t expensive when revenue is in the millions per month
Think about it. 50 million albums if on average 1 gigabyte per album is 50,000 terabytes which is 250k a month storage costs. If they have 5 million subscribers at 20 bucks a pop like qobuz pricing that’s 100 million a month revenue
1) Similar to a vinyl rig, IME it takes care to get streaming to be on par with local file playback. Your modem, router, switch, streamer and all their power supplies' quality (or lack thereof) matter. Cable and unit isolation also matters. 2) Can local file playback equal CD playback? Yes, but it depends on the quality of your DAC versus your CD/SACD player. This is my experience with a very resolving system and much experimentation. 3) Most people don't perform equal caliber CD player to streamer comparisons, with many folks assuming their CD player's USB input (using the internal DAC) is as good as the CD player's primary (more sonically optimized) playback stacking the odds against streaming.
Personally, Tidal's MQA leaves a lot to be desired. The only two streaming services even close to what I would consider as hi-fi would be Amazon Music and Qbuz. I feel like Qbuz has a little bit better dynamic range and sound stage, but Amazon's catalog is much larger. So I mostly listen to Amazon Music's Hi-Fi service. It's crazy how different music can sound from artist to album to producer to playback method to the equipment. It's a very deep rabbit hole.
Yes, Amazon Music HD streams lossless CD or better quality already and it works. Someone should do some A/B data comparison with the real CD data to see if anything is compressed or done on the Amazon side on particular tracks. If so, it’s human actions doing it and shame on that.
Tidal is HiFi no matter what anyone says. As for MQA, that's up for debate even though I don't mind it.
@@bkkersey93 No need to debate too much. Fact is that MQA is a proprietary LOSSY format. FLAC is an open LOSSLESS format. MQA was created for those who wanted more “quality” than MP3/AAC, but didn't want the file sizes of FLAC.
MQA is alright if your internet speed is very limited but there is no sense of it if you have a modern adequate connection that can also do much more demanding tasks such as 4K video streaming.
In my experience the streaming service is the main factor, Tidal is for me the lowest quaility, Qobuz was slightly better, I have not tried Apple music, however for me the best sounding streaming service is Amazon Music HD, I have music on CD and when compared to the same track from Amazon Music HD, I cannot really hear a difference, and I do have a hi resolution system. However, the quaility of the recording /mastering seems to be the main factor, a poorly mastered piece of music is poor from CD or Streamed, a 192kHz piece of music from Amazon HD vs a 44kHz CD recording is noticably better. How I see it is that there is good and bad in all the different mediums, but for me Tidal is by far the worst......
Pre digital everything was simpler.
Not simple, but simpler.
I can see no other logical explanation than compression here.
Im not an audiophiliac Luddite.
Dolby took the obvious and standard approach with their product back in the day.
If it is possible to manipulate wavelengths electronically by computer then its theoretically possible to do the same with noise/THD.
I think a new approach is needed.
We want natural 1st generation copies of the original performance with nothing ever added.
We want no THD or noise.
"Compression is suppression and causes audiophile depression"
Hi Paul - wish you a happy new year and by the way:
I am currently building a system that outputs all digital sources via just one DAC.
I need this "standard" music source for a small development around a ClassD amplifier.
Luckily, so far I have found that the sources, transport routes and compression methods have no influence on the sound if the file then got to me without any bit errors. To avoid jitter, etc., I use an old Intel server with 2 XEON CPUs and 138GB RAM, in which the music file has to be complete before I play it. The fact that CD players, streamers and whatever else sound different is due to the fact that different DACs and different electronics are used for playback.
So it will be impossible to hear the same music with no audible differences between the CD and the sound card.
🙂
Andreas
How do you do this? "Cache" the songs on your computer before you play them?
@@williammurray9055 Hi William
Yes, my program loads the file before releasing it. More importantly, I can send a constant amount of data to the DAC at a guaranteed consistent rate.
I use Linux for this, since Windows is not real-time capable.
For my project, I chose the CD-quality WAV file format. The bitrate is then
2 channels x 16 bits x 44.1 kHz = 1411.2 kbps
which are constantly sent to the DAC.
