Excellent overview, Craig. I've personally never had a problem with the sound of well-mastered 16-bit, 44.1 kHz audio -- the key phrase being *well-mastered,* not the garbage the record companies have been putting out ever since the "CD Loudness War"!
Yes, and storage capacity, bandwidth, and processing power, which are less of an issue these days, but were definitely more of one earlier on. Also, tech guys can get some pleasure finding ways to compress information at the cost of clarity - some forms of radio are a good example with literal bandwidth restrictions and further limitations. Compression techniques have improved a lot as well, but often times an engineer's job is to find something "good enough," though perhaps not ideal for the end consumer. Or even worse, sometimes quality is intentionally limited to maintain certain licensing rights or proprietary incentives. Ironically, some of these earlier "lo-fi" type distortions are now intentionally used as sound effects for a retro type style, for example some synthwave stuff.
20:38 *Vinyl TV, VWestlife* On loud (loudness war) recordings you can hear the difference between 48kHz and 96kHz, the upper frequencies on a very loud recording removes the harshness and narrow sound. This is WHY Hi-Res audio exists, to remove the narrow sound as best as possible giving you more headroom and space enough to compensate more fidelity gone 48kHz and surface noise to none. Unfortunately, the loudness war came to stay forever, even worse because radio volume standards won't move back to the past in time like. It's like going back from today's smartphones to first-gen cellphones to them.
Great video. Also remember that 6db of sound is represented for each part of the bit depth - so 16 bit x 6db = 96db range. 24 bit x 6db gives you 144 db range of sound.
God I love this video and this channel, so instructive, can't wait for your next upload ! Could you do a video on how you setup your digital music library, maybe with Foobar ? Could be very interesting to compare with my personal setup :) Kind regards from France
Thank you, I learned something today. My sound card can do up to 32/96. I have been recording my LPs at 24/96, and 45s and 78’s at 24/48. I now understand 24/96 is just a waste of space. I wasn’t aware the sampling rate was 1/2 of the frequency range. I’m 60, and always have to boost the treble significantly because the high end gets lost for me. I will stick to 24/48. Cd quality sounds fine for me, though.
Craig, great job. I did a hi res/digital sampling video where I let people listen to a vinyl rip at 24/96 and brought it all the way down to 192k mp3 - and only one person could hear a difference between the CD rate and the mp3, let alone the hires! Its why I use 44.1/24bit (since I make a CD backup, and feel why process from 48 to 44.1 if I'm going to CD anyway)...
Hi Craig! Great channel, I just subscribed last week. There is an audible difference between 48 kHz and 96 kHz, but it is not because of those inaudible frequencies above 16 kHz; it is because it just has more sense to take 96k “frames“ per second of a thing moving (for example) 12k times per second (12kHz should be audible for all of us). I´ve also worked with 192kHz files, but at that level the difference only became important when you start to apply dynamic processes, plug-ins, or any digital filter. This is a huge topic and it´s so easy to get yourself into endless arguments with other fellows... so in my personal opinion: 48k vs 96k IS audible, but not THAT audible, you should consider the size of the files you are storing in your drive and the device you are going to use to playback those files; as personal advice for the community, it is WAY more audible if you invest in a good preamp or an interface (if you are interested, search for internal clock´s topics), it can be expensive, but if you are sort of an audiophile it should be worth of your money. About the interface, I personally love RME´s Fireface UCX, but there are more options out there.
Hi! The difference pops out depending on your monitoring system and your ADDA converter, I´ve worked on recording and post quite a few years and there is an audible difference (if you have an adequate monitoring). I wrote frames between “ “ as an example meaning samples, and as you said (an I agree) it is useful for digital filters, but increased headroom is more related to bit depth. I am sure your opinion is well backed up, and I respect it. On my experience there is a difference and a completely different workflow managing 96kHz files that has nothing to do with “perception“ or even “placebo effect“. But as I also said, my point it´s not to discuss this, but focus on your chain of equipment. It will always sound WAY better if you get a great signal from your cartridge and preamp; and then, if it´s possible to made the conversion with a strong clock interface (like the one I mentioned) you should be able to achieve great sounding files. In case of recording a copy of one of my records, I think the best is 24 / 48kHz. But in case of recording anything else to work with afterwards, probably you should aim higher if your interface allow you. Always work on getting a strong signal and it will sound great! Regards!
Going back to what this video was about... 24 / 48 is more than enough for archiving vinyl. Considering that bit depth is directly related to dynamic range, a vinyl record might be equivalent to an 8 bit recording, or 12 on a good day. I would say that 16 bit is plenty, but 24 gets you out of the dog house in the community, and you sleep better. But as hard drives get cheaper, it certainly doesn't hurt to forward think, and allow lots or room for vinyl to do its thing. Cheers!
I personally contacted him, I too hope he's OK and asked if he planned on any new material. This guy is one of the few who really explains things in an understandable way and conveys why I like vinyl recordings.
With the higher frequencies above normal hearing range, some people have said that they can produce more natural harmonic renounces for a fuller and more realistic sound. I think there might be something to it, as I know you can get various wave interference patterns, but I couldn't say for sure one way or the other what affect it actual has on the overall quality and experience of listening to music, if any. I do think it's perhaps something to keep in mind though.
