Article: Digital Vinyl: Temporal Domain

[Le Concombre Masqué]

Why are you recording at 176 ? would you buy your 176 source stuff as hi-res or be perfectly happy streaming it "CD Quality" ? MQA ?

Okay in a perfect world with perfect reconstruction higher sampling rates get you higher bandwidth. They do get you better time resolution, but it is already so good that should not matter by a factor of like several hundred times at least. So you get bandwidth. You don't get a more accurate reconstruction of the signal below 20 khz. In the above example the differences you see are from no bandwidth above 20 khz at the 44.1 khz sample rate.

 

Now yes there is some signal above 20 khz. Can you hear it, does it effect the sound your ears hear if it is gone versus if it is present? I don't think so. People typically can't hear the difference unsighted. Our basilar membrane in the ear has a physical construction that appears to end response at 15 or 16 khz. It acts as a loosely tuned filter at that point with some rapidly diminishing response above that frequency which is why with enough sound level you can hear 20 khz when young. A small percentage of people with the most excellent hearing in their 20's can hear 23 khz. Now I don't know the answer, I bet at least when young the ear drum probably responds up a bit more, but the basilar membrane which is what activates the auditory nerve responds weakly if at all up around 20 khz.

 

Below is a comparison of a cochlea mathematical model vs actual results. It gets pretty close. Notice the peak somewhere around 16 khz and then around 20 khz it simply drops to no response.

 

[ATTACH=CONFIG]34052[/ATTACH]

 

Using that same model this page has animation of the BM response in the cochlea for several types of signals. One of those is a Dirac impulse. Click on that circle for the 'pulse sequence' and see what is predicted.

 

Traveling Waves

 

 

Bone conduction can stimulate some sense at higher frequencies. Sounds through the air are not energizing bone anywhere close to enough to matter unless sound level reached more than 150 db. You would be deafened by everything else if that happened and still not hear it. The BM stereocilia would be destroyed or highly damaged.

 

So that leaves imperfect implementation of our digital system. It is my opinion that is close enough to theory it is of no concern. Others feel differently. I would note that is feel differently.

 

I know it is a cliche. If you sit someone down and show them SD and HDTV there is no question which is better. Super plain and simple. Ditto for 4k TV. If there is some benefit to higher sample rates it is a very small benefit nowhere near the differences in TV resolutions. If it were half that large we would not have any argument whatsoever at this late date.

 

If you still want to go higher rate just in case, I certainly see no reason for more than 96 khz. I mean you aren't hearing 40 khz or artefacts from that even if instruments have some harmonics higher than that. Damned few microphones go even to 40 khz. Anything which has gone across tape is diminished up there. I see no basis for 192, 384 and more.

 

Now you mention catching quick impulses. Even when an impulse is quick your ear simply has no pathway to follow it fully. Higher harmonics are going to be filtered out just the same. The upper filter is pretty steep on your ear too. That is why some DACs don't sound tremendously awful if they don't have a reconstruction filter. The roll off of ear response IS the filter.

 

So I couldn't say with 100% certainty 44.1 khz is enough or even 48 khz. There are lots of reasons, not opinions, but well founded reasons to think it is. To think if something is missed it is either of no consequence or something so small as to be of highly limited difference in terms of how satisfying the playback is for you.

 

As for 16 or 24 bits you could make a case for either way. 16 bit isn't losing much though there are cases where it could be audible. Much more so than higher sample rates. Basically with modern gear go 24 and don't worry about it. Then again with modern popular music compression is so out of hand dithered 8 bit is good enough. What little hiss you might get would give that old cassette softness which might cover up some of the spitty sound of limiting kicking in.

[board]

There is a little mistake I think. Timing accuracy of digitally sampled audio is 1/(44,100*2*pi*65536) Works out to about 56 picoseconds. This is for 44,100 at 16 bit.

.

The 20,000 is the bandwidth available on a CD, although it should probably be 21,000. As far as I know, the limit for how small an increment can encode is 1 nanosecond.

