-

Hmm, very long article. Seems to be a collection of audio information; some of the usual "vinyl is better" BS. Some garbage about limits of 96dB dynamic range for 16 bit.

I skimmed it, probably missed a lot, but seems a bit dated (2001).

Was there something in particular you were considering?

the very topic of the link...
he is claiming that we can't differentiate 44kHz vs. 48kHz vs. 96kHz... but one COULD know the superior difference of audio at 24bit and not 16bit, but to me is all rubbish

This has been covered to death on this forum already. Do we really need to start a new thread on the subject?

Take the discussion here:
http://www.hydrogenaudio.org/forums/index....showtopic=40134

It's a load of pseudo science. The guy has a hobbyists interest, has heard some audiophile bull, but has no understanding of the theory.

Of course circa 1990 8-bit audio sounded awful. It was sampled at goodness know what rate, typically with no dithering, no anti-alias filters on record, and no anti-image filters on playback.

However, Correctly generated 44.1kHz sampled 8-bit audio (e.g. from foobar set to convert to 8-bits) has the one and only problem that's created by having too low a resolution: noise. No distortion, no loss of fine detail, no grittiness or grainy character - just background noise, plain and simple. Human listeners can hear through a surprising amount of noise.

_any_ correctly sampled digital audio has the same issues - a bandwidth limit and a noise limit. 16-bit audio has noise at -96dB FS RMS.

I'm sure you're aware of the extensive threads in the FAQ that cover this subject in horrendous detail.

Of course, given the limitations and issues with field recording, the lower noise floor of 24-bit digital audio is genuinely useful. You can set your levels conservatively, and easily avoid clipping while keeping everything way above the digital noise floor. With very good microphones and pre-amps, and a need to record without compression or processing or even a perfect idea of the levels which will hit the microphone, and wishing to avoiding clipping, 16-bits could be considered an issue.

Cheers,
David.

EDIT: Reading down, this person doesn't understand dither. He's bought the bogus argument that is "masks" distortion, rather than eliminating it. No wonder he's so confused.

EDIT2: Oh dear, he thinks Nyquist is something to do with the limits of human hearing. He thinks noise shaped dither is useful because the noise may be later filtered out.

EDIT3: And it's rather out of date. I think most of us are now managing to run IE and MS Office without it preventing us using our PCs successfully for audio!

@greynol

I'm not starting a topic on that subject.
Just wanted to verify that page.
Other people just replied nicely to the question, but you, why do you come into this with a kick in the bum? No one asked your opinion...

The page appears to be a source of infinite wrongness. Starting with the "imagine 4 levels" if not sooner.

Yes you did, you asked for comment on if the page was correct implicitly asking for the opinion of the people reading it on the piece. The page is Worse Than Wrong, it appears to someone not skilled in the art to be authoritative but in fact totally incorrect.

It appears the site (and the article) are oriented towards recording. More specifically "should I use a 16bit DAT or a 24bit DAT" (or HD recorder, etc.). In this context, he's mostly right. If you're recording at 16bit, and need to EQ an area of the track to make something more audible, you're going to get a lot of distortion awfully quickly.

He doesn't confuse Nyquist with human hearing range, he simply mentions the two in the same sentence. Although there's certainly some errors, the rest of the information is "accurate enough" for the subject matter of recording and the points he makes are valid.

With recording, there is a huge difference between dealing with 16bit sources and 24bit sources. The HA groupthink skepticism of 24bit audio is related to playback of properly mastered and dithered music. Don't compare apples and oranges.

Hang on a second benski,

There's no argument about the possible benefits of 24-bits for recording.

However, it takes one sentence to say that.

This link points to a whole page where almost every justification and technical "fact" is wrong, but presented as if the author understood the subject perfectly.

I don't think the net need more pages of misinformation, and certainly not ones that look so informative!

Cheers,
David.

