Why 24/192 is a bad idea?

Well the notion that you can sue anyone for antyhing is of course an urban legend. Rule 111 and abuse of process are just to examples. F.R.C.P 12b6 is anither example.

Of course fiiling and prevailing are two different arguments.
 
That's the problem! Care to have a listen to the files I posted? You are always asking for proof & files to be posted so time to listen to some files & PM me your results.

Sure, point me to the links. Though I don't know how listening to your files will prove that bit depth affects more than the noise floor. Can you explain further? It seems to me this is more suitable for mathematical or graphed evidence. Either the noise floor only is affected, or more is affected.

I have to add that I'm satisfied with my proof because nobody has yet refuted it other than to claim I'm wrong with no further evidence. "Just listen" is not evidence, nor is saying the sound is affected using vague audiophile terms like "bleached." Again, this is settled science, so the burden of proof is clearly on those who believe otherwise.

--Ethan
 
There in post 187 which I'll repeat here:

Test Files found here https://docs.google.com/open?id=0BzRUtXo0Yuj5SERtcEVKZlB5YWc
There are 4 files (each about 19 secs of audio) - some of them are the same files.
Listen to the whole file but particularly the rim shots in each for naturalness. Best to PM me with results so as not to pollute this thread or coach others.

I will reveal how the files relate to your question when you have listened to them & reported your results

Sure, point me to the links. Though I don't know how listening to your files will prove that bit depth affects more than the noise floor. Can you explain further? It seems to me this is more suitable for mathematical or graphed evidence. Either the noise floor only is affected, or more is affected.

I have to add that I'm satisfied with my proof because nobody has yet refuted it other than to claim I'm wrong with no further evidence. "Just listen" is not evidence, nor is saying the sound is affected using vague audiophile terms like "bleached." Again, this is settled science, so the burden of proof is clearly on those who believe otherwise.

--Ethan
 
Focusing a bit on downquantizing.
Xiphmont, are you talking about changing the bit depth, say from 16bit to 24 bit, and 24bit to 16bit (I am assuming is the downquantizing)?

Further quantizing from a higher bitdepth to a lower one, such as 24 bit to 16 bit.

Anyway it has given me a great idea for an article and just deciding whether to ping-pester Paul Miller or Keith Howard with the suggestion.

...do tell?

Monty
Xiph.Org
 
If the original sound of the instrument produces intermodulation products because of its ultrasonic content, then that is the true sound of that instrument in nature.
Lee

No instrument does so. Audible intermodulation would be purely a result of the recording and playback chain. Thus my assertion that ultrasonics can only harm (however rare that may be) and never help.
 
But FFTs are histograms, there's no 'floor'.

FFTs have an equivalent integrated bandwidth per bin, so there is indeed a floor, and we even know exactly how it was computed (as he gave us the window). The floor can be made to apparently move around depending on the window type and length, so this doesn;t mean anything absolute. However, we're not making absolute comparisons; the comparison is relative. For this reason, Ethan's FFT is a valid comparison.

Added to that, an FFT is a time exposure.

So is noise itself, so is hearing itself. In what way does the length of this FFT render the point invalid? I do hope we're not going to get into the validity of time/frequency duality.

How about providing evidence for your claimed love? Like putting up a 'properly implemented' DAC in response to my earlier request?

I'm out of town right now... but once I'm back home with my lab, I'll be able to fufill this request.

Monty
Xiph.Org

hmm, I need to avail myself of the automated sig feature.
 
It was said that the filter in the DAC is a walk in the park now because there is "digital oversampling" in the DAC. Just because something is in the DAC and is "digital" it does not mean at all that it is perfect. Indeed, we can't perform perfect processing there.

We can perform digital processing perfect to within any arbitrary specification.

Interpolation has accuracy limits and implementation constraints.

It's incorrect to call oversampling or resampling 'interpolation', as no interpolation is happening.

Leaving the terminology debate aside, 'implementation constraints' itself implies that these constraints are somehow of damning practical concern. Once you're 160dB-180dB deep before even resorting to floating point, I should think anyone's requirements for a DAC have been met.

What the designer chooses is a trade off. He makes the filter design easier but then builds in distortion that was created by the interpolator.