 

It's also worth remembering that the the quantization errors performed by the A/D converter are completely eliminated and turned into random noise by a process called dithering, which is contained in every A/D converter on the market. The early A/D converters from the 80s didn't contain dithering though, as this wasn't introduced until the mid 80s. So, bit depth is only signal-to-noise ratio. In other words, 24 bit doesn't offer higher resolution, just higher dynamic range (144 dB for 24 bit, 96 dB for 16 bit).

[board]

So then you would agree that demonstration of a consistent ability to distinguish between PCM44 and let's say DSD256 recordings would be proof of the ability to perceive ultrasonic content?

 

 

Room treatments for headphone users

 

Yes and no.

DSD is a bit tricky, as it's a different technology that has certain issues. If you haven't heard about it, Stanley Lipshitz and John Vanderkooy did a paper on 1-bit technology in DSD and basically concluded that it was broken:

http://www.essentialaudio.com/1-Bit-Is-Bad.pdf

 

I've heard about studies that concluded that people couldn't tell DSD and PCM apart anyways. I believe our ears are a lot more forgiving than most audiophiles like to think.

[esldude]

Why are you recording at 176 ? would you buy your 176 source stuff as hi-res or be perfectly happy streaming it "CD Quality" ? MQA ?

 

I didn't record the cymbals. John E. Johnson Jr who writes for one of the Home Theater mags is a drummer and did a series of measurements and reviews of dozens of cymbals. His recordings.

 

I do a little recording. I have a few times recorded at 96 khz and also at 192 khz. Why? The ability to do so was there. If there were a difference, a recording done by me directly off the mikes with zero in between processing should reveal it. While I didn't expect a difference I did try and see if there was one. I also sent some tracks to the younger folks in the band (one was 12 years old) who might have better hearing than I. I told them nothing other than I thought these are a bit different tell me which is better. The fly in the ointment is they don't have great high end gear. They did use some studio monitors and a DAC that could play the high res file. They all said they heard no difference.

 

So generally I record at 48 khz. Why 48 khz? Well 48 khz does have a wider transition band between the cut-off at 24 khz and response at 20 khz. That should allow the filters to work slightly better and have a little lower aliasing.

 

As for buying recordings I am in favor of as few transitions as possible. If I knew a recording was done originally at 176 khz then that is the one I would prefer. Although I have violated that idea when the price difference was ridiculous. Pricing by the bit for instance. I'll take the 48/24 version when the 176 or 192 version costs 4 times as much.

[board]

So far this article that we're debating has spawned 13 PAGES OF COMMENTS!

If the people writing the article had actually consulted people who understood how digital audio works, they wouldn't have written this article to begin with.

If someone who understood how digital audio works had posted the first comment, debunking everything in the article, the debate would have stopped right there.

The authors are just repeating what Kunchur wrote several years ago. Back then it was disproven, and he was discredited. The same should have happened here. For anybody who wants to understand how the technology works, I provide links at the bottom of this post :-).

 

But surprise, surprise! The authors didn't consult people who understood the technology but instead went to people who love vinyl and hate CDs but don't understand the technology. That has, unfortunately, become standard procedures for vinyl lovers, because the truth seems less important than propaganda that hails vinyl and sheds bad light on digital audio – CDs, in particular. Unfortunately, some of these people, like Dudley and Fremer, are in positions of power: Most audiophiles only have experience with listening to music, but they look up to Fremer and Dudley et al as figures of authority, so if those people spread misunderstandings and lies, those beliefs are rarely questioned by audiophiles.

Unfortunately, this has created a lot of fighting between the audiophiles and the scientists :-(.

 

There's nothing wrong with having a preference for something – we all do. Many albums I would never, ever buy on CD. But I would have a lot more respect for the vinyl lovers if they just said "yes, I know vinyl is less transparent, but I simply like the coloration it gives to music", instead of spreading bullshit and lies. If they said that, there would be no arguments. Saying the Nyquist theorem is wrong is simply downright stupid!

Luckily, other people who understand the technology have chimed in here, trying to sort things out, but an article like this does a lot of damage to the people who don't understand the technology and can't stomach reading through 12 pages of comments.

 

As mentioned before, the ONLY difference there can POSSIBLY BE between CD and hi-res (unless something has gone wrong) is that the hi-res file has information above 22 kHz and that it may have more relaxed filters. Nothing else. Zip. Zero. Nada.