This is the best video on this topic.. I’ve been digitizing my vinyl records 24 bit 48. However I never knew the difference with this bit rate.. 16 bit 41 is good? I am trying that out to see if I hear a difference.. thank you again for this video..
Great Video, what about the effect 2nd and 3rd harmonics have on the audible frequency range? I've read / watched videos showing audio content (and not just noise) on a vinyl LP going up way passed 25,000Hz. So with a good cartridge 24bit @ 48KHz seems like a better option. Many instruments produce sound over 20,000Hz, and good mics go above this level as well, so an analogue format should also reproduce this (if the cartridge and kit is able too).
interesting, thanks. re hearing different sample rates. when a filter is applied to a stereo file it distorts (tiny micro) to both bass and hi end. a by product is spikes or ripples across the harmonics within human hearing ranges. its subtle but when someone points it out you may then hear it. guess it depends on individual, im always lost in enjoyment of the piece not discrete artifacts. but i think thats whats going on.
A friend of mine believes 64kHz sampling is equivalent of vinyl quality, he hates 44.1. He recorded the council inspector using mic,tape deck to preamp the mic to computer sound card, the PC was set to 44.1 and when it was played back it sounded metallic . He didn't have enough time to set it to 48 or whatever
good video, not just because it's informative but i also like your way to approach some of the audiophile voodoo talk that can be found all over the internet. i bought a turntable with built in usb, wich also means it has a pre-amp built in wich is very handy but also very versatile because the chinch connectors can be set to output at line OR phono level (never need to worry if the amp has phono input or not) it also allows me to record a vinyl on my computer AND my tape deck at the same time, or connect a seperate audio interface should i ever want to go higher than 16/48 wich is the highest the built in usb will give me. (and totally enough if you ask me) maybe i could borrow such an interface to record two files in different quality from the exact same spin and then do the phase inversion thing for comparsion. anyways ... the recording is done with plain old audacity and then stored as FLAC and eventually transcoded to mp3 as well if i don't have this particular album in digital form already. some recent releases have the vinyl AND a CD in one box wich in my opinion is exactly the way to do it nowadays: the vinyl version because it's vinyl and a CD thrown in as a bonus for ease of use (even if it only comes in a simple cardboard sleeve and not in a full on fancy jewel case or digibook, but if i would have wanted that i could have bought the CD release to begin with)
A lot of people (including me), can hear soft sounds below the noise floor! If they measured vinyl below the noise floor, then the dynamic range specs would be higher than CD'S! But some people can't hear below the noise floor, thus the specs they give it!
Enjoyed this video My set up is the same. Preamp to AV / USB I will probably change my settings now and listen for the any change in quality Do you use any dolby or declicker or other settings. Hope your well
As long as the frequency response can comfortably encompass the frequency range of your source you should be fine, otherwise you may very well get audible artefacts even though you are dealing with frequencies that the human ear can't hear.
Does anyone have experience with having this done professionally / as a service? How are the results, cost etc? It is not because I wouldn't wanna attempt it myself; I am just thinking that if I do this, I want a really good result and of course right there, it would be limited by what my personal TT can deliver...
I don't normally put music on my computer and I wouldn't give you a $ a dozen for all the MP3's in the world, man. I mostly listen to Classic Rock on Sirrus in the car. But, "vinyl is final". GB
Most good software internally capture audio at 32bits precision due better rendering/downsampling or using declicker/denoiser etc... even if vinyl has around 70db dynamic range that is near close to 16bit. While saving audio need to be dithered properly if saving 32->24 or 32->16 bits and that is more crucial than other things. And setting samplerate more than 48khz is absolutely unnecessary and big waste of space... I strongly recommend to capture your vinyls using good software like iZotope RX or Adobe Audition, because other can harm your audio even more...
Actually you can't really record in 32bit depth as converters use 24bit (or 16bit for older converters). 32 bit are only used inside digital audio workstations for processing audio but you are never gonna be able to playback audio using all those bits as they are reduced to 24bit in DA converters and if they exceed that you're going to hear distortion. In fact if you select 32bit for recording an analog source like a turntable or a guitar you're just getting 24bits of recording plus 8 empty bits which are only good if you like wasting your hard drive storage space. Also for the (roughly) 70dB of dynamic range that vinyl is capable of delivering 24bit isn't probably worth it, as 16bit depth is capable to deliver 90 to 100 db of dynamic range.
I can only go up to 16bit 48K on my USB in direct from my table. Good enough for casual listing, I'm sure, but is it good enough for achieving and science? Hmm... probably not for me.
I wouldn't have a problem with it. 16 bits will give you much more dynamic range than any vinyl record can produce, even with the best turntables out there. You just can't drag a stylus along a groove and get as much dynamic range as 44.1 can give. I don't care what cartridge you have. Just try to record as loud as you can without going over. If you record low at 16 bit and then normalize, you do reduce your dynamic range quite a bit.