Of course, this doesn't work directly with compressed files, which is why I convert the data to WAV format in my memory before playing it.
This is not rocket science either, as there are already very good software components for the individual tasks.
🙂
Andreas
The differences being heard are most likely nothing to do with the streaming services. It is most likely that the music on the CDs listened to are older masterings which tend to be more dynamic. Streaming services tend to only get the most recent masterings which are almost always more dynamically compressed than the original mastering.
That it is the most likely answer not bandwidth or file compression.
What you said is what I've heard for years about streamed music. It makes sense as portability dealt music fidelity a serious blow years ago and settling for less became the norm.....till the young folk brought vintage equipment back.
I have a substantial vinyl collection and an even more substantial -CD collection, vut, these days I stream the vast majority of my music, it’s just so damned convenient. I have a highly resolving system incorporating a premium, state of the art independent DAC. I subscribe to multiple streaming services but, Apple seems to be the favourite when my streamer is searching for tracks, albums or play lists.
Yes, I have had friends and family round to help me do a blind test between a track played from a CD and that same track streamed from Apple. If I listen intensely, I occasionally hear minuscule differences. That doesn’t make one format better than the other, it just means they are different. I do however find that if my streamer is connected via my home broadband network, there is a notable deterioration in sound quality compared with using a mobile router. There could be 101 explanations for that phenomena.
By the way Paul, I can’t speak for other streaming services, but I do know that Apple has a library in excess of 76 million tracks.
Just as an aside, I’ve also blind tested friends and family, playing the same track on vinyl and CD. The majority preferred vinyl. That doesn’t make vinyl better, it just means that on my system, the majority of friends and family would choose vinyl.
The thing that drives me bonkers is relying on what sites like Tidal and Qobuz say about the bit-rate and quality without being able to easily check it. To take an audiophile classic, Michael Rabin's The Magic Bow, as released by EMI on CD (whether as Strings by Starlight or in their box sets of his recordings) has a sample rate that is not full CD quality. I can rip my CD and look at the spectrum analysis and readily see that. With a streaming service, it's nowhere near that easy.
I dont see how seeing the Spectral Analysis of a tune could give you any indication of the quality of that file, with test files with tones and such sure, but not music.
" The Magic Bow, as released by EMI on CD (whether as Strings by Starlight or in their box sets of his recordings) has a sample rate that is not full CD quality"
If it was released as an audio CD it must have a sample rate of 44,1 KHz. In theory they could do the analog to digital conversionen with a lesser sample rate and then afterwards convert that to 44,1 KHz by interpolation - but why should anybody do such wierd things?
I stream tidal a lot on high res. I got a Fiio dm13 (cd player) and started listening to some of my cds. Wow. It was apparent how much more dynamic, open, overall dramatically better. I love streaming for the variety and convenience, but the sound from the cd is way better.
Audiophile switches definitely matter. I was sceptical about it at first (bits are bits) but then I got a Sotm switch and was pleased by sonic improvements even with a stock ethernet cable.
Only when you are dumb and don't understand ethernet - your holy switch by definition can't distinct if a packet belongs to an audio stream or a word document
@@Harald_Reindl Only when you are more dumb not to realize there is electrical noise passing into your streamer together with the data stream. But the dumbest people think they know everything by not trying out by themselves and not hearing what others share
@@larazss3254 Muhaha electrical noise on the ethernet layer? How do you imagine that could change the decoded data? You morons know nothing about digital data and layers
@@Harald_Reindl Try it by yourself first hurensohn
@@larazss3254 why should I try idiotic snakeoil when I understand the technical details of all layers?
It depends on the songs....some songs sound great...some sound noisy....it all depends on the original studio recording quality I guess....and adding a DAC in between the phone and your output will bring the audio quality the to the next level.
Happy New Year to everyone at PS Audio.
This was an interesting question you answered today.
I recently switched from using my IFI Go Blue from exclusively in Bluetooth mode ( Apple iPhone using Amazon Music HD) ..and it’s nice .
I recently switched from Bluetooth to directly wired from the iPhone ( with the abomination of Apple connectors) and WOW…is it a night and day difference..nope ..but there’s more space , more dynamics and a sense of it just bringing more to my ears .