192 kHz or 384 kHz sampling rates do NOT produce an output where the raggedy steps more closely approximates the original analogue soundwave, because the output of a DAC are not stair-steps to begin with (see more about this further down). All they do is include higher frequencies. Anybody who claims otherwise don't understand the Nyquist sampling theorem.

And yes, if the original analogue signal contains energy above 22 kHz, then a higher sample rate could strictly speaking be said to be "closer to the source". If you want to retain that information for intellectual stimulation, then fair enough. As long as you state that. But as it's inaudible anyway IMO it's pointless to preserve those frequencies. But that's my opinion :-).

 

If you insist on preserving supersonic frequencies, hi-res digital is the way to go – NOT vinyl. Vinyl records are mostly cut using a low-pass filter that rolls off high frequencies, usually above 15 or 18 kHz. Any supersonic frequencies are so low in level to be inaudible – which they would be in any case, as they are above 20 kHz (masking also comes into play here). Yet, many spectrograms of hi-res recordings of vinyl records show energy above 22 kHz, but most of this is distortion or noise. Some vinyl records are cut off at even lower frequencies. An extreme example would be the original copy of Grand Funk's "We're an American Band" where anything above 5 kHz seems to be cut off!

When the Beatles remasters vinyl records were released, someone presented a spectogram of a hi-res recording of one, which showed content above 22 kHz, and people were raving "Yes! Yes! Yes! I knew it! They are cut from analogue tape or hi-res digital!". The mastering engineer then stepped in and said that they were cut from 44.1 kHz files, so the content above 22 kHz could simply not be anything else than noise or distortion from the vinyl material or the turntable rig.

 

Art Dudley's claims are fantasies and just fancy and romantic sounding jargon that proves nothing. He's free to believe anything he wants, but when he makes his fantasies public, and they're factually wrong, he will be corrected.

Granted, digital audio is a very complicated subject, and it took me a long time to understand. There are still issues I struggle with, and I don't fashion myself to be an expert on the matter, but apparently I know more about it than most people working for Stereophile!

But then again, how could I expect that someone who's willing to pay $3000 for a power cord could possibly understand anything about technology?

 

So, Art Dudley, and the authors of this article, please provide the results of the level-matched blind-tests you have passed between vinyl or analogue tape vs. said tapes/records properly transferred to 16/44.1 digital. Then please also provide the results of the blind-tests you have passed between hi-res and the same material properly downsampled to 16/44.1.

Until you do that, the claims are just empty claims.

 

A small side-note on something that very few people bring up: I was vinyl only for 15 years, but after upgrading my stereo I have compared almost 800 albums on vinyl and CD.

If there's one category of CDs that are the easiest to generalize about it's CD from the 80s. I'm not a fan of 80s music, but I did compare a reasonable amount of CDs from that period. So far I have only found two CDs from the 80s that were better than their vinyl counterpart (as well as one album where the two medias were practically identical). The rest sounded thin, shrill, cold, clinical and hard.

I still have several hundred albums that either simply sound better on vinyl, or that I simply prefer. Unfortunately, there are many CDs, both old and new, that sound bad, and many albums from the analogue era sound lovely, as they are well-produced, unlike a lot of music nowadays.

Art Dudley, Michael Fremer and so many other CD haters probably heard CDs back in the 80s and concluded they were crap (I know Fremer did). I would have done the same! And whenever you hear of comparisons being done, they are usually done with CDs from the 80s.

I attribute these issues to the poor A/D converters and the poor mastering practices (for CDs) in the 80s. Even pro-digital people (Stanley Lipshitz, Bob Ludwig, Bob Clearmountain, etc.) have also mentioned these issues in public.

 

Nowadays, converters have become cheap, small and well-functioning. I ripped a CD, then took the CD and recorded the output from my CD player with a very long (at least 5 meters) cheap standard analogue cable into the line-in on my cheap Soundblaster soundcard, adjusted the volume levels and did a blind test: 4 out of 12 correct. Worse than flipping a coin!