The way I understand it, the bit depth is basically the resolution or accuracy of the audio within its full potential range (per sample), so I would assume that a higher bit depth would theoretically always result in a more accurate digital representation of an analog source (and likely digital conversions as well). Of course there would come a point where having a higher resolution wouldn't result in a distinguishable quality gain, contending with noise, margins of error, human hearing potential, etc. True analog though would theoretically be infinite in its resolution, well, perhaps down to quantum levels. Either way, it's good to know that you think 16bit is sufficient as long as the full loudness range is properly utilized and I'm not cutting off dynamic range potential. (You seem to have a pretty good ear for this sort of thing and real world experience.) My Windows 7 computer can "listen" to the USB turntable source (I have the AT-LP120-USB as well) at up to 16bit 48K for recording into Audacity (I'm not sure if this is an OS, driver, or USB limitation - I might try this on my Debian/KDE laptop too), then for file encoding archiving, I'm using FLAC level 8 compression, 24bit 48K. Kind of wish the USB would go to 24bit as well for a better match (weakest link in a chain, and all that), but the USB out actually sounds better (brighter and with higher clarity) to me than the preamp out on this table, so that's why I'm sticking with it instead of just using a regular line-in on my computer. The noise on the USB with the table connected and powered is surprisingly low as well, which is nice, and perhaps more significant overall than the 16bit limit. 16bit, 48K is listed as being "DVD quality," at least. Anyway... Looking forward to a followup video to this if you end up making one. I'm curious to see how you go about all this more on the software, recording, and encoding side of things. Cheers.
Many people need to realize that the insanely high sample rates are complete nonsense. I looked at a 24/192 song that I down sampled to 24/48 and there was absolutely no difference. Looking at the frequency spectrogram of the music revealed that more than three quarters of it were not being used and just taking up memory, the more people know what proper settings to use the more space they'll save on their hard drives. Also I don't mean to sound rude but why do your eyes constantly move through your videos?
Sounds like it was a bad mastering or "up conversion" of some sort then. I can make gigabytes worth of nothing too, and have in fact to use as testing loads, but that's a different matter.
@@ZANTHERA It is funny you should say that something that is actually being done cannot be done ;) Unlike what you say, it is not complete nonsense. Of course nobody thinks they just magically created more information that was not there to begin with. The reason for up-sampling is that the filters that get applied e.g. by your CD player in processing CD audio are responsible for the harsh etc sound people complain about when comparing CD to vinyl. Now, up-sampling does not make the music better (as you say, of course), but when processing the signal, the impact of filters by your CD player gets moved out of the audible range for human ears. That is why they sell you CD players that do up-sampling to 192. In the 80s, early 90s, powerful chips were not available and/or too expensive for the mass market, but today it is not an issue anymore. So replacing a CD player from 20-25 years ago will make a real impact (or any older tech player). E.g. I still have a Musical Fidelity CD1 from early/mid 90s, which had cost 2000 bucks back in the day, and if you compare that to a current player of the same price range, they beat the pants off of it. Shame, last time I checked, they were throwing MF CD 1s for like a 100 bucks at you on Ebay... I recently paid almost as much for a replacement laser unit :) (I don't use it, I just keep it because it looks awesome). At the same time in the 90s, I had also bought a pre-owned Linn Basik turntable, which I still have :) Of course I agree with you that having such up-sampled files laying about on your hard drive makes no sense...
96 gives you more than twice the frequency response needed for human hearing at birth. I guess that's what confuses me. Why do we need all that extra frequency range when 48 still gives well above human hearing. Has anyone ever been able to explain this?
Hi Craig, I, like you have been a musician since the 1970's and a home recordist and songwriter since the 1980's. My most recent recording set up uses a MOTU 896 digital interface which is capable of recording 8 channels at up to 24bit 96KHZ sample rate, Logic Audio studio, a MacBook pro and an external hard drive. I usually record my bands at 24bit 44.1KHZ due to not wanting to tax my lap top with plug ins. However, I did an experiment with all the settings on the MOTU 896 at various sampling rates. I monitored on a £50 (approx $50) pair of Senheiser headphones that I know very well. I used a Fender Malibu acoustic guitar, and four microphones plugged into the MOTU 896 ; Rode NT-1 condensor mic, Superlux "pencil style" drum overhead condensor mic from my drum mic kit, a SM-57 dynamic mic and a Tandy (Realistic) PZM mic. On an acoustic source, my guitar and voice, the change is sound quality was quite marked. At the higher sample rates you could hear more of the room sound "the air", the acoustic space in the recorded sound and at 24/96 it was as if the headphones disappeared and it was more like hearing myself play and sing live in the room and not from a recording. There was no additional high end apparent, because as you quite rightly say we cannot hear this directly. However, I have read what happens is that we perceive the higher frequency range because the higher harmonic octaves that we can't hear directly, of the sound recorded interact with the frequency range we can hear. The effect of this is you can hear the qualities of the acoustic space and the "air" due to the higher harmonics captured at 96KHZ. This is not a subtle change, it is quite striking, even on a fairly budget pair of headphones. I don't think 96khz would make as much of a difference when recording a loud, close miked rock band, but on acoustic instruments it does. That is also why I record my band at 44.1KHZ. Plus, I would need a high quality acoustic recording room to make the most of 96KHZ. I may do it as an experiment and then mix down to CD quality to see if there is a difference in the mixed track. These higher harmonic octaves that 96KHZ and 192KHZ capture are present in analogue recordings from the 1960's onwards and also on vinyl. So for capturing vinyl sound in all it's glory I would recommend using the higher sampling rate. This is also why, I believe, vinyl sounds better than CD. Because the analogue reproduction captures the sound of the "air" and acoustic space. Whereas, the limited range of CD leaves a somewhat "flat" one dimensional sound. I have read interviews with recording engineers where they claim that even 192KHZ makes a difference, albeit a diminishing return from 44.1 KHZ, And having tried 96KHZ recording myself, I believe them.