Take care and may ‘23 be a good one for all.
I believe that all Bluetooth has limitations, it's probably written in the manual for the DAC. 🙂
Yeah, almost all of the Bluetooth codecs are lossy
Let's assume for a minute that you're using a computer, the stream is bit-for-bit the same as a CD rip, and it still sounds different. We have to ask what else is different? Is there two different pieces of software you are using to play the FLAC vs the stream? If so, are they using the same driver stack to communicate with the DAC? (WASAPI exclusive, ASIO, kernel streaming, DirectSound) Are they resampling to the DAC's sample rate using different methods?
When I used Tidal, there was definitely a difference between the CD rip and the stream. And not just the mqa files, as those we know are not bit-for-bit. I have since switched to Qobuz, (I will hold my tongue about Tidal for now) and, after making sure Qobuz and Foobar were both using ASIO, I could no longer hear a difference. I suppose there should be should have been some difference as Foobar was set up to do its own resampling, but it wasn't big enough to notice right away so I did not do any further testing.
Why would you use asio vs kernal streaming? Kernal streaming on windows 11 with newest Audirvana edition is amazing
yeah, it is about exclusive and non-exclusive access of the player application to the audio hardware. if exclusive mode is used, streaming sounds exactly like the CD. if non-exclusive mode is used, streaming sounds muddy and lacks dynamic. it is a night and day difference.
The problem is always the same. Jitter and electrical noise. I used filters, audiophile switches with clk and in the most cases they clean the sound, better silences, better dynamic but always seamed to be lost some naturally, musicality, until i hookup my server directly by optical fiber the sound it started to have a much greater cleanliness with some silences like black holes. With this experience, I came to the conclusion that I was doing bandages before and now I have treated the disease. The optical fiber, being immune to RFI and EMI, is not affected, taking the signal directly to the server always clean. Today is very difficult to me distinguish the streaming from flacs or wav.
I really think it all comes down to system synergy to get a good streaming setup to sound great. It took me a bit but I found what I wanted by trading out a couple DACs and Streamers to find the right combination. Now Qobuz sounds better than CD in most cases on my current system. Even the correct USB cable makes a pretty big difference. All about system synergy and what sounds good to your ear 😊
@@davidgoffin1537 I started off with a Bluesound and then tried it with a R2R DAC. That DAC failed and I wanted a better streamer so ended up with Gustard R26 into a Zen Stream with external power supply. This combo is perfect for the type of sound I like.
I am pretty sure there is ZERO compression in transfer or storage on Qobuz, just the lossless file provided by the artist. At least that's what I gathered from David Solomon from Qobuz (who also used to work for tidal).... I know MQA is compressed and Lossy. My thoughts. ALSO - Ensure you are listening to the SAME Mastering.
There are 2 factors to consider.
1. The fidelity of the data. Is it bit perfect from cd transport, a network device or a commercial streaming service. The first two are identical. The streamer may be using a lossy compression or volume leveling
2. The quality of the 'transport' which determines timing jitter and noise levels that accompany the data. The 'AirLens' or other devices will supply the data with less noise and a given phase noise. Given a bit perfect data stream, the data stream should be better than from a conventional CD Transport given your DAC doesn't have the same function.
I rip cds to a network hdd and play using a streamer with very good galvanic isolation and a SOTA clock. Playback is far better than from a CD player or transport. If I access a streaming service and find inferior sound quality, then it is the fidelity of the data that has been impacted. AirLens as Paul describes it would be similar. But it cannot correct a data stream that is not bit perfect as supplied from a streaming service.
Paul, as I was listening to your explanation of storage and transmission, could this be the real reason behind MQA? I'm a Qobuz user, so I don't have a dog in this hunt, but it does raise an interesting idea that as time goes on and music services are pushed to store even more recordings.
I'm not Paul but yes storage costs. Its not just an x number of petabytes in 1 server suite somewhere in a datacenter. That data needs to be redundant so there is a backup cluster. Making it 1.5x the cost. And of couse, its not only 1 dc but multiple around the globe (bc. uptime/failover). So again 10x the cost. And of course every file has different quality versions. So again x-times the costs. And then comes data transferring between the datacenters and eventually the transferring towards the customers. Anyway, every MB you can save on file size is big money over a longer time period like 1 yr.