You can try this yourself, but if you're too lazy you don't even have to record anything on your own: You can go to Ethan Winer's website and download three different songs for his loop-back test. They all come in four versions: The original, then recorded back into his $250 Focusrite soundcard (one generation). Then after 5, 10 and 20 generations. See if you can hear a difference.

Also, if you take a vinyl record and record it to CD, you will most likely not be able to hear a difference. If you transfer a CD onto a vinyl disc you will most likely be able to hear a difference (although some, and only some, vinyl records will of course be transparent within the limits of audibility, minus the surface noise). If you take an analogue master tape and copy it to another analogue tape, most people can hear the difference.

That just goes to show that digital audio is transparent and analogue is far from. Claiming vinyl or analogue tape is closer to the source than CDs is ludicrous. But when these "figures of authority" say so, it becomes gospel. Let me use Michael Fremer's own favourite phrase against him: "You're talking out of your ass!"

 

I hope everybody will take 1 minute and 30 seconds to watch the beginning of this following video by Sam Harris. The rest of the video is mostly about religion, and you don't need to watch more than the first 1:30 minutes. I link to this, as I feel this is how it is debating with vinyl-lovers who don't trust facts nor science.

 

 

For anybody who wishes to understand how digital audio works, here are a couple of links:

 

First, watch the video by Monty above. It's absolutely crucial!

 

Then a lot of answers to myths about vinyl being technically superior to CDs:

Myths (Vinyl) - Hydrogenaudio Knowledgebase

 

Then an easy to understand article about the real difference between 16 and 24 bit (spoiler: it's not resolution):

24bit vs 16bit, the myth exploded!

 

Then an article from the same person about the real difference between CDs and hi-res (spoiler: it's not resolution either):

Hi-Rez - Another Myth Exploded!

 

 

Then lastly, a quote by Christopher Hitchens:

"What can be asserted without proof can be dismissed without proof."

 

This article here has shown no proof and can be dismissed.

[esldude]

The 20,000 is the bandwidth available on a CD, although it should probably be 21,000. As far as I know, the limit for how small an increment can encode is 1 nanosecond.

 

It's also worth remembering that the the quantization errors performed by the A/D converter are completely eliminated and turned into random noise by a process called dithering, which is contained in every A/D converter on the market. The early A/D converters from the 80s didn't contain dithering though, as this wasn't introduced until the mid 80s. So, bit depth is only signal-to-noise ratio. In other words, 24 bit doesn't offer higher resolution, just higher dynamic range (144 dB for 24 bit, 96 dB for 16 bit).

 

The pertinent number however is rate of sampling not response. Yes dither in theory might make timing resolution as low as you want. In practice in some sense it increases timing resolution though not to infinity I think. In any case it is a non-problem for fidelity to human ears.

[esldude]

May we listen to Stephen Colbert. He said last night that millenials had to stop pushing vinyl. Don't they know we would have given anything to get free music whenever we wished streaming over our rotary phones?

 

I may have garbled the quote, but you get the idea.

[Fokus]

The early A/D converters from the 80s didn't contain dithering though, as this wasn't introduced until the mid 80s.

 

Some didn't. But the various Sony PCM16** and F1 convertors did have dither from day one.

 

https://www.gearslutz.com/board/attachments/mastering-forum/25537d1161190991-old-sony-pcm-users-analogue-pcm-1983-.gif

[Jud]

I sample rate converted to 44.1 and then back to 176 so the scale would match on both. Upper is the original 176 khz and lower is the 44 khz.

 

[ATTACH=CONFIG]34044[/ATTACH]

 

 

[ATTACH=CONFIG]34045[/ATTACH]

 

 

You can see in the latter image that the very beginning of the transient wasn't slow to start on 44 khz. You do see some peaks are rounded and lower in level due to the energy in ultrasonic frequencies that were not captured.

 

It is hard to say where the real fundamental modes of the cymbal are. The decaying tail has strong output at 6700 hz and 8 khz with a big drop off after that. The harmonics do extend to maybe 50 khz prior to 100 milliseconds. The microphone used is said flat to 40 khz and I am sure has some output beyond that. After about 100 milliseconds the strong harmonics beyond 10 khz suddenly die out leaving pretty much only below 10 khz sound.