Hey John, thanks for your detailed explanation. I like detail. I fully understand and agree with what your talking about with the ultra-sonic frequencies interacting with the sonic frequencies we hear. If two frequencies above 20khz, but very close together, both in frequency and volume interact, they can cause a "beat note" or sub harmonic. The question is... What is the precise sampling rate needed to allow this to happen just above our range of hearing? At some point, even the sub harmonics created at some sampling frequency would be above our range. Notes/tones have to be fairly close together to cause peaks or troughs that might generate sub sonic tones we can hear. So there you have the number one question that needs to be answered with math. But the one that I always go to... Placebo. It is scientifically proven that our perception of things can definitely differ, depending on what we might expect. This has been so strongly proven, that it's used all the time in blind tests for medications. Unless a test, even with audio, is done in a blind manner, the results are totally invalid for study, albeit proof of the placebo effect. I would be overwhelmingly shocked and impressed if I witnessed, in person, a person who could tell the difference between 48 and 96 without knowing which was which. And the test has to be done many times so an average can be taken. Someone guesses it once, that's not enough. I have read a lot about it too, and I haven't really looked hard enough to find any peer-reviewed studies that conclude that we can perceive, or be affected by anything over our highest hearing threshold. I have read that subjects who were put to such a test between high bitrate mp3 and CD, came out in favor of the CD in regard to emotional response, not sound quality. That might be something to look into more. But for me, I never trust an A/B comparison unless A and B are unknown to the subject being asked, and the more tests, the more "bit depth" we have... :) No audio storage method is transparent, but our ears are limited. Maybe one day we will be able to store audio in it's pure analog form, with no loss. Only then, will this subject be solved. I'm thinking Plank Level Storage, or PLS. I have a Patton..... /Kidding :)
All I can say is that I tried the recordings myself at different sample rates up to 96KHZ. I was not expecting to hear any difference, I was quite sceptical, to be honest, but the change in the audio with respect to the sense of hearing the acoustic space was quite striking and surprising. It was far too obvious to be put down to "Placebo". Why do you think audio interfaces go to 96khz and now 192 or beyond?? Because the improvement in audio quality, with an acoustic source, is quite obvious. It is not a subtle enough change to be subjective. It's quite obvious! After hearing this, you then realise that CD is a quite restricted audio format. Better than MP3 yes, but not true HiFi. CD was set at 16 bit and 44.1KHZ because they wanted to fit Beethoven's 9th, which is 77 minutes long. 16/44 was the best resolution they could manage with early 1980's digital technology. It's not the best quality that can be achieved from Digital audio. Try it yourself with a good quality Mic and an acoustic source, your voice or a guitar or percussion. Distance mic by some inches and you will hear as you go up the sampling rate that the room sound, dept and sense of the "air" is more apparent. And as for which sampling frequency to stop at, I can't answer that question by my own experience but I guess it would be a case of diminishing returns. There was an interview by a well known recording engineer who talked about 192 and even 384KHZ on the Sound On Sound magazine website.
Excellent overview, Craig. I've personally never had a problem with the sound of well-mastered 16-bit, 44.1 kHz audio -- the key phrase being *well-mastered,* not the garbage the record companies have been putting out ever since the "CD Loudness War"!
Yes, and storage capacity, bandwidth, and processing power, which are less of an issue these days, but were definitely more of one earlier on. Also, tech guys can get some pleasure finding ways to compress information at the cost of clarity - some forms of radio are a good example with literal bandwidth restrictions and further limitations. Compression techniques have improved a lot as well, but often times an engineer's job is to find something "good enough," though perhaps not ideal for the end consumer. Or even worse, sometimes quality is intentionally limited to maintain certain licensing rights or proprietary incentives. Ironically, some of these earlier "lo-fi" type distortions are now intentionally used as sound effects for a retro type style, for example some synthwave stuff.
VWestlife agree, loudness war is destroying music!
They sound just as bad as TV ads DR of around 3dB
20:38 *Vinyl TV, VWestlife* On loud (loudness war) recordings you can hear the difference between 48kHz and 96kHz, the upper frequencies on a very loud recording removes the harshness and narrow sound. This is WHY Hi-Res audio exists, to remove the narrow sound as best as possible giving you more headroom and space enough to compensate more fidelity gone 48kHz and surface noise to none.
Unfortunately, the loudness war came to stay forever, even worse because radio volume standards won't move back to the past in time like. It's like going back from today's smartphones to first-gen cellphones to them.
Can't wait to see Part 2, I hope it's coming soon? I have some vinyl now that I would love to have archived. Thanks again Craig.
Man your videos are so informative that I love it!! Everything in a simple way and understandable. Keep the good work!!
Great video. Looking forward to watching part 2.
Thank you for the in depth explanation of all those confusing numbers. Can't wait for part 2.