Digital resampling is practically free of distortion (again, we're talking about 180dB down, not 60dB).

by resampling to a fixed target clock, they sharply reduce source induced jitter. But then they have a similar problem in that they need to constantly figure out the ratio of incoming clock to outgoing and that adaptiveness (sp?) creates its own forum of distortion.

No, it does not. Taking ASYNC mode ISOCHRONOUS USB as an example, the DAC always clocks its samples to a high accuracy sample clock. The samples are taken from a buffer (FIFO) that holds a few ms of audio. If the DAC clock is a little faster than the host clock, the fill level of the FIFO drops slowly and the host is requested to increase the number of bytes per packet to keep up. If the FIFO is slowly filling, the host is requested to slightly reduce packet size.

There is no distortion caused by this mechanism as it's completely divorced from the DAC. It can go wrong if the FIFO underruns or overruns, but we all agree this is an unmistakable catastrophe.

Monty
Xiph.Org
 
What do you find vague about what I already quoted from the paper "They found the pre-echo due to a filter ripple of ± 0.2dB with a span of 23Hz corresponded with echoes of -32dB at ±40ms - which was found to be quite perceptible even with untrained listeners

I'm still trying to figure out exactly what they meant by that myself. Passband ripple doesn't cause preecho. And just about anything time-domain isn't going to cause 'echos', it's going to cause Gibbs effect.

So I'm scratching my head too. I'm sure it meant something, just can't figure out what.

[edit: oh, wait, no I think I see what they're saying. They are complaining about Gibbs after all.

BTW, this is not the same effect as adding a true preecho (leading copy) of the original signal at -39dB, so demonstrating a true preecho is audible is not equivalent. I'm not sure that's what jkeny meant though.]

Monty
Xiph.Org
 
Last edited:
More seriously, there have been NO significant arguments advanced here to support the original idea that 24/192 music is a bad idea.

The point of the original article was that ultrasonics are never a benefit. Given that they're never a benefit and can be detriment (even if it's a small detriment, and only occasional), why would we ever want ultrasonics?

I think most audiophiles would say "If one out of 1000 people can tell the difference between normal and high-res, that's good enough reason to use the higher rate." I hope you're not shocked to find I agree; 1 in 1000 is more than enough for me. I only insist the result be reproducible and verifiable.

However, we currently have a case where more than 1 in 1000 people (?) will hear the difference--- only because the playback is introducing distortion (from ultrasonics) that _did not exist in the original performance or the original recording_; it is purely an artifact of imperfect playback.

I already said that the format should not be based on assumed limitations of downstream equipment, and I believe that. However, we have a case where the ultrasonics provide no benefit whatsoever, and can actively harm playback fidelity. God help anyone with a tube amp.

[edit: actually... who here would like to try out the IMD test samples on their tube amps? :-]

Monty
Xiph.Org
 
Last edited:
It is ironic that the proof point of this is in one of the other references in your article, namely the Meyer and Moran test of high-resolution audio:
[...]
Putting aside the point that the paper does indeed provide counter evidence of the "harm" that you say exists with respect to ultrasonics, here is the salient point related to mine:

M&M were using fantastically expensive equipment. There was no audible IMD. That's not a big surprise.

Absence of harm is not the same thing as a benefit.

"A NOTE ON HIGH-RESOLUTION RECORDINGS
Though our tests failed to substantiate the claimed advantages of high-resolution encoding for
two-channel audio, one trend became obvious very quickly and held up
throughout our testing: virtually all of the SACD and
DVD-A recordings sounded better than most CDs—
sometimes much better.
Had we not “degraded” the sound
to CD quality and blind-tested for audible differences, we
would have been tempted to ascribe this sonic superiority
to the recording processes used to make them.
Plausible reasons for the remarkable sound quality of
these recordings emerged in discussions with some of the
engineers currently working on such projects. This portion
of the business is a niche market in which the end users are
preselected, both for their aural acuity and for their willingness to buy expensive equipment, set it up correctly,
and listen carefully in a low-noise environment.
Partly because these recordings have not captured a
large portion of the consumer market for music, engineers
and producers are being given the freedom to produce
recordings that sound as good as they can make them,
without having to compress or equalize the signal to suit
lesser systems and casual listening conditions.
These recordings seem to have been made with great care and
manifest affection, by engineers trying to please themselves and their peers.
They sound like it, label after label.
High-resolution audio discs do not have the overwhelming
majority of the program material crammed into the top 20
(or even 10) dB of the available dynamic range, as so
many CDs today do."