 

Here is the spectrogram. It goes to the gray background at -80 db and uses 1024 bin FFT as I have it set here.

 

[ATTACH=CONFIG]34046[/ATTACH]

 

Thanks. What's interesting is that (at least as my eyes are seeing it - would be nice to get an overlay view if that's possible) that as you say the rise to peak doesn't seem slow; the differences I think I'm seeing are in the decay. (And these don't look to be large differences.)

 

Edit: What conversion filter did you use?

[Jud]

There's nothing wrong with having a preference for something – we all do. Many albums I would never, ever buy on CD. But I would have a lot more respect for the vinyl lovers if they just said "yes, I know vinyl is less transparent, but I simply like the coloration it gives to music", instead of spreading bullshit and lies.

 

 

* * *

 

If there's one category of CDs that are the easiest to generalize about it's CD from the 80s. I'm not a fan of 80s music, but I did compare a reasonable amount of CDs from that period. So far I have only found two CDs from the 80s that were better than their vinyl counterpart (as well as one album where the two medias were practically identical). The rest sounded thin, shrill, cold, clinical and hard.

I still have several hundred albums that either simply sound better on vinyl, or that I simply prefer. Unfortunately, there are many CDs, both old and new, that sound bad, and many albums from the analogue era sound lovely, as they are well-produced, unlike a lot of music nowadays.

 

So many vinyl lovers' preferences may come to some extent from the fact that at least part of the time they're listening to better mastering. I completely agree with the proposition that mastering trumps resolution or even vinyl vs. digital.

 

 

And then folks get sucked in by plausible-sounding hypotheses that are incorrect, but they don't examine too closely since these hypotheses are in line with their preconceptions. I'm still in agreement, in fact perhaps more so than you: that Lipshitz and Vanderkooy article is steadfastly cited by lots of people whose DACs are doing sigma-delta modulation internally (yourself as well, perhaps?), and they don't seem to mind.

[Le Concombre Masqué]

ie Van Gelder's remasterings sound great at 24/44.1

I didn't record the cymbals. John E. Johnson Jr who writes for one of the Home Theater mags is a drummer and did a series of measurements and reviews of dozens of cymbals. His recordings.

 

I do a little recording. I have a few times recorded at 96 khz and also at 192 khz. Why? The ability to do so was there. If there were a difference, a recording done by me directly off the mikes with zero in between processing should reveal it. While I didn't expect a difference I did try and see if there was one. I also sent some tracks to the younger folks in the band (one was 12 years old) who might have better hearing than I. I told them nothing other than I thought these are a bit different tell me which is better. The fly in the ointment is they don't have great high end gear. They did use some studio monitors and a DAC that could play the high res file. They all said they heard no difference.

 

So generally I record at 48 khz. Why 48 khz? Well 48 khz does have a wider transition band between the cut-off at 24 khz and response at 20 khz. That should allow the filters to work slightly better and have a little lower aliasing.

 

As for buying recordings I am in favor of as few transitions as possible. If I knew a recording was done originally at 176 khz then that is the one I would prefer. Although I have violated that idea when the price difference was ridiculous. Pricing by the bit for instance. I'll take the 48/24 version when the 176 or 192 version costs 4 times as much.

[jabbr]

Hope I won't ruin your patience but I'm getting more and more confused :

 

to my understanding every demo and thus SN theorem relies on both the band limitation and the periodical nature of the signal, the 2 (X2, twice bandwidth) factor being related to the fact that, taking into account the given sinusoidal shape by nature of the signal, what was up will be mirrored down, that there's only one solution for a given position at a determined timing in a limited bandwidth.

 

This is why I tried to break the image by bursts at rates higher than sampling.

 

Led to out of bandwidth.

We now are looking at signals that when recorded at 176 yield degraded versions at 44.

So, at the end of the day (actually very beginning here !) is 44 perfectly enough as per Monty etc or do actual sound emissions require larger bandwidth for better capture and can this capture be accurately be reproduced at 44? To my eyes you demonstrate Yes they require larger bandwidth, No 44 can not be faithful.

I'm confused for it would thus appear that if it's not a matter of number of samples it's a matter of bandwidth ; isn't it silly ?

Let me try to explain. The Fourier Transform itself is defined for continuous signals (without sampling). It states that for *any* continuous time dependent signal, there exists a transform into the frequency/phase dimension such that the signal can be described *only* by its constituent frequency/phase coordinates -- that means that *any* time dependent signal can be described entirely by a set of sinusoidal waves having a defined frequency, amplitude and phase.

 

This is mathematically true for every signal. *Why* its true involves some math: https://en.wikipedia.org/wiki/Fourier_transform but trust that this is true.

 

Now there is the discrete fourier transform (DCT) which is defined for discrete samples. This is equivalent to the general fourier transform when the sampling rate is infinite. So what to do? In practice suppose we take samples in the Ghz range. We can then examine the discrete fourier transform and suppose we see that *for that specific signal, e.g. a jazz band* that the amplitudes of the frequency components >20 kHz are *zero*. That tells us that *that particular signal* is bandwidth limited to 20 kHz. Now suppose we hit the cymbal ? is there a component > 20 kHz ... let's say there is, now for the sake of discussion up to 40 kHz and then zero above 40 kHz. That means that we can capture the transients in that signal by sampling at say 88 kHz. Or if there are components at 70 kHz then 176 kHz will contain those transients.

 

Do you follow?

 

So you can just look at the discrete fourier transform of the recording and see what the minimal sampling rate would be to capture any transients or *anything else* in that signal. You can determine the actual bandwidth of that signal.

 

Shannon-Nyquist https://en.wikipedia.org/wiki/Nyquist%E2%80%93Shannon_sampling_theorem just tells us what the sampling rate *needs to be* in order to capture the full fidelity of any particular signal (given its particular bandwidth).

 

So again *assuming the signal is bandwidth limited to 20 kHz* 44 kHz is adequate to capture any transients that exist *in that signal*, to answer your question. Not silly.

[esldude]

Thanks. What's interesting is that (at least as my eyes are seeing it - would be nice to get an overlay view if that's possible) that as you say the rise to peak doesn't seem slow; the differences I think I'm seeing are in the decay. (And these don't look to be large differences.)

 

Edit: What conversion filter did you use?

 

Don't have a simple quick way to overlay them precisely. I instead inverted one and mixed them to show the difference. The FFT at the bottom is for the first 100 milliseconds. The top view is spectrogram and the lower one of the zoomed in part where you can see the samples. I used Sox for the resampling to 44 and back via Audacity. Below 20 khz their is some noise at -160 db. When I play this I hear nothing. When I play it at 25% normal speed it sounds like if you tap the bottom of a 12 ounce soda can with a pencil.

 

[Jud]

Let me try to explain. The Fourier Transform itself is defined for continuous signals (without sampling). It states that for *any* continuous time dependent signal, there exists a transform into the frequency/phase dimension such that the signal can be described *only* by its constituent frequency/phase coordinates -- that means that *any* time dependent signal can be described entirely by a set of sinusoidal waves having a defined frequency, amplitude and phase.

 

This is mathematically true for every signal. *Why* its true involves some math: https://en.wikipedia.org/wiki/Fourier_transform but trust that this is true.

 

Now there is the discrete fourier transform (DCT) which is defined for discrete samples. This is equivalent to the general fourier transform when the sampling rate is infinite. So what to do? In practice suppose we take samples in the Ghz range. We can then examine the discrete fourier transform and suppose we see that *for that specific signal, e.g. a jazz band* that the amplitudes of the frequency components >20 kHz are *zero*. That tells us that *that particular signal* is bandwidth limited to 20 kHz. Now suppose we hit the cymbal ? is there a component > 20 kHz ... let's say there is, now for the sake of discussion up to 40 kHz and then zero above 40 kHz. That means that we can capture the transients in that signal by sampling at say 88 kHz. Or if there are components at 70 kHz then 176 kHz will contain those transients.

 

Do you follow?

 

So you can just look at the discrete fourier transform of the recording and see what the minimal sampling rate would be to capture any transients or *anything else* in that signal. You can determine the actual bandwidth of that signal.