Great video Craig, the first one I see of yours.... concise, precise and I love the drumstick-pointer!!...subscribed!
Cheers Craig, That was really brilliant.
This is a great channel - keep it going ;)
You really helped me alot with the video on balancing the tone arm & stylus pressure.
Really a great topic ! And thank you for your channel as well !.. Now we need part 2 :-)
Great video Craig. Really informative. you taught me more than music today. :)
This is a very instructive an amazing video! Thanks!
Great video. Also remember that 6db of sound is represented for each part of the bit depth - so 16 bit x 6db = 96db range. 24 bit x 6db gives you 144 db range of sound.
Excellent and crystal clear explanations, I just hope nothing happened to that person ! Waiting for the rest...
Awesome video! Thanks man, you really explain it well! just as if you were talking to me personally.
God I love this video and this channel, so instructive, can't wait for your next upload ! Could you do a video on how you setup your digital music library, maybe with Foobar ? Could be very interesting to compare with my personal setup :)
Kind regards from France
Thank you, I learned something today. My sound card can do up to 32/96. I have been recording my LPs at 24/96, and 45s and 78’s at 24/48. I now understand 24/96 is just a waste of space. I wasn’t aware the sampling rate was 1/2 of the frequency range.
I’m 60, and always have to boost the treble significantly because the high end gets lost for me.
I will stick to 24/48. Cd quality sounds fine for me, though.
Really enjoy your videos. Looking for part 2 also :) Hope all is well. Let us know please :)
Great, thank you!
Craig, great job. I did a hi res/digital sampling video where I let people listen to a vinyl rip at 24/96 and brought it all the way down to 192k mp3 - and only one person could hear a difference between the CD rate and the mp3, let alone the hires! Its why I use 44.1/24bit (since I make a CD backup, and feel why process from 48 to 44.1 if I'm going to CD anyway)...
Most of my recordings were done at 48/24 as I felt more was just wasted space.
Hi Craig! Great channel, I just subscribed last week.
There is an audible difference between 48 kHz and 96 kHz, but it is not because of those inaudible frequencies above 16 kHz; it is because it just has more sense to take 96k “frames“ per second of a thing moving (for example) 12k times per second (12kHz should be audible for all of us). I´ve also worked with 192kHz files, but at that level the difference only became important when you start to apply dynamic processes, plug-ins, or any digital filter.
This is a huge topic and it´s so easy to get yourself into endless arguments with other fellows... so in my personal opinion: 48k vs 96k IS audible, but not THAT audible, you should consider the size of the files you are storing in your drive and the device you are going to use to playback those files; as personal advice for the community, it is WAY more audible if you invest in a good preamp or an interface (if you are interested, search for internal clock´s topics), it can be expensive, but if you are sort of an audiophile it should be worth of your money.
About the interface, I personally love RME´s Fireface UCX, but there are more options out there.
Hi! The difference pops out depending on your monitoring system and your ADDA converter, I´ve worked on recording and post quite a few years and there is an audible difference (if you have an adequate monitoring). I wrote frames between “ “ as an example meaning samples, and as you said (an I agree) it is useful for digital filters, but increased headroom is more related to bit depth.
I am sure your opinion is well backed up, and I respect it. On my experience there is a difference and a completely different workflow managing 96kHz files that has nothing to do with “perception“ or even “placebo effect“.
But as I also said, my point it´s not to discuss this, but focus on your chain of equipment. It will always sound WAY better if you get a great signal from your cartridge and preamp; and then, if it´s possible to made the conversion with a strong clock interface (like the one I mentioned) you should be able to achieve great sounding files.
In case of recording a copy of one of my records, I think the best is 24 / 48kHz. But in case of recording anything else to work with afterwards, probably you should aim higher if your interface allow you.
Always work on getting a strong signal and it will sound great!
Regards!
That is an EXCELLENT video. I learned a lot from it, including some of the information in this video.
Going back to what this video was about... 24 / 48 is more than enough for archiving vinyl. Considering that bit depth is directly related to dynamic range, a vinyl record might be equivalent to an 8 bit recording, or 12 on a good day. I would say that 16 bit is plenty, but 24 gets you out of the dog house in the community, and you sleep better. But as hard drives get cheaper, it certainly doesn't hurt to forward think, and allow lots or room for vinyl to do its thing. Cheers!
Is part 2 ready yet? Thanks for all your work Craig!
Where's pt. 2? it's been 4 months, are you okay man?
I personally contacted him, I too hope he's OK and asked if he planned on any new material. This guy is one of the few who really explains things in an understandable way and conveys why I like vinyl recordings.
So......, any updates?
No - nothing I'm afraid. Doesn't look good, hope he's OK.
He seems okay, and is still posting on his other UA-cam channel "CraigTube".
192 is crazy indeed. Damn hdtracks
I connect mine into the mic port on my old HP laptop works great.
With the higher frequencies above normal hearing range, some people have said that they can produce more natural harmonic renounces for a fuller and more realistic sound. I think there might be something to it, as I know you can get various wave interference patterns, but I couldn't say for sure one way or the other what affect it actual has on the overall quality and experience of listening to music, if any. I do think it's perhaps something to keep in mind though.
please, please, please... release the second part
Thanks for your work
2 months ago you posted this, now when will you post part 2? i'm still waiting. also how to you eliminate noise floor being too loud?