So as you see, as a practical matter, what you are advocating does result in poorer sound.

What you are advocating is being unable to sell better masters to an audiophile without the added expense of senseless recording overkill, because said audiophiles have been 'educated' that they need a gold plated Hummer with artillery mount to drive to the corner store.

Is this a benefit to the industry? to anybody?

Not because of any technical reasons but for a business one.

I actually agree with your business point to an extent, but it's due to you painting yourself into a corner (as a whole industry, not you personally Amir) and you're here continuing to paint yourself tighter and tighter into the same corner.

More than a few of your potential customers are on Apple Insider, Audio Asylum, Computer Audiophile, etc, stating they're holding out for 384kHz because 192khz just has too many compromises. They're serious. These are loons of course, but this de-education is going going to come back to bite somehow.

As I have said and continue to repeat is that I do not have a dog in this hunt.

So you've jumped in with a cat, a squirrel and six musical gerbils, to argue.... what exactly?

You seem to be running interference in this without articulating any reason as to why.

Might have something to do with running an educational foundation. I also really hate it when people are wrong on the internet.

The question is moot. It has been and I have said it multiple times. It matters not if we can prove technical superiority when actual superiority is abundantly clear. And that there is already a marketplace for it.

That's what the Catholic Church told me, but the Buddhists got all huffy.

I started with experimental results. The experimental results say your statement is incorrect. The theory provides some explanation, but it's the experiments that talk.

You're not the only person who would love to show me up on these points--- after all, other industry insiders do have plenty of dogs in this race. Several PHOsters suggested the idea of renting out a suite or finding a local studio at SXSW next year to do informal but experimentally sound listening tests. It would be off-record, no official results, no 'lording it' over one side or the other. Personal enrichment only. Interested?

Monty
Xiph.Org
 
Last edited:
Of course fiiling and prevailing are two different arguments.

Yes, of course! And in practice, anti-SLAPP laws have helped to an extent in personal litigation.

However, patent law is not like civial or criminal law; he who has the patent is generally presumed to be the authoritative party in a lawsuit; the defendant is guilty until he proves himself innocent, and that can only happen if the trial is seen through to victory.

A successful defense in court is the only way you can legally state that you certainly don't infringe a patent. That's why so much language around patents sounds evasive.
 
FFTs have an equivalent integrated bandwidth per bin, so there is indeed a floor, and we even know exactly how it was computed (as he gave us the window).

Ah, the text was too small for me to pick it out, I've enlarged that now and agree, with the information provided its possible to calculate a equivalent bin bandwidth.

The floor can be made to apparently move around depending on the window type and length, so this doesn;t mean anything absolute. However, we're not making absolute comparisons; the comparison is relative. For this reason, Ethan's FFT is a valid comparison.

Its not 'proof' that 'the only thing that changed is the noise floor'. Just posting up two FFTs which look identical isn't proof that the two signals involved were identical. These don't look identical even when the noise 'floor' is factored out - the smearing effect of the window is different.

So is noise itself, so is hearing itself. In what way does the length of this FFT render the point invalid? I do hope we're not going to get into the validity of time/frequency duality.

The noise need not be constant over the acquired window. In the case of quantization noise there's the general assumption at high levels of signal that tnere's no correlation between it and the wanted signal. However this is an approximation, the correlation as I understand it increases as the signal level decreases. There needn't be 'a floor' - what the FFT shows us is the averaged nose (in this case over 65k samples).
 
For me to "miss out" on something means that the person mixing the content did indeed hear ultrasonics. And that he had a positive experience as a result of it. In that regard, we better all group together to quantify and measure it so that we can reproduce it at home. Would you like to join us in that? Or should we conclude that it is best to not go there? ;) :)

Oh, my last reply was general, not even specific to ultrasonic content. After all, if ultra-sonic IMD is good, then so is standard harmonic distortion and IMD from content in the audible range. So you have to make sure your speakers reproduce at least as much of that distortion as standard studio monitors, and avoid speakers that are too linear.