 

Shannon-Nyquist https://en.wikipedia.org/wiki/Nyquist%E2%80%93Shannon_sampling_theorem just tells us what the sampling rate *needs to be* in order to capture the full fidelity of any particular signal (given its particular bandwidth).

 

So again *assuming the signal is bandwidth limited to 20 kHz* 44 kHz is adequate to capture any transients that exist *in that signal*, to answer your question. Not silly.

 

One thing further to what jabbr said, for our masked vegetable friend :

 

"Periodic" in the sense of sine wave-like signals is also not necessary. The Sampling Theorem also works fine for stuff that doesn't produce nice sine waves, like percussion or inharmonic attacks, such as the pluck of a string or (for brass players) "tonguing."

 

 

Sent from my iPhone using Computer Audiophile

[jabbr]

One thing further to what jabbr said, for our masked vegetable friend :

 

"Periodic" in the sense of sine wave-like signals is also not necessary. The Sampling Theorem also works fine for stuff that doesn't produce nice sine waves, like percussion or inharmonic attacks, such as the pluck of a string or (for brass players) "tonguing."

Yes, absolutely true. The fourier transform is applicable to all time dependent signals including transients. The classic transient is called the "impulse" and the "impulse response" of a system is extensively studied and discussed. https://en.wikipedia.org/wiki/Impulse_response

 

The response of a system to an impulse or transient is one of the best ways to study a system, be that an audio system e.g. when measuring rooms for correction kernels, or even to study what happens to the Twitter space when Trump issues an impulse tweet about Obama. ***

 

*** I'm entirely serious.

[mansr]

One thing further to what jabbr said, for our masked vegetable friend :

 

"Periodic" in the sense of sine wave-like signals is also not necessary. The Sampling Theorem also works fine for stuff that doesn't produce nice sine waves, like percussion or inharmonic attacks, such as the pluck of a string or (for brass players) "tonguing."

 

One way of looking at it is that all signals are periodic, but some have infinite period time.

[mansr]

So many vinyl lovers' preferences may come to some extent from the fact that at least part of the time they're listening to better mastering. I completely agree with the proposition that mastering trumps resolution or even vinyl vs. digital.

 

 

And then folks get sucked in by plausible-sounding hypotheses that are incorrect, but they don't examine too closely since these hypotheses are in line with their preconceptions. I'm still in agreement, in fact perhaps more so than you: that Lipshitz and Vanderkooy article is steadfastly cited by lots of people whose DACs are doing sigma-delta modulation internally (yourself as well, perhaps?), and they don't seem to mind.

 

That paper talks about inherent problems with 1-bit sigma-delta converters, and it is quite correct. It explicitly states that multi-bit sigma-delta is perfectly fine. Almost all modern DACs use multi-bit converters. Guess why.

[The Computer Audiophile]

Hi Guys - I just edited the article with the following statement at the top.

 

 

The following article contains information that has been deemed incorrect by leading digital audio engineers. I attempted to corroborate the findings of this article by asking several digital audio experts. I was unable to find anyone who could back up the statements made, with any scientific data or theory. Consider the following article retracted.

 

I am leaving the text of this article up on CA because it has enabled a good discussion to take place. By leaving it up, people can read what was claimed and read the followup arguments that the prove it incorrect. To remove the article completely only opens up a space for this to happen again, and again, and again.

 

I take full responsibility for the publishing of this article. I should have had a technical editor check it before publication. I apologize to the CA Community for the error in judgement.

 

- CC.

[Jud]

That paper talks about inherent problems with 1-bit sigma-delta converters, and it is quite correct. It explicitly states that multi-bit sigma-delta is perfectly fine. Almost all modern DACs use multi-bit converters. Guess why.

 

Yes, as Miska is fond of saying, it shows you get bad results when you use crappy SDM.

[Jud]

Hi Guys - I just edited the article with the following statement at the top.

 

 

 

The following article contains information that has been deemed incorrect by leading digital audio engineers. I attempted to corroborate the findings of this article by asking several digital audio experts. I was unable to find anyone who could back up the statements made, with any scientific data or theory. Consider the following article retracted.

 

I am leaving the text of this article up on CA because it has enabled a good discussion to take place. By leaving it up, people can read what was claimed and read the followup arguments that the prove it incorrect. To remove the article completely only opens up a space for this to happen again, and again, and again.

 

I take full responsibility for the publishing of this article. I should have had a technical editor check it before publication. I apologize to the CA Community for the error in judgement.

 

- CC.

 

 

I think that's great, frankly. On how many other sites would the editorial content (1) take account of the comments, (2) ever bother to say "whoops, we goofed," and (3) bother to leave (2) in a position of some prominence?

[corente]

Hi Guys - I just edited the article with the following statement at the top.

 

 

The following article contains information that has been deemed incorrect by leading digital audio engineers. I attempted to corroborate the findings of this article by asking several digital audio experts. I was unable to find anyone who could back up the statements made, with any scientific data or theory. Consider the following article retracted.

 

I am leaving the text of this article up on CA because it has enabled a good discussion to take place. By leaving it up, people can read what was claimed and read the followup arguments that the prove it incorrect. To remove the article completely only opens up a space for this to happen again, and again, and again.

 

I take full responsibility for the publishing of this article. I should have had a technical editor check it before publication. I apologize to the CA Community for the error in judgement.

 

- CC.

 

Thanks for leaving the discussion on CA. Please, be sure not to delete it because I have learnt a lot and I will be able to learn a lot more, reading again the thread and giving myself time to go to the linked articles. I think it is very valuable.

And also many thanks to everybody that has given their time and efforts to describe technical questions in an understandable way (besides I am not an English speaking person)

I do not post in CA but I usually follow it and really enjoy discussions like this.

[jabbr]

That paper talks about inherent problems with 1-bit sigma-delta converters, and it is quite correct. It explicitly states that multi-bit sigma-delta is perfectly fine. Almost all modern DACs use multi-bit converters. Guess why.

 

Ha ha, well if you don't believe anything over 20 kHz is important, and according to their graph the noise level roughly -150dB with DSD64 then what's the problem? Oh right...

 

In any case if we upsample then that rising noise level get's pushed out farther and farther. An excellent single bit sigma delta converter which can be entirely analyzed because the circuit is entirely published is the Signalyst DSC1. You would see that the digital FIR filter suppresses noise > 80 kHz or so and then the 4 stage SK filter with a corner at 100 kHz suppresses the generated noise which at DSD512 is pushed into the MHz range.

 

To my reading however, the Grimm article is comparing PCM 24/192 to DSD 64 and if we are talking purely NOS DAC, that is a reasonable comparison.

[mansr]

Ha ha, well if you don't believe anything over 20 kHz is important, and according to their graph the noise level roughly -150dB with DSD64 then what's the problem? Oh right...

 

In any case if we upsample then that rising noise level get's pushed out farther and farther. An excellent single bit sigma delta converter which can be entirely analyzed because the circuit is entirely published is the Signalyst DSC1. You would see that the digital FIR filter suppresses noise > 80 kHz or so and then the 4 stage SK filter with a corner at 100 kHz suppresses the generated noise which at DSD512 is pushed into the MHz range.

 

To my reading however, the Grimm article is comparing PCM 24/192 to DSD 64 and if we are talking purely NOS DAC, that is a reasonable comparison.

The Lipshitz/Vanderkooy paper is about sigma-delta modulators, not the D/A stage. While excellent performance is possible, there is no denying that 1-bit modulators of 3rd order or higher are inherently unstable and exhibit various forms of distortion. A multi-bit modulator does not (necessarily) have these issues. Given the choice, a multi-bit architecture is therefore preferable.

[Fokus]

Hi Guys - I just edited the article with the following statement at the top.

 

Well done. This is appreciated.

[Jud]

The Lipshitz/Vanderkooy paper is about sigma-delta modulators, not the D/A stage. While excellent performance is possible, there is no denying that 1-bit modulators of 3rd order or higher are inherently unstable and exhibit various forms of distortion. A multi-bit modulator does not (necessarily) have these issues. Given the choice, a multi-bit architecture is therefore preferable.

 

Would you say "various forms of distortion" are characteristic of the DSD recordings on offer? What about audibility of such distortions if/where they do exist?

 

 

Sent from my iPhone using Computer Audiophile