This is the best video on this topic.. I’ve been digitizing my vinyl records 24 bit 48. However I never knew the difference with this bit rate.. 16 bit 41 is good? I am trying that out to see if I hear a difference.. thank you again for this video..
Great Video, what about the effect 2nd and 3rd harmonics have on the audible frequency range? I've read / watched videos showing audio content (and not just noise) on a vinyl LP going up way passed 25,000Hz. So with a good cartridge 24bit @ 48KHz seems like a better option. Many instruments produce sound over 20,000Hz, and good mics go above this level as well, so an analogue format should also reproduce this (if the cartridge and kit is able too).
Hi Craig -- Love your videos! I was wondering if you have any tips on avoiding inner groove distortion? Thanks!
interesting, thanks.
re hearing different sample rates. when a filter is applied to a stereo file it distorts (tiny micro) to both bass and hi end. a by product is spikes or ripples across the harmonics within human hearing ranges. its subtle but when someone points it out you may then hear it. guess it depends on individual, im always lost in enjoyment of the piece not discrete artifacts. but i think thats whats going on.
This is great stuff! Thank you! What about Part 2? :D
I go with 32/96. I remove clicks, fade out and very light noise reduction.
A friend of mine believes 64kHz sampling is equivalent of vinyl quality, he hates 44.1. He recorded the council inspector using mic,tape deck to preamp the mic to computer sound card, the PC was set to 44.1 and when it was played back it sounded metallic . He didn't have enough time to set it to 48 or whatever
You say 2 part, will there be a part 2?. Miss seeing new content here.
good video, not just because it's informative but i also like your way to approach some of the audiophile voodoo talk that can be found all over the internet.
i bought a turntable with built in usb, wich also means it has a pre-amp built in wich is very handy but also very versatile because the chinch connectors can be set to output at line OR phono level (never need to worry if the amp has phono input or not)
it also allows me to record a vinyl on my computer AND my tape deck at the same time, or connect a seperate audio interface should i ever want to go higher than 16/48 wich is the highest the built in usb will give me. (and totally enough if you ask me)
maybe i could borrow such an interface to record two files in different quality from the exact same spin and then do the phase inversion thing for comparsion.
anyways ... the recording is done with plain old audacity and then stored as FLAC and eventually transcoded to mp3 as well if i don't have this particular album in digital form already.
some recent releases have the vinyl AND a CD in one box wich in my opinion is exactly the way to do it nowadays: the vinyl version because it's vinyl and a CD thrown in as a bonus for ease of use (even if it only comes in a simple cardboard sleeve and not in a full on fancy jewel case or digibook, but if i would have wanted that i could have bought the CD release to begin with)
A lot of people (including me), can hear soft sounds below the noise floor! If they measured vinyl below the noise floor, then the dynamic range specs would be higher than CD'S! But some people can't hear below the noise floor, thus the specs they give it!
I use my Digital recorders to capture the analogue music. that is also an option.
Enjoyed this video
My set up is the same. Preamp to AV / USB I will probably change my settings now and listen for the any change in quality
Do you use any dolby or declicker or other settings. Hope your well
As long as the frequency response can comfortably encompass the frequency range of your source you should be fine, otherwise you may very well get audible artefacts even though you are dealing with frequencies that the human ear can't hear.
hey, when are you gonna record part 2 of this tutorial? thx.
where is part 2?
intresting what frequency have real musical instrument and if human ear hear 20hz - 20khz why they create hi range frequency
Craig one minor correction, 2 to the 32nd power is around 4.3 billion, not 43 billion. Great video still, clear explanations.
One correction: the Compact Disc has up to 96db of dynamic range, not 90db.
Does anyone have experience with having this done professionally / as a service? How are the results, cost etc? It is not because I wouldn't wanna attempt it myself; I am just thinking that if I do this, I want a really good result and of course right there, it would be limited by what my personal TT can deliver...
When does part two come out?
Where is part 2 ):
I don't normally put music on my computer and I wouldn't give you a $ a dozen for all the MP3's in the world, man. I mostly listen to Classic Rock on Sirrus in the car. But, "vinyl is final". GB
Part Deux? I see MANY comments that are awaiting the next installment....
Most good software internally capture audio at 32bits precision due better rendering/downsampling or using declicker/denoiser etc... even if vinyl has around 70db dynamic range that is near close to 16bit. While saving audio need to be dithered properly if saving 32->24 or 32->16 bits and that is more crucial than other things. And setting samplerate more than 48khz is absolutely unnecessary and big waste of space...
I strongly recommend to capture your vinyls using good software like iZotope RX or Adobe Audition, because other can harm your audio even more...
part 2?
Actually you can't really record in 32bit depth as converters use 24bit (or 16bit for older converters). 32 bit are only used inside digital audio workstations for processing audio but you are never gonna be able to playback audio using all those bits as they are reduced to 24bit in DA converters and if they exceed that you're going to hear distortion. In fact if you select 32bit for recording an analog source like a turntable or a guitar you're just getting 24bits of recording plus 8 empty bits which are only good if you like wasting your hard drive storage space. Also for the (roughly) 70dB of dynamic range that vinyl is capable of delivering 24bit isn't probably worth it, as 16bit depth is capable to deliver 90 to 100 db of dynamic range.
did you ever get new table
20:44
Pun!
I can only go up to 16bit 48K on my USB in direct from my table. Good enough for casual listing, I'm sure, but is it good enough for achieving and science? Hmm... probably not for me.
I wouldn't have a problem with it. 16 bits will give you much more dynamic range than any vinyl record can produce, even with the best turntables out there. You just can't drag a stylus along a groove and get as much dynamic range as 44.1 can give. I don't care what cartridge you have. Just try to record as loud as you can without going over. If you record low at 16 bit and then normalize, you do reduce your dynamic range quite a bit.
The way I understand it, the bit depth is basically the resolution or accuracy of the audio within its full potential range (per sample), so I would assume that a higher bit depth would theoretically always result in a more accurate digital representation of an analog source (and likely digital conversions as well). Of course there would come a point where having a higher resolution wouldn't result in a distinguishable quality gain, contending with noise, margins of error, human hearing potential, etc. True analog though would theoretically be infinite in its resolution, well, perhaps down to quantum levels.
Either way, it's good to know that you think 16bit is sufficient as long as the full loudness range is properly utilized and I'm not cutting off dynamic range potential. (You seem to have a pretty good ear for this sort of thing and real world experience.)
My Windows 7 computer can "listen" to the USB turntable source (I have the AT-LP120-USB as well) at up to 16bit 48K for recording into Audacity (I'm not sure if this is an OS, driver, or USB limitation - I might try this on my Debian/KDE laptop too), then for file encoding archiving, I'm using FLAC level 8 compression, 24bit 48K. Kind of wish the USB would go to 24bit as well for a better match (weakest link in a chain, and all that), but the USB out actually sounds better (brighter and with higher clarity) to me than the preamp out on this table, so that's why I'm sticking with it instead of just using a regular line-in on my computer. The noise on the USB with the table connected and powered is surprisingly low as well, which is nice, and perhaps more significant overall than the 16bit limit.
16bit, 48K is listed as being "DVD quality," at least.
Anyway... Looking forward to a followup video to this if you end up making one. I'm curious to see how you go about all this more on the software, recording, and encoding side of things.
Cheers.
Ok Craig.....I think I got it...but how many bites in a McDonalds? :-)
Depends on how hungry you are :)
sorry if this sounds rude, is Craig blind ? ( partially sighted )
Yes, he is partially sighted and registered blind, he mentioned that in a previous Home Brew Wednesday video.
Is this dude ok? Maybe he is not well :(
I suppose, after Craig had removed PreAmp out of his turntable, he can no longer digitize anithing. That's why there's no Pt2..))
Many people need to realize that the insanely high sample rates are complete nonsense. I looked at a 24/192 song that I down sampled to 24/48 and there was absolutely no difference. Looking at the frequency spectrogram of the music revealed that more than three quarters of it were not being used and just taking up memory, the more people know what proper settings to use the more space they'll save on their hard drives.
Also I don't mean to sound rude but why do your eyes constantly move through your videos?
Sounds like it was a bad mastering or "up conversion" of some sort then. I can make gigabytes worth of nothing too, and have in fact to use as testing loads, but that's a different matter.
You can't really up convert music, once it gets as good as it can be it cannot be made better. Afterwards it's just taking up more memory.
Yeah, that's basically what I was saying. You can do some cleanup type stuff, remastering, etc., but that's a little different.
@@ZANTHERA It is funny you should say that something that is actually being done cannot be done ;) Unlike what you say, it is not complete nonsense. Of course nobody thinks they just magically created more information that was not there to begin with. The reason for up-sampling is that the filters that get applied e.g. by your CD player in processing CD audio are responsible for the harsh etc sound people complain about when comparing CD to vinyl. Now, up-sampling does not make the music better (as you say, of course), but when processing the signal, the impact of filters by your CD player gets moved out of the audible range for human ears. That is why they sell you CD players that do up-sampling to 192.
In the 80s, early 90s, powerful chips were not available and/or too expensive for the mass market, but today it is not an issue anymore. So replacing a CD player from 20-25 years ago will make a real impact (or any older tech player). E.g. I still have a Musical Fidelity CD1 from early/mid 90s, which had cost 2000 bucks back in the day, and if you compare that to a current player of the same price range, they beat the pants off of it. Shame, last time I checked, they were throwing MF CD 1s for like a 100 bucks at you on Ebay... I recently paid almost as much for a replacement laser unit :) (I don't use it, I just keep it because it looks awesome). At the same time in the 90s, I had also bought a pre-owned Linn Basik turntable, which I still have :)
Of course I agree with you that having such up-sampled files laying about on your hard drive makes no sense...
If you record from an analogue source, 24/96 sound hugely better than 16/44. It's not even a subtle difference.
96 gives you more than twice the frequency response needed for human hearing at birth. I guess that's what confuses me. Why do we need all that extra frequency range when 48 still gives well above human hearing. Has anyone ever been able to explain this?
Hi Craig, I, like you have been a musician since the 1970's and a home recordist and songwriter since the 1980's. My most recent recording set up uses a MOTU 896 digital interface which is capable of recording 8 channels at up to 24bit 96KHZ sample rate, Logic Audio studio, a MacBook pro and an external hard drive. I usually record my bands at 24bit 44.1KHZ due to not wanting to tax my lap top with plug ins.