As for ultrasonic measuring issues, I think it's a great idea. I'm not sure where to start though. Can you provide a 24/192 sample where the ultrasonic IMD is audible so we can start working from there?
 
Oh, my last reply was general, not even specific to ultrasonic content. After all, if ultra-sonic IMD is good, then so is standard harmonic distortion and IMD from content in the audible range. So you have to make sure your speakers reproduce at least as much of that distortion as standard studio monitors, and avoid speakers that are too linear.

As for ultrasonic measuring issues, I think it's a great idea. I'm not sure where to start though. Can you provide a 24/192 sample where the ultrasonic IMD is audible so we can start working from there?

Seems like there's a basic misunderstanding here. Ultrasonic IMD in digital recordings is whatever we make it to be during production. All of the available relevant formats have low enough inherent distortion that it need not be an audible issue at all.

But, the same potentially pure signal passes through audio systems whose performance varies all over the map. Consider the HF nonlinear distortion in this multi-killbuck$ amplifier:

207cayin.13.jpg


The above test was made at 4 watts. Other tests suggest posted as part of the same review suggest that distortion is several times greater at a mere ten watts, and may also be twice as great at 40 KHz, We're talking 10% or more nonlinear distortion and still several dB below below clipping!
 
Its not 'proof' that 'the only thing that changed is the noise floor'. Just posting up two FFTs which look identical isn't proof that the two signals involved were identical.

Well, strictly speaking they probably weren't as the white noise was probably different each time (unless it was the exact same offset of the exact same signal each time). I know that's not what you meant, it was mostly just a comment for the peanut gallery.

If you had anything other than white noise + a signal, that would be visible in the FFT. I suppose you could have a very slowly time variant white noise, but there's no reason to think that would happen in the example above.

These don't look identical even when the noise 'floor' is factored out - the smearing effect of the window is different.

I think you're being thrown by the scale. They do in fact look identical to me when overlayed.

The noise need not be constant over the acquired window. In the case of quantization noise there's the general assumption at high levels of signal that tnere's no correlation between it and the wanted signal. However this is an approximation, the correlation as I understand it increases as the signal level decreases.

IIRC, It increases relative to the original signal, not in an absolute sense. The absolute value only goes down from what you see in the graph above (Ethan, is this correct? I'd doublecheck now, but I'm about to get on a plane).

There is no correlation whatsoever for appropriate dithered quantization (whether it's white or shaped), which is the point of the dither.

Monty
Xiph.Org
 
No instrument does so. Audible intermodulation would be purely a result of the recording and playback chain. Thus my assertion that ultrasonics can only harm (however rare that may be) and never help.

My original post was poorly written. If the ultrasonic components of an instrument's sound contribute to its unique tonal signature, then all that information should be preserved. We should not cut it off, since a linear system reproducing the signal would then produce a different sound of the instrument without that content.

Lee
 
My original post was poorly written. If the ultrasonic components of an instrument's sound contribute to its unique tonal signature, then all that information should be preserved.

Yes, that is how I interpreted your original message. The ultrasonic components are inaudible, do not affect the audible components, and contribute nothing to an instrument's tonal signature.

The only case we're worried about here (and worry is the right word) is that preserving these useless ultrasonic components in a recording can cause them to interfere with playback fidelity due to limitations in a DAC, power amplifier or loudspeaker. The best case is that ultrasonics contribute nothing, as they would have contributed nothing in the live performance. The worst case is that ultrasonics cause additional IMD in the playback due to inadequacies of the playback equipment; this is a situation analagous to clipping, wow+flutter, etc. It is purely detrimental. It can have no positive contribution.

Monty
Xiph.Org
 
Last edited:
Yes, that is how I interpreted your original message. The ultrasonic components are inaudible, do not affect the audible components, and contribute nothing to an instrument's tonal signature.

The only case we're worried about here (and worry is the right word) is that preserving these useless ultrasonic components in a recording can cause them to interfere with playback fidelity due to limitations in a DAC, power amplifier or loudspeaker. The best case is that ultrasonics contribute nothing, as they would have contributed nothing in the live performance. The worst case is that ultrasonics cause additional IMD in the playback due to inadequacies of the playback equipment; this is a situation analagous to clipping, wow+flutter, etc. It is purely detrimental. It can have no positive contribution.