@@the_dude182 Hi, yes that all makes sense. Originally, I thought MQA was an interesting idea for Europe where unlimited data plans weren't as widespread as in the US. Occasionally, I see albums deleted from Qobuz, and when I factor storage that makes sense. As a user, I really don't want to be bothered with MQA. I have an iFi Signature DAC (v1) that sounds much better when I change the firmware to the non-MQA version. I listen to much more music because of Qobuz than I ever have in the past.
This may be subjective, but I don't think streaming service matters. It may have more to do with implementation. For example, SACD/DSD on a locally implemented hardware based system should be the choice for all. To test streaming,I directly compared Qubuz to Tidal and found MQA was more to my taste. I got a similar level of the thing audiophiles would look for: soundstage and imaging and details never heard before, but my implementation is specific to me. To lean in more on this topic, I would like to take Paul's suggestion on using a hardware cd player and compare that to local file storage and retrieval methods using network and computers (which is what I suspect maybe the limitting factor) to see if there are differences. Update a/b direct CD transport to streamed local file via Roon. No noticeable SQ difference (good for me as I can be happy with my method of storage). However, a/b/c test cd player vs 44.1/16 bit hifi vs tidal Master 44.1/16 bit. Master had the best SQ, 2nd best was cd player, and the tidal hifi quality track sounds the worst. As always, I'm never really done, but there seem to be many dynamics when comparing.
I dont THINK anymore Paul says buy a Mcintosh so thats what I will do. Thats the great thing he is Honest why push his own stuff i mean just look at the speakers in his house.
In the early days of computer audio, there was a holy grail of Pure Music version. It only ran on an older version of Mac OS and if you upgraded the Mac OS, it broke that particular release of Pure Music. That version sounded amazing. I stuck with that version of MacOS (not upgrading/updating) for as long as I could. The developer of Pure Music was asked about what he did - and from memory, he said he worked on power optimisation - reducing power spikes. It was a monumental effort and I don't think he ever tried doing that again.
Coincidentally, when the CTO of ESS was talking about how his company transitioned from computer audio (with great SNR) to audiophile audio, he found that in the blind tests he conducted with various engineering samples, the audiophile designers always chose the versions where the various states in the digital-audio conversion had minimal power spikes as it transitioned from different states - and not how quickly it reached the final state.
On my own testing, many years ago, I had a Wadia iTransport that you could dock an iPhone (then a 3GS or 4?) to, and it would output the digital audio files through a SPDIF output that you could send to any DAC. At that time, my DAC of choice was the Ayon CD-5S which coincidentally had a built in CD transport (I2S). I played the same file (ripped from a CD) through an iPhone on the iTransport and also directly through the built in CD player. The CD player won. But on the advice of a friend, he asked me to rip it in WAV. I had previously used Apple Lossless. This time, the WAV version got me a lot closer to the CD. Everyone told me that converting from FLAC to WAV in real time took so little CPU cycles that it couldn't possibly affect the sound - but maybe there was just those transient power spikes, however short, that made a difference.
I remember the original Digital Lens on the old PerfectWave Bridge used to implement something similar, didn't it Decode to WAV ahead of time and put it in a buffer to be sent out. Is something like that in the new Air Lens?
One thing I notice is that any song I 'stream' has a particular quality but when I download that same exact song and play it from the download it sounds noticeably a lot better. All my settings are set to stream and download at the highest quality available and I have very fast internet. Not sure why there is such a noticeable difference but it seems to have to be in the 'streaming' process itself.
I noticed this with spotify, the down loads sound way more dynamic and punchy compared to streaming the same song, even tho i have it set to stream on highest quality and download in hi quality also.
Nonsense
@@Harald_Reindl what makes you an expert? Or are you just beaking for spite?
@@anthonycyr9657 the fact that I deal with TB large data files as my day job where a single bit flip would be fatal
@@Harald_Reindl lies..
"Streaming" implies a continual stream of data. It means that "perfect data" is not the only goal, also it has to meet the non interruption goal. When the transmission is less than perfect, it may drop data or lower the quality (ie bitrate). All of this is to ensure instant and not interrupted playback. Otherwise you would have to wait that the entire file is downloaded to be reliably played.