However, I did an experiment with all the settings on the MOTU 896 at various sampling rates. I monitored on a £50 (approx $50) pair of Senheiser headphones that I know very well. I used a Fender Malibu acoustic guitar, and four microphones plugged into the MOTU 896 ; Rode NT-1 condensor mic, Superlux "pencil style" drum overhead condensor mic from my drum mic kit, a SM-57 dynamic mic and a Tandy (Realistic) PZM mic.
On an acoustic source, my guitar and voice, the change is sound quality was quite marked. At the higher sample rates you could hear more of the room sound "the air", the acoustic space in the recorded sound and at 24/96 it was as if the headphones disappeared and it was more like hearing myself play and sing live in the room and not from a recording. There was no additional high end apparent, because as you quite rightly say we cannot hear this directly. However, I have read what happens is that we perceive the higher frequency range because the higher harmonic octaves that we can't hear directly, of the sound recorded interact with the frequency range we can hear. The effect of this is you can hear the qualities of the acoustic space and the "air" due to the higher harmonics captured at 96KHZ.
This is not a subtle change, it is quite striking, even on a fairly budget pair of headphones.
I don't think 96khz would make as much of a difference when recording a loud, close miked rock band, but on acoustic instruments it does. That is also why I record my band at 44.1KHZ. Plus, I would need a high quality acoustic recording room to make the most of 96KHZ. I may do it as an experiment and then mix down to CD quality to see if there is a difference in the mixed track.
These higher harmonic octaves that 96KHZ and 192KHZ capture are present in analogue recordings from the 1960's onwards and also on vinyl. So for capturing vinyl sound in all it's glory I would recommend using the higher sampling rate. This is also why, I believe, vinyl sounds better than CD. Because the analogue reproduction captures the sound of the "air" and acoustic space. Whereas, the limited range of CD leaves a somewhat "flat" one dimensional sound.
I have read interviews with recording engineers where they claim that even 192KHZ makes a difference, albeit a diminishing return from 44.1 KHZ, And having tried 96KHZ recording myself, I believe them.
Hey John, thanks for your detailed explanation. I like detail.
I fully understand and agree with what your talking about with the ultra-sonic frequencies interacting with the sonic frequencies we hear. If two frequencies above 20khz, but very close together, both in frequency and volume interact, they can cause a "beat note" or sub harmonic. The question is... What is the precise sampling rate needed to allow this to happen just above our range of hearing? At some point, even the sub harmonics created at some sampling frequency would be above our range. Notes/tones have to be fairly close together to cause peaks or troughs that might generate sub sonic tones we can hear. So there you have the number one question that needs to be answered with math.
But the one that I always go to... Placebo. It is scientifically proven that our perception of things can definitely differ, depending on what we might expect. This has been so strongly proven, that it's used all the time in blind tests for medications. Unless a test, even with audio, is done in a blind manner, the results are totally invalid for study, albeit proof of the placebo effect. I would be overwhelmingly shocked and impressed if I witnessed, in person, a person who could tell the difference between 48 and 96 without knowing which was which. And the test has to be done many times so an average can be taken. Someone guesses it once, that's not enough.
I have read a lot about it too, and I haven't really looked hard enough to find any peer-reviewed studies that conclude that we can perceive, or be affected by anything over our highest hearing threshold. I have read that subjects who were put to such a test between high bitrate mp3 and CD, came out in favor of the CD in regard to emotional response, not sound quality. That might be something to look into more. But for me, I never trust an A/B comparison unless A and B are unknown to the subject being asked, and the more tests, the more "bit depth" we have... :)
No audio storage method is transparent, but our ears are limited. Maybe one day we will be able to store audio in it's pure analog form, with no loss. Only then, will this subject be solved. I'm thinking Plank Level Storage, or PLS. I have a Patton..... /Kidding :)
All I can say is that I tried the recordings myself at different sample rates up to 96KHZ. I was not expecting to hear any difference, I was quite sceptical, to be honest, but the change in the audio with respect to the sense of hearing the acoustic space was quite striking and surprising. It was far too obvious to be put down to "Placebo". Why do you think audio interfaces go to 96khz and now 192 or beyond?? Because the improvement in audio quality, with an acoustic source, is quite obvious. It is not a subtle enough change to be subjective. It's quite obvious! After hearing this, you then realise that CD is a quite restricted audio format. Better than MP3 yes, but not true HiFi. CD was set at 16 bit and 44.1KHZ because they wanted to fit Beethoven's 9th, which is 77 minutes long. 16/44 was the best resolution they could manage with early 1980's digital technology. It's not the best quality that can be achieved from Digital audio.
Try it yourself with a good quality Mic and an acoustic source, your voice or a guitar or percussion. Distance mic by some inches and you will hear as you go up the sampling rate that the room sound, dept and sense of the "air" is more apparent.
And as for which sampling frequency to stop at, I can't answer that question by my own experience but I guess it would be a case of diminishing returns. There was an interview by a well known recording engineer who talked about 192 and even 384KHZ on the Sound On Sound magazine website.
Cerastes, that article is wrong! I've done the tests, which I recommend that you do too.