Monty
Xiph.Org

I understand what you're stating. To make a blanket statement that preservation of the entire recorded signal (in this case, specifically the ultrasonic content) is purely detrimental, seems one dimensional in approach to me. There is a lot of literature describing ultrasonic hearing via bone conduction, etc. These effects can certainly be an integral part of the sensory experience of listening to a live instrument. In a simple standard auditory test-tone audiometry world, ultrasonic frequencies have no place in human hearing. However, in a more completely-encompassing view of auditory perception, both sub and ultrasonics play an important role in how we perceive sounds around us.

Lee
 
There is a lot of literature describing ultrasonic hearing via bone conduction, etc.

There isn't much data to support it though. Like I said, getting published doesn't make a paper right, having it reproduced and supported by additional research [eventually] does. The press is especially bad at getting this part.

These effects can certainly be an integral part of the sensory experience of listening to a live instrument.

I don't think 'certainly' is the right word here. There are plenty of things about a live performance that a recording can't capture; the ambiance, the context, the mood, the setting.... but these have little to do with the sound.

I remember as a kid I loved the sound of the cello and double-bass and wanted rather badly to learn how to play (and could afford no such thing, so instead I became a singer--- no expensive instrument involved, and I could get lessons in public schools). Eventually, a friend let me play a few notes with coaching on his bass... and I was terribly disappointed that it felt nothing like I expected. The warm, earth-shaking depth that seemed to be there just a few feet away wasn't there when I held the instrument in my own hands. It was there for the audience, not the player. To say that was a great disappointment is putting it mildly.

Some years ago, I decided I needed to pick up an instrument before I turned thirty or I never would. I still had the wonderful resonance of the double bass in mind when a friend in a pit introduced me to the first bassoon player. He pulled out a fresh reed for soaking, gave me a few tips on how to force out the merest note, I worked on getting the naked reed to voice, we put the reed on the bocal, he gave me the fingering for a B flat... and told me to give it a go. A bassoon may sound like a farting duck to to audience but the all the depth the bass lacked was there in my hands. I felt like I was holding a running diesel engine in my bare hands.

I don't think the vast majority of recordings do the bassoon any justice. But it's not the imperfection of the equipment, it's that the recording lacks everything but the sound, no matter how perfect the sound is. More bits and higher frequencies don't replace the feeling of an truck engine tucked under your arm (or that one more fff whole note and your embrochure is done for, and the oboes will never let you hear the end of it).

In a simple standard auditory test-tone audiometry world, ultrasonic frequencies have no place in human hearing.
Lee

But who says tests concern only tones? Tones have a prominent role in testing only due to the fact they're easy to handle and make many tests possible that otherwise wouldn't be. Testing does not and never has relied only on tones (or noise, or synthetic signals). The 24/192 vs 16/44 tests I linked in the article didn't use synthetic audio.

However, in a more completely-encompassing view of auditory perception, both sub and ultrasonics play an important role in how we perceive sounds around us.

Sub: yes. No one argues about this, because they do matter, and there's something actually perceptible there for tests to verify.

Ultrasonics, I do not concur. They have nothing audible [or otherwise perceptible] to contribute. Moreso, it is easy to convince me otherwise... Construct a reproducible test that demonstrates their contribution. There have been quite a few attempts to do this, none have suceeded so far.

Monty
Xiph.Org
 
Yes, of course! And in practice, anti-SLAPP laws have helped to an extent in personal litigation.

However, patent law is not like civial or criminal law; he who has the patent is generally presumed to be the authoritative party in a lawsuit; the defendant is guilty until he proves himself innocent, and that can only happen if the trial is seen through to victory.

A successful defense in court is the only way you can legally state that you certainly don't infringe a patent. That's why so much language around patents sounds evasive.

I'm not a patent lawyer but generally speaking anyone who has statutotry proterction has the "upper hand." Once a patent is granted the holder is pretty much own his own. By the time it winds it way through the courts the holder is often broke.
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu