A few important notes: 1) Whilst the video explained 1-bit delta-sigma and PDM at a basic level, most available delta sigma DACs are actually 'multibit' delta sigma. Meaning rather than modulating to high speed 1-bit, they modulate to high speed 5-7 bit as this allows you to avoid the biggest challenges with making a highly linear R2R DAC and you also don't need to be operating as fast or have as much difficulty with the modulator design as a full 1-bit DAC. 2) DACs that ARE 1-bit usually don't actually have a 1-bit PDM converter. Instead they usually use a moving-average FIR filter approach, this will be explained further in a future video. 3) Some products on headphones.com such as open box units may not fall under the 365 day return policy, please check the product page for specifics!
Strictly speaking DACs don't need oversampling. 16 bit signal -> 16 bit DAC -> reconstruction filter, no oversampling required. But when the DAC has less bits than the signal you need dithering to preserve the dynamic range, and when you need dithering it's a good idea to oversample to have less noise in the audible frequncies, even less with noise shaping.
The RME DAC that you showed on the picture in thumbnails and on the start of video are one of the best DACs I ever listened to, the sound it produces is so precise, it really amazed me and the functionality is very good, very well thought out (loudness and equalizer in it are really good) highly recommended if you have good amp, headphones and speakers and are willing to spend a little more money, it's trully a beauty to listen to
This was the most easy to understand explanation about the topic I've seen to date! Thanks Golden! Now I have something to send to my friends when they start questioning my sanity...
I legitimately learned a ton from this. Some stuff was in more detail than I needed but you broke it down very well and the reap at the end was perfect. Thank you very much. I think I’m going to try a DAC for the first time soon
actually its a terrible video, find an explanation from a physicist or electrical engineer...that 1s and 0s stuff is nonsense....look up what 8 or 10 or 12.... bit means
I just spent over £240 on a 4 meter balanced headphone cable from Oidio Sound. My current cable had worn through the outer layer at the right earcup, so they needed replacing. And I bought a Midgard headphone amp, which has 4 pin balanced output, but my current headphone cable is unbalanced with a 3.5 mm mini jack. Balanced cables do make an audible difference by rejecting noise, this is especially important for very long cable runs. Will they sound better? Absolutely, as the placebo effect can be very strong. Would I be able to tell which cable is connected, if I don't know if it's the old or the new one? Probably not. But I wanted a 4 m cable, so I don't need to use a headphone extension cable, and I wanted to move up from unbalanced to balanced. Most other companies selling headphone cables for my Dan Clark Audio Aeon 2 Noire either didn't sell longer than 3 m, or they were more expensive, or I'd have to pay import tax and fees. So while I am in the, "paid more on the cable, than the device it's connected to" camp, I'm happy with my decision, and looking forward to hooking it up, and listening to balanced headphone audio.
Your timing is just perfect, i just started 2 days ago to get into Headphones and sidekicks and already stumpled upon the DAC and Amplifier Problem. I'm ready to learn more! Btw, my current Headphone of Choice for my next shopping tour is the Sennheiser HD600 or 650. Chosen by your very big Tier List video from 8 Months ago ;)
Vids like this are excellent as there is always someone new to the Hobby/Industry. The visuals really make it for me. I'm always having to refresh my memory as to why I want one over the other and the visuals really made it pop. Good Job!
I presume/hope that the 3 “Biggest improvements per $ spent” list will be revisited in Part 2, with strikethough added to Cables, and replaced with Amplifier (possibly moved to #2, with Headphones #1). Loved the joke though, for those of us who saw it for the joke it was! 😆
3:15 I just want to point out that oversampling isn't strictly necessary to reconstruct the analog signal. So long as you have a precise low-pass filter at half the sample rate, the signal reproduction will be accurate. However, oversampling allows you to control how that cutoff happens much more precisely. I don't know exactly how common DACs that don't use any oversampling whatsoever are nowadays, but you'd find them for example in toys that include a sound module.
This is true, though the problem is it's not practically possible to make an analog filter that'd do this, hence why we typically do it digitally as we can make a very steep filter that way. With an analog filter you'd either need to attenuate content in the audible band too much or you'd leave too much above the Nyquist frequency.
Schiit Audio makes multibit DAC's that can turn off their oversampling filter and run in No Over Sampling (NOS) mode. My Bifrost 2/64 can do that. I have also seen other companies DAC's, which have the option to do no over sampling. I've not tried it out yet, as I only got my DAC and Midgard headphone amp unboxed and hooked up yesterday. But most people prefer Over Sampling, especially on the Schiit Audio DAC's, which use their proprietary time and frequency optimised oversampling digital filter.