Hmm, I just skimmed the first part, but the argument seems to be similar to the argument for the need for more bit depth in image recording. In that realm, though, it's actually a real problem both because of the nature of that domain as well as the fact that they're still working at only 8 to 12 bits per channel for the most part.

The distribution of values is much more felicitous in audio than it is in imaging, though, so I think this is less of an issue. Additionally, the bit depth is already at a higher level than available in my 12 bit camera . In imaging, where it's a lot easier to actually perceive the distribution of values, 16 bits per channel technology is doing quite well.

This seems like typical hi-fi fare to me, though. Give plausible arguments with information that seems technically accurate and then draw pure conjecture from it through affective language and appeals to existing biases in the reader. It'd be nice if there were more in the way of critical analysis and inquiry in the hi-fi hobbyist community, but I've gotten the impression that there just isn't an interest, let alone a place for it in much of the hobby.

Who, believing in the gods, and their work in the world, wants to be distracted from such exciting mystery by facts and reason?

Bourne - everyone's ears develop at their own pace. Applied practice of reliable listening tests with controlled data, one audio sample against another, is needed to begin to pick up fine differences such as bit depth nuances.

Dan Heend is a digital recording pioneer who's article is still relevant today. It's hard to find good teachers in this field these days but Dan is apparently one of them. I'm interested to hear what he has to say on the topic of jitter.



No one understands what you've written here.

I suspect most of us who visit here are capable of offering out own opinions when we want to; we have no need of a spokesman. I also suspect that most of us, as usual, can understand David fairly well, even when not always as versed in the more technical aspects of the topic. Of course, I can only really speak for myself. My hypothesis is simply based on many observations over an extended time.

On the contrary, I suspect that most of the regulars here at HA understand David quite well because these are topics that have been discussed frequently and at much greater depth than David's summary. If you are unable to understand it then that is what you should say. And if you are unable to understand it then I suggest that you stick around because I am sure that these topics will come up again.

Please do not speak for the group. If you do not comprehend something ask for assistance. Enough said on this.

etc. etc. etc. no need to quote any more bogusness.

<yawn>

It is very rare that I post anything on this topic these days, as I have no need whatsoever to prove to anyone what I know about audio music recording/engineering/production. The many professional musicians and engineers I've worked with through the years know I stand on terra firma.

However, I thought I'd crawl out from under my audio engineering rock for a moment to respond to this nonsensical seemingly personal attack, dated as it may be, given that internet search engines seem to have a way of capturing things into eternity and making them seemingly relevant in the present.

About the 24-bit Field Recording FAQ:
While specifically dated 2001 and candidly notated as such for the purposes of managing the readers' expectations of topicality, the words I wrote in the [now formerly] hosted document you refer to were indeed correct, lucid, well considered statements and have been proven valuable and valid to countless field recordists. Your's and a couple of other scarely qualified comments here notwithstanding, I have received through the years more positive comments from novices and professional audio engineers alike than I could ever have responded to. To imply that I don't know what I'm talking about is to imply that guys like Ken Pohlmann, Bob Katz, and so many others I've read and personally spoken to don't know what they're talking about.

Having applied all of the concepts stated in my document in depth and detail throughout my PROFESSIONAL, non-hobbyist audio engineering/recording career, I find it hard to accept criticism from an audio plugin developer who doesn't understand that the difference between *any* live sound and *any* recording thereof can be quantified in some manner by the use of either words "distortion" or "noise." Further for anyone to say that 8-bit audio suffers from only noise and not distortion is a most ridiculous statement that really doesn't even deserve the lengths I've gone through to address it up to this point. Needless to say, I prefer to devote my time to debating topics like this with audio professionals with proven knowledge, reading comprehension, and experience, so I have nothing further to say on that, since it would seem obvious who is lacking a clue.