Also there is the problem of what CoDec is used, what source was used, what format sound is stored, etcetera. I don't really know if some of this applies to high quality streams offeredby some.
DSD streaming is another story.
Nonsense - Audio is that small that large parts or the whole track is buffered
@@Harald_Reindl You're focusing on bandwidth. You should be focusing on checksum, or lack there-of. Streaming has no checksum so, even if the bandwidth is monumental, OP is correct that some of the packets of data can be lost.
@@stevens1041 are you drunken? I earn my money with IT and if the ethernet layer would randomly drop packets music would be the smallest problem! I can safely transfer gigabyte large virtual machines where a single bit-flip would be fatal and you clown pretend your music will be damaged? The problem with you audiofools is that you lack any education and technical knowledge
@@stevens1041 Spotify instead uses TCP. Firstly, having a reliable transport protocol simplifies protocol design and implementation. - Oh and when you are using UDP a trained monkey can make sure data integrity and will do so
" All of this is to ensure instant and not interrupted playback. Otherwise you would have to wait that the entire file is downloaded to be reliably played."
If you have an Internet connection which deliveres data between 200 and 300 KBytes per second and for playback you need 150 KBytes per second you not need to wait for the entire file being downloaded. If your Internet only delivers 160 KBytes in average but transmission rate is varying betwen 140 KBytes and 200 KBytes the software should wait until the buffer is filled with some amount of data to make the chances high that there are no interruptions in the playback.
Playing back CD is like unpacking something that was delivered from across the street VS. streaming is like unpacking something that was delivered across the country via going thru different carriers and repackaging processes in between. Even both packages come from the same source, you can still likely to tell a (big) difference if you have sharp senses (ie resolving systems).
Nonsense - if that would be true files which don't survive a bit flip would randomly arrive corrupted and in that case every compressed and encrypted file transfer would be impossible
Hello, I master records for a living. Files digitally captured during playback from Tidal of masters that I've done have nulled perfectly against my house archives of the same masters. So, AFAIK, the data is the same as my source. Any differences in sound would have to be dependent on the end-user's system or network.
What if we rip the CDs? And we “stream” them from our NAS or from an attached hard drive? Does it still have noise? Is so, how do we play music files for the best sound? Maybe using something like the Eversolo dump a8. It can have an internal M2 NVMe SSD… would that be better than any thing that is using Ethernet and/or usb connections?
many commenters here show their ignorance on digital processing. All digital networks transfer data absolutely perfectly, no matter the route or distance. This is essential or so many of todays technologies would be impossible.
If the CD and streamed versions sound different it is probably due to different Masters. Is your CDs sound consistently better than CD quality streaming then you have a setup or equipment issue.
Agreed. Rips of Japanies cd's i have sound better than Australian release. bypassing windows mixers upgrades quality too.
The only thing I have found to help ....is an iFi SPDIF iPurifier, but obviously that only works if your streamer is hardwired.
Part of the issue with streaming is the way the packets of data are transmitted across the various hops (switches, routers, etc.) they traverse on their way to your computer, and how your computer handles those packets when they arrive. When you download a complete file, the packets are normally transmitted using TCP, which dictates that your computer check to make sure the packets are in order and none are missing. If there is something missing or out of order, the computer will rearrange the packets and/or request the server resend missing packets. Because of time restraints, streaming and audio/visual communication packets are sent over UDP, which means your computer receives a blast of packets that is often many fewer than the server sent (packets are frequently dropped at each hop the data traverses), and sometimes they arrive out of order, and your computer does what it can to rebuild the data (as you are listening) with what it receives. There is less of an issue the closer you are to the server (fewer hops to lose packets on), and the better your internet connection is (more bandwidth and faster data retrieval). Hopefully soon, routers, switches, servers, and networks will be more efficient, internet speeds will continue increasing, and dropped or out-of-order packets will no longer be a concern for music streaming. In the meantime, simply downloading the .flac files to local storage should make a huge difference.
Oof my friend you are only partially informed. The switching has nothing to do with the audio quality as you should know. Also streaming is buffered so generally this is not an issue
"Because of time restraints, streaming and audio/visual communication packets are sent over UDP,"
That is not correct. Time restraints are a problem if you want to transmitte audio or video close to real time - for example the live transmission of a sport event or telephony over Internet.
Streaming is not real time - the data is always buffered before it is played and as long as your Internet access is faster than needed for the playback there is enough time to request a resend of defect packets or to reorder the packets in the correct sequence.
Many streaming provides use TCP/IP - and even if the use UDP that does not mean that there is no error handling. The difference to TCP is that the application have to decide on its own how to handle errors.
The first Audiophile streaming service was Yodeling in the Mountains.
I think there's multiple detrimental effects that streaming has. One is the timing in music is extremely important. Networks aren't set up to have it occur at 44khz. Never any reason to. Speeds vary and interfere with it. The other is error correction that used an algorithm to 'guess' at what sound to make when it senses a corrupt bit. You don't hear it skip but that doesn't mean there's no error. To bad we can't shut off error correction on cd players etc. To then realise how often it goes into error correction. I use a NAS and don't stream from the net but for you tube on occasion. Even local flags don't sound as good as a cd.
Moron the timing is irrelevant in a ethernet setup - you buffer the whole track, decompress the whole FLAC and play it like it's a local wav file
Your DAC retimes everything as it reads from the input buffers. Any jitter across the network is irrelevant so long as the buffer doesn't stall.
"One is the timing in music is extremely important. Networks aren't set up to have it occur at 44khz."
The Internet does not care in anyway about sampling rates of digital music. The Internet transmit packets containg binary data.
It is like downloading a e-Book with 1000 pages (which takes only a couple of seconds). The timining of that download has nothing to do with the speed you are later reading this book.
Streaming differs from file downloads only in the way that you are able to start reading the "book" already even if the download is not completed yet.
After the first pages are transmitted you can start reading - and while your are reading more and more following pages are transmitted in the background.
"The other is error correction that used an algorithm to 'guess' at what sound to make when it senses a corrupt bit."
There are different methods to handle errors. TCP/IP requests a resend of a packet if it detects that the received packet has errors (which is done using a CRC checksum).
Audio CDs have instead additional bits which are allowing to correct such errors "on the fly" - only in the case there are too many errors which could not be fixed any more some "guessing" happens.
@@thomaswalder4808 so why does local always sound better? If it gets a bad packet error correction kicks in. Streaming is asynchronous. Doesn't ask for it back. Can't. The packets work on the data being computer correct. Not audio correct when it's recived. It's a digital signal passed on an analog square wave form. That's subject to jitter and timing delays. The end unit resolves the errors. What you are saying is there's no need for error correction at the dac. Huh?
@@steveaustin7306 "so why does local always sound better?"
You believe it sounds better because you expect problems with streaming. That is known as expectation bias.
Think about you have two Big Macs in front of you which were made at the same time - but somebody tell you that the second one was made 3 days ago and stored without cooling and heaten up again in a microwave.
Which one would taste better?
"If it gets a bad packet error correction kicks in."
Correct - but your understanding of "error correction" is wrong. In case a bad packet is received the streamer detect that it is bad because its CRC-checksum is wrong. In this case the streamer requests the packet again from the streaming server. Where is no "guessing".
"Streaming is asynchronous. "
Correct - therefore streaming always needs a buffer on the receiver side. This buffer is filled with packets arriving from the Internet and readen synchronous by the DAC for playback.
A normal Internet connection could deliver the packets much faster than needed for playback - so this buffer is normaly always full. In case a bad packet is received there is plenty of time to request this packet before the buffer would be empty and the playback must be interrupted.
In the early days of the Internet where most data was send over analog telephon lines such bad packets were common. Today it is extremly rare.
My Router keeps statistics of bad packets - I not had a single bad packet in over one year....
But as described - even if that occur it does not impact the playback at all.
"Doesn't ask for it back. Can't."
Why "can't"?
In TCP/IP it is unavoidable that bad packets are requested again.
"The packets work on the data being computer correct. Not audio correct when it's recived."
Receives packets are either correct or not. The CRC checksum used to determine if a packet is correct or not does not distinct between audio data or other data. It even does not know what type of data is in the packet.
"It's a digital signal passed on an analog square wave form."
There are many ways to encode digital data in an analog wave form. In the very early days sine waves with two different frequencies were used. A sine wave with one frequency was used to encode a "0" - a sine wave with the other frequency was used to encode a "1"
Ethernet uses different methods like "Pulse-amplitude modulation" or "Quadrature amplitude modulation" - these do not look like square waves...
But that is not the point anyway as long as the digital information could be extracted from that analog wave (however it looks) in most cases.
If it can't then packets have to be retransmitted...
"That's subject to jitter and timing delays."
Yes - and therefore a buffer is needed before the DAC which get rid of all that jitter and timing delays - and even packets arriving in the wrong order.
And every streamer has such a buffer
"The end unit resolves the errors."
Correct - so what is the problem?
"What you are saying is there's no need for error correction at the dac. Huh?"
That depends on what you understand by the term "dac".
Normally a dac is just a digital analog converter which has some digital input lines (for example 16 ) and converts the value represented by this digital input lines to an analog voltage in a defined range.
In the HiFi world DAC often means not only the Digital to Analog Converter itself but also some additional processor to convert different digital audio encoding standards and protocolls (like USB or SPDIF)
But yes - there is no need for error correction of "bad packets" as this is already done in the streamer which also fixes "timing delays" and "Internet jitter" using a buffer as described above.
I use Logitech Media Server + Squeezelite (running on PiCorePlayer) as my music streamer. I find that the settings of streaming buffer makes a hell of a difference to the resulting sound quality. I use 64000Kb stream and 192000Kb output and I find that the difference between locally stored flac files and tidal (non MQA) virtually vanished
Paul, you just explained why MQA is used by Tidal (besides being a copyright-protection-system which only works with expensive DACs)! 😉
Qobuz to me sound very good with jazz which I tend to listen to 95 percent of the time. Can't tell any difference frm CDs. But once you go popular music things begin to fall apart.
Some music sounds better on CDs others on Qobuz.
Streaming can sound as good and better than CD. It requires a fair amount of investment, but it can be done. Given the number of steps involved in getting a streamed file to a DAC, I'm surprised by that. But in comparing 16/44 streamed files to the same CD's, my ears tell me that the streamed files sound just as good. PS. That's via a high bandwidth wired (very important) ethernet connection and a high quality bridge.
Agreed 👍
You can't stream cd brother.
Huge conflict of interest for those depending on 'music on plastic' to make a living to defend 'streaming' and vice-versa. Thus, abstract opinions matter not, only the audio experience of an individual. I dont make my living of music, and like both formats. Music on plastic because of its physical/nostalgic/ritual component, streaming (via a Node) because of convenience. The TIDAL masters sound extremely good. As good as my best Music on plastic stuff, which includes some UHQR stuff. Voilà my truth.
Qobuz lets you download and store their files offline (Tidal as well, but in the mobile device app as far as I know) - when I playback these offline files, my DAC claims to be reading the same bit rate etc. off the file as advertised by Qobuz. I do this wired (on a short path) - of course, the files originally came to me from the Qobuz servers and data stream. Is my DAC being fooled when reading the resolution of the offline files?
One thing I did notice, when I downloaded all my identical Qobuz and Tidal playlists as offline files (in the highest res offered) to my phone, the Qobuz library amounted to many, many more GB of storage. I don't know, maybe this is because Tidal hi-res is MQA and that packet is smaller than Qobuz FLAC? Either way, I thought it a bit suspicious as fair number of my playlists were at CD resolution only as hi-res was not available for some of the artists I like on either platform (or the original recordings were early digital at 16/44) .
That makes sense, which would explain why few services actually offer FLAC and HD Audio. However, whole streaming wars, is the same as video wars...
It also really depends which equipment your listening on as well.. No one will notice if your listening on a pair of Apple Earpods, but on a high end systems, you would..'
It's no different than a playing an mp4 video,, with 1.6 Gig compressed file size.... No one would notice on a laptop or a 30' inch display, but will be pixelated on a 50' inch..
It depends on the user. that's all it is... This is why i don't let anyone tell me which is better, because its more simple than that.. Personal choice only.
Hi Paul, you just hitting the nail, let saying there is B3 CD player and Ripping, WiFi, etc. This B3 engineering and manufactured in UK by Martin Brennan.
Should you be able to give more information about this B3 CD player. Are they compression file as well?
Compression is the files we compressed to smaller size which we lost the original size- as well sound quality lost.
I use my PlayStation 4. Extremely cheap I know, but it can read FLACs, you can plug your favorite USB DAC onto it... Plus you can use Spotify and UA-cam if you want to. It's not the most silent thing in the world, but even with open back headphones I can't hear the fans spinning
FLAC was year 2002, dude.
@@rabarebra So? It's still the lossless codec that has the most compatibility in the market today. Much more than WAV.
if you dont hear adifference, its time to get your ears checked. I never stream music. downloading in 320kb mp3 or FLAC/WAV yes. but the most of all i prefer even my Vinyl albums (with good dynamics, not every label has, like "cough" CBS) above my cd's. they also good tho, but if you can get an album in ufcd MSFL ultradisc, i recommend those to buy if you like good dynamics
I'm wondering if the loss in quality is during streaming from a service (qobuz), or is it in the files they storage itself? in the first case the sound quality would be good if you download/buy the musicfile from the service! anyone knows more about this?
Equipment differences, streamers vs cd players, internal power supplies, caps etc, then DAC’s internal vs external are among the first layer of controllable differences, so it is hard to compare apples and oranges. I find even if I stream and upsample to 192Khz using the dac inside my cd player - an ARC Ref CD9 - the cd sound’s obviously better in every way. I’ve enjoyed streaming well recorded tracks and they sound good but cd’s sound so much more dynamic.
Interesting video! not sure it would be ethical if streaming services that claim to offer cd quality cut costs in that manner. I lately put much of my music in lossless ALAC format on my iPhone, which I airplay to an old apple tv which feeds it into my amp's DAC. Don't have a CD player so this is the best digital experience I can get. Any thoughts on this method? Would going over wifi and through all these devices somehow deteriorate the quality of lossless audio and would a cd player improve my listening experience?
The biggest influence on the quality of music, if the master copy. Someone once told me, that when a track is released multiple master copies are mixed and distributed, with each mix specifically catering for it’s intended recipient, be that cd, vinyl, radio or 1 or more of the many streaming services. So, when you hear differences between the same track played from cd or a streaming service, it is likely the difference in the master recording you are noticing!
While I don't really consider myself a true audiophile (my systems are resolving, but not to the point where I can discern huge differences), I can definitely tell a difference between listening to my original Brothers In Arms CD on my 90's Denon versus a 320kb version of that album streamed. Granted, I don't have the best D/A, but you can tell that there is a lack of dynamics with the streamed version versus the CD, no doubt about it. This example is only one of many. As I said, there is not a HUGE difference, but if I am in a "listening" mood, it is CD every time without question.
That’s also my experience! It’s much better than it used to be some years ago, but still worse than CDs
"I'm not gonna start..." *immediately starts*
Speaking of streaming's short comings...on paper flac and the original wav file look the same, and on null tests (which I've done) they null, but flac lacks some high end. There's something that the null test cannot measure, but it can still be heard and felt. Not what many people want to hear, but it's the truth. People who suffer from hearing loss cannot hear the difference. Too bad for them, they're missing out. Many prefer meta data over sound quality.
If you have RCA inputs in your amp ( main in) which is direct amp signal , you can put outputs 6.3 mm banana from sound card which already have DAC preamp into that rca main in amp inputs and you WILL get better sound quality from spotify or youtube .
A little fun with the KEF driver in the background since I watched a video from the Norwegian hifi magazine "Stereo+" youtube channel wisiting the KEF factory😝
A lot of the differences people hear is in their equipment or how they have things connected up.Your streamer DAC and cd player DAC are not going to sound the same.Try running your cd player through the same DAC using it as as a transport.You may be surprised.
This video is via hexibase. He is educated on this and many other topics. A Great video I think you will find useful
So many variables introduced when using the streaming option, from server storage to bandwidth differences. The vast majority of listeners will be happy with convenience, at the expense of ultimate quality.