for the pragmatist out there, two aspects kind of give the tone for all decisions to follow :) a) when they measure dynamic range, noise floor, etc, it's kind of important to keep in mind what kind of values are we experiencing in the real world and b) when they present frequency response and fq curvers it's kind of important to keep in mind our hearing is much much worse than it appears on paper. don't let the critics, reviewers and especially 'influences' ruin your feelings towards music and sound in general. the most accomplished mastering engineer still uses the same headphones (when he uses them) he liked a few decades ago :). i'm not saying nice things are bad, but i am saying the "nicest" things are useless. -------------------- analyzing the dynamic range of numerous music files, the actual difference between the loudest and softest sections isn't nearly as drastic as some 'theoreticians' insinuate. music doesn't sound and certainly does not feel great w/ insane 'volume swings'. not even 'explosions' do, even those are limited/gated and EQed to make them 'nice sounding'. let's take an example that should be easy to imagine: at the concert hall, the piece starts w/ a loud cymbal crash. so from 'complete' silence we jump to the fast attack of a high fq percussion instrument at full volume. but what's the full volume, how many dBs is it above the silence? for the percussionist the sound is louder than for the conductor and much louder than for a listener positioned mid concert hall. say you want to experience what the percussionist hears for whatever reasons. then what do you do when the rest of the orchestra starts playing? move away from the cymbals or bring the whole orchestra close to your ears? :) for most of us, the most pleasant experience is to listen to the whole orchestra from a distance, give the sound enough 'room' to nicely blend together and in most cases also take the character of the acoustics of the room. i hope this makes sense... because most recording engineers put great effort in balancing things out, then the mixing and mastering engineers refine that balance even more and in the end, we rarely (almost never) experience music that jumps in volume to the limit of the 16bits offered by the "old CDs". dynamic range is a bit different in real world than on specs sheets... your 'regular music' has less than 20dB of 'intensity range' and 'loud music' (be it EDM or metal) far less than that. when it comes to our hearing, perfectly healthy hearing of a post puberty human, none of us have the fq range everyone quotes, 20 to 20000Hz. it just doesn't happen and that's perfectly fine because again, analyzing the fq range of numerous music files, there isn't much information in the last half octave of the audible spectrum. if you're over 30yo, anything above 15kHz is pretty quiet but you're not missing much at all as counterintuitively, 'clarity' in music (recordings) is not defined by those high fqs. in fact, at such high fq our ears have a difficult time asserting pitch, location and even intensity. as the wavelength of a sound becomes shorter and shorter we're simply not equipped to properly analyze it. for instance, if you were to play a 15kHz pure tone from a speaker and would slightly move your head any direction you'd hear tremendous fluctuations of not just intensity but also timbre as the sound bounces around your head, ears and everything around you; you'd be picking up all sorts of 'interferences' from all those bouncing 'sound beams' and the fluctuations are drastic for just a few mm change in position. have you ever wondered w/ is it so difficult to locate a cricket that made its way inside your room? because we're not owls :):), we didn't need that sort of 'accuracy' and our hearing didn't care to develop in that way. although we can hear high fq sounds, our brains more or less ignore most of the data those sounds carry. (old men installing super tweeters is super funny -until they attempt to convince you you're not experiencing the Beatles at their full potential if you don't own ultrasonic transducers, then it becomes annoying at best.) ps: there's a good reason why not all studios use the same monitors/headphones, why not all artists use the same make/model of in-ears monitors... everything is a compromise and nothing can be objectively accurate in sound reproduction. they key is to find a balance that pleases You. and when it comes to 'accuracy', being familiar w/ a certain device yields far finer precision than 'the latest' or 'most expensive' model of whatever device. that's what sound engineers concluded since forever and i'm talking about ppl who can and did hear a bolt being lose on the drum kit of a loud metal band, not your average youtuber :)
This video was very confusing as I failed at math and science when I was in high school in the 90's and have learning disabilities but I was interested in it, I just had to replay and rewind a lot to try to follow and understood it vaguely.
I bought a JVC CD player in 1995 that had "1 bit DAC" as one of its main features and selling points. It was even printed on the front of it beside the CD tray. Knowing a little bit about binary (no pun intended, or was it!) I thought at the time it was an unusual property to boast about, because 1 is the least number of bits you can have. Obviously I didn't understand how the DAC worked and the significance of "1 bit DAC". Now I do!
You still don't understand much. This video is very oversimplified, though it gets the core concepts right. But, the people who need such a simple explanation, probably don't know Ohm's law even and, as the result, most viewers likely didn't really understand how that resistor ladder works. BTW, -80 db SNR is perfectly fine for consumer headphone audio. THD is more important.
@@noop9k You're right. I don't have a good understanding of electrical engineering in general and specifically how digital audio works. It's only recently I have watched some videos on DACs and it's fascinating that we have been listening to digital audio for over 40 years but we're still trying to perfect the DAC! It's amazing that there is still room for improvement after all this time.
@@MillisecondFalcon The problem is that none of the electrical components are perfect. They all react not instantaneously, don't have perfectly linear response, clock sources are not perfectly precise. And also a single capacitor and a single resistor are not a very good frequency filter and certainly not enough to have a good DAC. Math that proves that 44100 Hz is sufficient for representing all audible frequencies assumes perfect antialiasing filter. IRL really good analog antialiasing filter is a complex and expensive thing made of many basic components. To relax the requirements for analog components they use much higher frequencies (oversampling). Basically, they move some of the hard work into the digital domain, so cheaper and simpler analog components can be used..
Spent most of yesterday comparing a Denafrips Ares II to a Topping E50. I was expecting to hear no meaningful difference or hopefully some presentation difference in the Ares. Before volume matching I heard all kinds of magic in the soundstage and imaging. Just like the reviews had said. By ear I was off by 1,5dB (on the Topping L70 volume scale). After matching things changed. No difference whatsoever in either, the bass was notably weaker or rounder on the Ares and the highs were easier, maybe smoother. I measured it and there is a treble roll-off. Around -1dB@7kHz, -2dB@9kHz and -5dB@15kHz. I EQ'd the same for the Topping and the effect was the same. In conclusion, I'd take an E50 (or E30, same thing really) over the Ares. Should have tested an Apple dongle as well which was about the same as the E50.
Hi. Apart from the tonality difference and dynamics, you will most likely not find any difference in dacs using easy to drive headphones or especially iems. You will need a speaker system to notice it, or something like a susvara. That's why most dac reviews compare them using speakers. Watch jays iyagi video where he tested 100 dacs. Also, you need to test much more dacs for more experience to know what to listen to and where you need to pay attention.
@@maximgodunov7717 I've tested dozens and from the Apple dongle to Jotunheim there was really nothing worth of note when proper volume matching was done. In power amplifiers there are massive differences with the integrated DACs I've noticed, but I doubt adding room issues and degrading the signal helps with hearing the differences. Perhaps power amps strenghten the coloration or something. I also used an HE1000se which should be essentially as revealing as anything. The bass issues and treble roll-off are still notable problems.
@@7in1 Depends if one considers them a problem or desired coloration. Imo it's a massive issue. Volume matched single seating measurement from a headphone showed that and it matches my perception so I can't think of another explanation.
@@hartyewh1 I'm talking about standalone dacs used with speakers. They are way more revealing compared to headphones because they have low impedance and sensitivity. Can't head a difference in soundstage if they is no soundstage to begin with. This is how most dacs are tested by reviewers- take a look at Jay's Iyagi 100 dac video
I've come to accept that the analog output section inside the DAC that brings up the signal to 2VRMS line level out is mostly more critical than perhaps the implementation surrounding the decoding method in DACs over a certain $ value. These gain stage(s) impart more characteristic on the DAC output than you would think. Having a socket to replace factory selected opamps with discrete versions can transform a DAC's perceived performance considerably. Also allowing you to tailor to your taste. IMHO this is no different than upgrading phono section on your turntable. Perhaps turntable cartridge selection is analogous to selecting a DAC topology. These things all matter if you decide to chase that rabbit like Alice.
Ah, the very nice T+A DAC 200. Great DAC for the money, which is not meant as faint praise but rather an acknowledgment that incredibly expensive stuff is out there and while it might be slightly better, it's 10X the cost. (I have a DAC I prefer but doesn't handle my occasional DSD fare. So both are in place and the DAC 200 may be the best DSD I've heard. I've heard a fair number but not enough to be high end certified 🙂
Hey, trying to find a headphone amp to power my Supermix 4s as well as headphones (which might get upgraded later on to higher ohms) I already have a DAC, well an audio interface "goXLR mini" but it produces hiss on my IEMs and I might as well get a desk amp. I've limited or googled around to see myself against two options, Schiit Magni and Topping L30, going for 175 euros and 90 euros. Which would you recommend? Am I losing out on that much between the two to justify paying the 85+ euros?
Hey brother!Can you help me with a question since you know far more than me?I prioritize ANC and comfort and I have these options;(edifier WH950NB, sony wh750b, 1more sonoflow, anker soundcore q30 and space one)what option is better?you can sugest something else if you know 😊
This is a nitpick but I think Jetpack Joyride is a better representation of ΔΣ than Flappy Bird because Flappy Bird doesn't continue to go up when you feed it a continuous input and Jetpack Joyride is also very well known, albeit not quite as famous as Flappy Bird might be.
So if i take the audio from my monitor (xbox series x source) via 3.5mm to Y 2x Rca connectors into the topping e70 Dac , the monitor has already a Dac built in , so my dac reconverts the signals another time ? Hmm , I'm very confused , how it works
Soo uhm i need to replay this video quite a few times to understand. but from what i get. harmonic distioriton is bad. and my pc (motherboard dac) isn't good and is creating distortion. so i need an external dac/amp to make the sound better?
@BrettGlass if you hook up your PC or phone to an external DAC, the conversion happens in the external DAC, not the one inside your PC or phone. So it's not sending the signal from a poor quality DAC to a better one, it's sending the PCM data down the USB cable, or coaxial or optical cable to the external DAC. This means the sensitive DA conversion happens in an external box, which is shielded, to protect it from the electrically noisy environment inside your PC. When I was plugging my headphones into the headphone socket on my motherboard, it was using a DAC on the motherboard, and the background hiss was quite high. I then got an external audio interface, a Steinberg UR22 Mk II, this used the DAC in the interface. This fixed the hiss and background noise problem, but it's a terrible interface, with frequent audio dropouts, which causes random pops in the background. I then bought a Bifrost 2/64 which doesn't have background hiss, or audio dropouts, so I don't hear pops when watching videos or playing music. With the Steinberg I'd hear little pops and clicks when I'd start playing or pausing audio files, sometimes several in 4-5 seconds, they were faint, but sometimes they were loud. External DAC's are usually better than motherboard or phone DAC's, and some are better than others. Many people have complained about the Steinberg audio dropouts problem.
Yes, I did that from 2016 until yesterday, using my Steinberg UR 22 mk II audio interface for my microphone and as a DAC. It had much better sound than my motherboard audio, as it didn't have the high level of background hiss problem. However it's a terrible interface as it has problems with audio dropouts, which showed up as faint pops and clicks when playing or pausing video and audio files. Frequently when watching a film, I'd hear a pop, then when I'd jump back 5 seconds, the pop wasn't their, so it was an audio interface problem, not a pop on the sound file. Yesterday I hooked up a Bifrost 2/64 DAC, and I don't have the pops when playing audio or watching videos. There are audio interfaces with decent sound, which don't have the audio dropouts problem, I'm not sure which, but do recommend you avoid Steinberg.
@ thanks for the detailed reply . I have the ur24c and a cheap Samson QH4 4-Channel Headphone Amplifier . I mainly use it to power my 250 ohms headphones it works but i feel actual dac and amp stack can give better sound clarity but I’ll try in the future .
Headphones plug into a headphone amplifier. That’s what you’re hearing. Even the cheapest DAC can accurately recreate the signal copied 100 times. Any noise or distortion is from the analog headphone amplifier, not the DAC.
First they're showing 10bits (2:26) while talking about 16 bits, then (3:00) 16 bits are represented as 7 bits + 9 bits shown under those 7 bits... come on. 3:07 now we see 16 bits divided into two bytes just to confuse you a little bit more, I'm exited for the second half of this video.
It’s unfortunate and confusing that PWM is labeled as “digital” here. At no point in the flow is the high/low signal interpreted as a sequence of numbers. It’s analog throughout.
Instructions unclear, I mortgaged my house for cables. (I have no money left for headphones, speakers, DACs, or even a source, but I think I'm getting the most bang for buck, so I'm happy with my purchase decisions. If I stare hard enough at the cables I hear something that sounds incredibly accurate and also enter a trance state that I assume is essentially achieving nirvana. 10/10 would go bankrupt again.) P.S. I imagine you can send a DC, full-voltage, strictly on/off signal to a speaker but that it won't necessarily sound good, unless you like square waves and/or what I'm boldly going to call "rectangle waves". Or maybe the sound of your transducer committing sudoku.
A few important notes:
1) Whilst the video explained 1-bit delta-sigma and PDM at a basic level, most available delta sigma DACs are actually 'multibit' delta sigma. Meaning rather than modulating to high speed 1-bit, they modulate to high speed 5-7 bit as this allows you to avoid the biggest challenges with making a highly linear R2R DAC and you also don't need to be operating as fast or have as much difficulty with the modulator design as a full 1-bit DAC.
2) DACs that ARE 1-bit usually don't actually have a 1-bit PDM converter. Instead they usually use a moving-average FIR filter approach, this will be explained further in a future video.
3) Some products on headphones.com such as open box units may not fall under the 365 day return policy, please check the product page for specifics!
im the most sigma tho
4) Don't buy them from DCS /hj
@@LucaTheStar dcs is goated they doubled my debt for free
Strictly speaking DACs don't need oversampling. 16 bit signal -> 16 bit DAC -> reconstruction filter, no oversampling required. But when the DAC has less bits than the signal you need dithering to preserve the dynamic range, and when you need dithering it's a good idea to oversample to have less noise in the audible frequncies, even less with noise shaping.
@GoldenSound Did you happen to spot a DM on Discord about the KA11? I wonder if you have different info then what I had come across.
I'm a little worried that the cable joke won't be obvious enough to some people 😬
I laughed my ass off at that ahahaha
What joke?
Snake oil reference for anyone coming from outside the "audiophile" world.
@@or1on89 same lmao
Need more in depths about the cable, I'm curious.
I honestly would like a whole series of Golden just explaining stuff. He's a really great teacher and has a way to make every topic interesting
Came for the flappy bird analogy, stayed for the cable joke.
We Are Learnding.
Why no comments 😢😢😢
Peak audio UA-cam: Sigma Delta DACs being described using Flappy Bird.
Such a great way to explain PDM in less than 10 seconds to a smooth brain like me.
@@juanblanco7898 Absolutely!
Flappy bird analogy is just perfect
and a very old one, for that matter.
The best ELI5 I've heard so far. Will share to my office mate if they ask "What is that weird box on your desk?" for the fourth time.
I still don't understand, need an ELI2
The RME DAC that you showed on the picture in thumbnails and on the start of video are one of the best DACs I ever listened to, the sound it produces is so precise, it really amazed me and the functionality is very good, very well thought out (loudness and equalizer in it are really good) highly recommended if you have good amp, headphones and speakers and are willing to spend a little more money, it's trully a beauty to listen to
This was the most easy to understand explanation about the topic I've seen to date! Thanks Golden!
Now I have something to send to my friends when they start questioning my sanity...
I legitimately learned a ton from this. Some stuff was in more detail than I needed but you broke it down very well and the reap at the end was perfect. Thank you very much.
I think I’m going to try a DAC for the first time soon
Classic golden teaching us of dacs and amps.
This was an absolute treat to watch.
What a great video. I've been in this hobby for over 10 years and only NOW understand how a DAC works lol.
actually its a terrible video, find an explanation from a physicist or electrical engineer...that 1s and 0s stuff is nonsense....look up what 8 or 10 or 12.... bit means
@DD-gi6kx Care to link me something that tells me what you're on about..?
So relieved to know the $3k cable I got for my $100 headphones was the correct choice 🙏
I just spent over £240 on a 4 meter balanced headphone cable from Oidio Sound.
My current cable had worn through the outer layer at the right earcup, so they needed replacing. And I bought a Midgard headphone amp, which has 4 pin balanced output, but my current headphone cable is unbalanced with a 3.5 mm mini jack.
Balanced cables do make an audible difference by rejecting noise, this is especially important for very long cable runs. Will they sound better? Absolutely, as the placebo effect can be very strong. Would I be able to tell which cable is connected, if I don't know if it's the old or the new one? Probably not.
But I wanted a 4 m cable, so I don't need to use a headphone extension cable, and I wanted to move up from unbalanced to balanced. Most other companies selling headphone cables for my Dan Clark Audio Aeon 2 Noire either didn't sell longer than 3 m, or they were more expensive, or I'd have to pay import tax and fees.
So while I am in the, "paid more on the cable, than the device it's connected to" camp, I'm happy with my decision, and looking forward to hooking it up, and listening to balanced headphone audio.
Great video. BRB buying a $1000 cable for my fiio dac
7:26 Never expected to see Haruhi on the Headphone Show, and yet here we are.
How could i have missed that...
I wonder if that was Cameron's pic. If so, we also need a watch recommendations in addition to the listen ones in the podcast.
Watch odd taxi and Dangers in my heart. Solid choices@@YTHandlesWereAMistake
Waooo. Thanks. So great. I knew all of this, but it is the first time I have seen it described so good. Great job
10/10 video. Very clear and easy to follow
Your timing is just perfect, i just started 2 days ago to get into Headphones and sidekicks and already stumpled upon the DAC and Amplifier Problem. I'm ready to learn more! Btw, my current Headphone of Choice for my next shopping tour is the Sennheiser HD600 or 650. Chosen by your very big Tier List video from 8 Months ago ;)
Get the 660s 2 , far superior than both
Great video! Always love the goldensound knowledge drops
Vids like this are excellent as there is always someone new to the Hobby/Industry. The visuals really make it for me. I'm always having to refresh my memory as to why I want one over the other and the visuals really made it pop. Good Job!
Cameron strikes again! Thank you for this!! Looking forward to the next entry in this series.
this is awesome: accurate, informative, concise
Extremely good explanations! I'll share this next time i have the opportunity to teach anyone about DACs
This video crushes for simplicity, communication, and entertainment. Way to go GoldenSound
I presume/hope that the 3 “Biggest improvements per $ spent” list will be revisited in Part 2, with strikethough added to Cables, and replaced with Amplifier (possibly moved to #2, with Headphones #1).
Loved the joke though, for those of us who saw it for the joke it was! 😆
3:15 I just want to point out that oversampling isn't strictly necessary to reconstruct the analog signal. So long as you have a precise low-pass filter at half the sample rate, the signal reproduction will be accurate. However, oversampling allows you to control how that cutoff happens much more precisely.
I don't know exactly how common DACs that don't use any oversampling whatsoever are nowadays, but you'd find them for example in toys that include a sound module.
This is true, though the problem is it's not practically possible to make an analog filter that'd do this, hence why we typically do it digitally as we can make a very steep filter that way.
With an analog filter you'd either need to attenuate content in the audible band too much or you'd leave too much above the Nyquist frequency.
Schiit Audio makes multibit DAC's that can turn off their oversampling filter and run in No Over Sampling (NOS) mode. My Bifrost 2/64 can do that. I have also seen other companies DAC's, which have the option to do no over sampling.
I've not tried it out yet, as I only got my DAC and Midgard headphone amp unboxed and hooked up yesterday. But most people prefer Over Sampling, especially on the Schiit Audio DAC's, which use their proprietary time and frequency optimised oversampling digital filter.
for the pragmatist out there, two aspects kind of give the tone for all decisions to follow :)
a) when they measure dynamic range, noise floor, etc, it's kind of important to keep in mind what kind of values are we experiencing in the real world and
b) when they present frequency response and fq curvers it's kind of important to keep in mind our hearing is much much worse than it appears on paper.
don't let the critics, reviewers and especially 'influences' ruin your feelings towards music and sound in general. the most accomplished mastering engineer still uses the same headphones (when he uses them) he liked a few decades ago :). i'm not saying nice things are bad, but i am saying the "nicest" things are useless.
--------------------
analyzing the dynamic range of numerous music files, the actual difference between the loudest and softest sections isn't nearly as drastic as some 'theoreticians' insinuate. music doesn't sound and certainly does not feel great w/ insane 'volume swings'. not even 'explosions' do, even those are limited/gated and EQed to make them 'nice sounding'.
let's take an example that should be easy to imagine: at the concert hall, the piece starts w/ a loud cymbal crash. so from 'complete' silence we jump to the fast attack of a high fq percussion instrument at full volume. but what's the full volume, how many dBs is it above the silence? for the percussionist the sound is louder than for the conductor and much louder than for a listener positioned mid concert hall. say you want to experience what the percussionist hears for whatever reasons. then what do you do when the rest of the orchestra starts playing? move away from the cymbals or bring the whole orchestra close to your ears? :) for most of us, the most pleasant experience is to listen to the whole orchestra from a distance, give the sound enough 'room' to nicely blend together and in most cases also take the character of the acoustics of the room.
i hope this makes sense... because most recording engineers put great effort in balancing things out, then the mixing and mastering engineers refine that balance even more and in the end, we rarely (almost never) experience music that jumps in volume to the limit of the 16bits offered by the "old CDs". dynamic range is a bit different in real world than on specs sheets... your 'regular music' has less than 20dB of 'intensity range' and 'loud music' (be it EDM or metal) far less than that.
when it comes to our hearing, perfectly healthy hearing of a post puberty human, none of us have the fq range everyone quotes, 20 to 20000Hz. it just doesn't happen and that's perfectly fine because again, analyzing the fq range of numerous music files, there isn't much information in the last half octave of the audible spectrum. if you're over 30yo, anything above 15kHz is pretty quiet but you're not missing much at all as counterintuitively, 'clarity' in music (recordings) is not defined by those high fqs. in fact, at such high fq our ears have a difficult time asserting pitch, location and even intensity. as the wavelength of a sound becomes shorter and shorter we're simply not equipped to properly analyze it.
for instance, if you were to play a 15kHz pure tone from a speaker and would slightly move your head any direction you'd hear tremendous fluctuations of not just intensity but also timbre as the sound bounces around your head, ears and everything around you; you'd be picking up all sorts of 'interferences' from all those bouncing 'sound beams' and the fluctuations are drastic for just a few mm change in position.
have you ever wondered w/ is it so difficult to locate a cricket that made its way inside your room? because we're not owls :):), we didn't need that sort of 'accuracy' and our hearing didn't care to develop in that way. although we can hear high fq sounds, our brains more or less ignore most of the data those sounds carry.
(old men installing super tweeters is super funny -until they attempt to convince you you're not experiencing the Beatles at their full potential if you don't own ultrasonic transducers, then it becomes annoying at best.)
ps: there's a good reason why not all studios use the same monitors/headphones, why not all artists use the same make/model of in-ears monitors... everything is a compromise and nothing can be objectively accurate in sound reproduction. they key is to find a balance that pleases You. and when it comes to 'accuracy', being familiar w/ a certain device yields far finer precision than 'the latest' or 'most expensive' model of whatever device. that's what sound engineers concluded since forever and i'm talking about ppl who can and did hear a bolt being lose on the drum kit of a loud metal band, not your average youtuber :)
Superb explanation! I understand this, but if I ever need to explain it to others, this is great!
Bro is a scientist
What a brilliant explainer - thanks GS! Looking forward to part 2. No pressure 😉😁
Fantastic explanations Golden. You struck a great balance of simplicity and accuracy.
Actually brilliant description. Thank you! Science communications is clearly your forté.
This video was very confusing as I failed at math and science when I was in high school in the 90's and have learning disabilities but I was interested in it, I just had to replay and rewind a lot to try to follow and understood it vaguely.
This Video was amazing. Thank you!
I need to watch this again after I haven't had a couple of beers! I'm more confused now than I was before I saw this.
Wonderful and very useful video!!
Nice! Good to see a clear explanation .... although a bit fast paced.
you are just slow, i watched this at 2x speed LOL
Amazing stuff this
Explaining PWM with flappy bird is freaking genius! hahahaha
I bought a JVC CD player in 1995 that had "1 bit DAC" as one of its main features and selling points. It was even printed on the front of it beside the CD tray. Knowing a little bit about binary (no pun intended, or was it!) I thought at the time it was an unusual property to boast about, because 1 is the least number of bits you can have. Obviously I didn't understand how the DAC worked and the significance of "1 bit DAC". Now I do!
You still don't understand much. This video is very oversimplified, though it gets the core concepts right. But, the people who need such a simple explanation, probably don't know Ohm's law even and, as the result, most viewers likely didn't really understand how that resistor ladder works.
BTW, -80 db SNR is perfectly fine for consumer headphone audio. THD is more important.
@@noop9k You're right. I don't have a good understanding of electrical engineering in general and specifically how digital audio works. It's only recently I have watched some videos on DACs and it's fascinating that we have been listening to digital audio for over 40 years but we're still trying to perfect the DAC! It's amazing that there is still room for improvement after all this time.
@@MillisecondFalcon The problem is that none of the electrical components are perfect. They all react not instantaneously, don't have perfectly linear response, clock sources are not perfectly precise. And also a single capacitor and a single resistor are not a very good frequency filter and certainly not enough to have a good DAC.
Math that proves that 44100 Hz is sufficient for representing all audible frequencies assumes perfect antialiasing filter. IRL really good analog antialiasing filter is a complex and expensive thing made of many basic components. To relax the requirements for analog components they use much higher frequencies (oversampling). Basically, they move some of the hard work into the digital domain, so cheaper and simpler analog components can be used..
@@noop9kand that's why they reckless audio at 192 or 96khz and then digitally resample to a fraction of that with a better filter.
That was brilliant!
Truly Great explanations, and you made me laugh 3 different times.. Great video !
Spent most of yesterday comparing a Denafrips Ares II to a Topping E50. I was expecting to hear no meaningful difference or hopefully some presentation difference in the Ares. Before volume matching I heard all kinds of magic in the soundstage and imaging. Just like the reviews had said. By ear I was off by 1,5dB (on the Topping L70 volume scale). After matching things changed. No difference whatsoever in either, the bass was notably weaker or rounder on the Ares and the highs were easier, maybe smoother. I measured it and there is a treble roll-off. Around -1dB@7kHz, -2dB@9kHz and -5dB@15kHz. I EQ'd the same for the Topping and the effect was the same.
In conclusion, I'd take an E50 (or E30, same thing really) over the Ares. Should have tested an Apple dongle as well which was about the same as the E50.
Are you claiming that Denafrips DAC is having such huge FR problems that even you could measure them? It's hard to believe to be true
Hi. Apart from the tonality difference and dynamics, you will most likely not find any difference in dacs using easy to drive headphones or especially iems. You will need a speaker system to notice it, or something like a susvara. That's why most dac reviews compare them using speakers. Watch jays iyagi video where he tested 100 dacs. Also, you need to test much more dacs for more experience to know what to listen to and where you need to pay attention.
@@maximgodunov7717 I've tested dozens and from the Apple dongle to Jotunheim there was really nothing worth of note when proper volume matching was done. In power amplifiers there are massive differences with the integrated DACs I've noticed, but I doubt adding room issues and degrading the signal helps with hearing the differences. Perhaps power amps strenghten the coloration or something. I also used an HE1000se which should be essentially as revealing as anything. The bass issues and treble roll-off are still notable problems.
@@7in1 Depends if one considers them a problem or desired coloration. Imo it's a massive issue. Volume matched single seating measurement from a headphone showed that and it matches my perception so I can't think of another explanation.
@@hartyewh1 I'm talking about standalone dacs used with speakers. They are way more revealing compared to headphones because they have low impedance and sensitivity. Can't head a difference in soundstage if they is no soundstage to begin with. This is how most dacs are tested by reviewers- take a look at Jay's Iyagi 100 dac video
Can I use my phone as a music source on a dac with my headphone amp?
Most excellent...!!! Great video and information...!!! Well done...!!!
This was fun to watch. Thanks!
I've come to accept that the analog output section inside the DAC that brings up the signal to 2VRMS line level out is mostly more critical than perhaps the implementation surrounding the decoding method in DACs over a certain $ value. These gain stage(s) impart more characteristic on the DAC output than you would think. Having a socket to replace factory selected opamps with discrete versions can transform a DAC's perceived performance considerably. Also allowing you to tailor to your taste. IMHO this is no different than upgrading phono section on your turntable. Perhaps turntable cartridge selection is analogous to selecting a DAC topology. These things all matter if you decide to chase that rabbit like Alice.
Ah, the very nice T+A DAC 200. Great DAC for the money, which is not meant as faint praise but rather an acknowledgment that incredibly expensive stuff is out there and while it might be slightly better, it's 10X the cost. (I have a DAC I prefer but doesn't handle my occasional DSD fare. So both are in place and the DAC 200 may be the best DSD I've heard. I've heard a fair number but not enough to be high end certified 🙂
I bought a ayima a5 pro max its pretty solid on my focals
This is really helpful!
I have an Anthem AVM-90
Which has an excellent DAC and an EverSolo-A6-master Edition. Do I need another DAC? Thx!
Excellent Video!
If the external DAC is receiving a distorted signal from the laptop then how will it improve the signal if the source is already bad?
Hey, trying to find a headphone amp to power my Supermix 4s as well as headphones (which might get upgraded later on to higher ohms)
I already have a DAC, well an audio interface "goXLR mini" but it produces hiss on my IEMs and I might as well get a desk amp. I've limited or googled around to see myself against two options, Schiit Magni and Topping L30, going for 175 euros and 90 euros. Which would you recommend? Am I losing out on that much between the two to justify paying the 85+ euros?
Awesome video
What DAC (r2r or delta sigma) is producing more true digital signal from original analog, and why?
Can I request for a part 2 video on DAC filters please?
What is the connection to the phone or computer? USB? Bluetooth? Need better cables?
What a great and well explaind video.
amazing explanation, thank you
Thanks it's a good one especially cable
What do you actually think about rme dac and amp?
Well Done....🎶
RME ADI series is 🐐
Yeah they are the best, it's sound is so precise
well done! fun and educational - tough to do.
Hey brother!Can you help me with a question since you know far more than me?I prioritize ANC and comfort and I have these options;(edifier WH950NB, sony wh750b, 1more sonoflow, anker soundcore q30 and space one)what option is better?you can sugest something else if you know 😊
This is a nitpick but I think Jetpack Joyride is a better representation of ΔΣ than Flappy Bird because Flappy Bird doesn't continue to go up when you feed it a continuous input and Jetpack Joyride is also very well known, albeit not quite as famous as Flappy Bird might be.
That flappy bird game was great! 👍🏼
So if i take the audio from my monitor (xbox series x source) via 3.5mm to Y 2x Rca connectors into the topping e70 Dac , the monitor has already a Dac built in , so my dac reconverts the signals another time ? Hmm , I'm very confused , how it works
Soo uhm i need to replay this video quite a few times to understand. but from what i get. harmonic distioriton is bad. and my pc (motherboard dac) isn't good and is creating distortion. so i need an external dac/amp to make the sound better?
Love the flappy bird analogy!
Wonderful video
I think I missed something...Why run the signal from a poor DAC to another DAC? Or do you somehow bypass the original some how???
@BrettGlass if you hook up your PC or phone to an external DAC, the conversion happens in the external DAC, not the one inside your PC or phone.
So it's not sending the signal from a poor quality DAC to a better one, it's sending the PCM data down the USB cable, or coaxial or optical cable to the external DAC.
This means the sensitive DA conversion happens in an external box, which is shielded, to protect it from the electrically noisy environment inside your PC.
When I was plugging my headphones into the headphone socket on my motherboard, it was using a DAC on the motherboard, and the background hiss was quite high.
I then got an external audio interface, a Steinberg UR22 Mk II, this used the DAC in the interface. This fixed the hiss and background noise problem, but it's a terrible interface, with frequent audio dropouts, which causes random pops in the background.
I then bought a Bifrost 2/64 which doesn't have background hiss, or audio dropouts, so I don't hear pops when watching videos or playing music.
With the Steinberg I'd hear little pops and clicks when I'd start playing or pausing audio files, sometimes several in 4-5 seconds, they were faint, but sometimes they were loud.
External DAC's are usually better than motherboard or phone DAC's, and some are better than others. Many people have complained about the Steinberg audio dropouts problem.
Predovic Point
Can I use audio interface as a dac ?? And connect it to an amp and call it a stack ???😅?
Yes, I did that from 2016 until yesterday, using my Steinberg UR 22 mk II audio interface for my microphone and as a DAC.
It had much better sound than my motherboard audio, as it didn't have the high level of background hiss problem. However it's a terrible interface as it has problems with audio dropouts, which showed up as faint pops and clicks when playing or pausing video and audio files.
Frequently when watching a film, I'd hear a pop, then when I'd jump back 5 seconds, the pop wasn't their, so it was an audio interface problem, not a pop on the sound file.
Yesterday I hooked up a Bifrost 2/64 DAC, and I don't have the pops when playing audio or watching videos.
There are audio interfaces with decent sound, which don't have the audio dropouts problem, I'm not sure which, but do recommend you avoid Steinberg.
@ thanks for the detailed reply . I have the ur24c and a cheap Samson QH4 4-Channel Headphone Amplifier . I mainly use it to power my 250 ohms headphones it works but i feel actual dac and amp stack can give better sound clarity but I’ll try in the future .
I'm not dumb. I'm passive.
😂
Why R2R is praised more than Delta Sigma, if both are nicely doing same job but different style, isn't there should be the same sound?
Great video
Can you use this to meet blokes on grindr?
Thank you
Headphones plug into a headphone amplifier. That’s what you’re hearing.
Even the cheapest DAC can accurately recreate the signal copied 100 times.
Any noise or distortion is from the analog headphone amplifier, not the DAC.
Fantastic,thank you 👍👍
Drivers are smooth-brained. That's why I get along with them so well. 😌
love this video
Cables, you almost got me rage baited there 🤪
First they're showing 10bits (2:26) while talking about 16 bits, then (3:00) 16 bits are represented as 7 bits + 9 bits shown under those 7 bits... come on. 3:07 now we see 16 bits divided into two bytes just to confuse you a little bit more, I'm exited for the second half of this video.
Excellent
1:40 Based Pixelzwithaz reference ❤
Brilliant!
Worth upgrading from Topping D90 to Dcs Lina??
GOOD ONE /< DCS SUCKS 💩💩💩💩💩💩💩💩💩💩💩💩💩💩💩💩
✅️🔥🚦💥💯💯💯💯💯💯💯💯💯
Good one DCS IS 💩💩💩💩💩💩
Buy RME Adi series DAC that's one of the best for the price
It’s unfortunate and confusing that PWM is labeled as “digital” here. At no point in the flow is the high/low signal interpreted as a sequence of numbers. It’s analog throughout.
great vid!!
will we cover up the why is the cables also have its own price?? 👀
"i MeAN theY aRE JuSt cabLeS" im looking forward for it, cheers.
SMW2 ❤️
Haha... I love that Cables joke.
Instructions unclear, I mortgaged my house for cables.
(I have no money left for headphones, speakers, DACs, or even a source, but I think I'm getting the most bang for buck, so I'm happy with my purchase decisions. If I stare hard enough at the cables I hear something that sounds incredibly accurate and also enter a trance state that I assume is essentially achieving nirvana. 10/10 would go bankrupt again.)
P.S. I imagine you can send a DC, full-voltage, strictly on/off signal to a speaker but that it won't necessarily sound good, unless you like square waves and/or what I'm boldly going to call "rectangle waves". Or maybe the sound of your transducer committing sudoku.
The speed that a motor turns is also a Sample Rate Unfortunately this point has been misstated
And cables lol good one
So in short: Some DACs are actually DADs: Digital to (another type of) Digital converter. Thanks delta sigma.
I'm this video, it sounded like GoldenSound was speaking from within the physical body of DMS.