In the 7+ years I paid to host my FAQ, I never once had the need to update or modify the fundamental principles and explanations upon which it was based. Not once as a result of any qualified comment or suggestion from other audio engineers in the field have I felt the need to change the date and update the underlying fundamentals. Given my well established penchant for accuracy and completeness, you can bet if anything needed to be clarified or corrected further, and it was brought to my attention in the past 7+ years, I would have made such corrections.

The document was neither created nor hosted for any purposes of personal ego--but rather philanthropically by me as a gift to all aspiring recordists and audio pros, giving back in the same tradition in which I was privileged to have gotten a start from kind audio pros like Jay Serafin back in '94.

Sure, there were some things in it related to the limitations of computers suitable for laptop recording at the time, and the limitations of available non-laptop-based recording devices.... things that many might have hoped I would have kept up to date. Through the years my initial intentions of keeping those types of items up to date were impeded by my lack of time and my desire to allocate my time not looking back, but rather forward to many musical, audio engineering, and computer projects. Rather than take the document down, I made a conscious decision to continue to host it as is, figuring that most would still find the majority of it quite valuable and useful... and I've continued to receive mail to that effect.

My partner and I record, mix, and master live music for legendary bands and musicians who recognize the talent we have for producing a high quality product doesn't come from casual hobbyist noviceness or blowhard "audiophile bull," as you put it.

David, judging from your persistence in beating a topic to death and bashing me in the process, perhaps the obvious chip on your shoulder is weighing on your arms preventing you from typing a valid point. You just keep writing your plugins, and leave the audio engineering to the pros.... because you certainly can't possibly have any more room in your mouth for another foot.

My most sincere apologies to all for dredging up this old stinky thread, and sincere thanks to those like Pandabear who posted with rationality in my defense.

Onward to better people, places, and things.

-DH
High Fidelity Microsystems
HFM Studios
www.hottunatunes.com

P.S. If anyone wants to host that old and yet somehow relevant doc, lemme know.

P.P.S. David, if you'd like to discuss this further, please keep it to email, so I can ignore your emails you while other kind folk here are likewise spared the drivel.

Let's see, I could believe David Robinson, or I could believe you. No contest, David is right and you are wrong.

In this case I knew the answer even before David confirmed it, but having an expert of David's stature agree with you is a great ego booster.

I was curious what got Dan Heend so upset after all these years, so I found an achieve of the site:

http://web.archive.org/web/20070103154338/...w.24bitfaq.org/

Reading through the first couple paragraphs, I have to agree with 2Bdecided. While Dan appears to have plenty of experience recording, his understanding of the theory is severely lacking.

"A passage that is 6dB louder than another passage is said to be twice as loud as the other passage."

First obvious mistake - 6dB is four times louder, not two. Sound level goes as the square of pressure.

Rather then antagonistically pile on minor arguments about how one defines loudness (pressure verses perception), maybe we could wait for 2Bdecided respond.

Agreed

So it IS possible for people who get their Theory wrong to still excel in Practice. Very interesting.

I believe I have adequately summarized the content of his post.

ccryder: If you wish to be taken seriously, use science, not rhetoric and anecdotes. You look foolish, and if there is any truth to your claims (the 24-bit FAQ appears to liberally sprinkle nonsense on top of some correct theory), you need to address the critique of your claims, not make an appeal to experience or an ad hominem attack on the abilities of another.

That depends... you know... sometimes to excel in something is related to the money put into it. (or in this case, the amount of bits)

There are several small errors around the document, like "4-bit recording would have 16 discrete possible amplitude levels.". a 4bit recording just have 8 amplitude levels, because a signal has a positive and a negative part. This one is done several times.
(And what about "Perhaps many are more familiar with 8-bit audio from real-time internet sources like RealAudio". that was audio compressed at 16kbit/s, not just "8-bit" !)

But one of the things that made me wonder is how 24bits (as opposed to 16bits) actually makes vinyl lovers happier. AFAIR the SNR of a vinyl is lower (i.e. less range) than that of a CD.
Either one doesn't like digital audio (and argues that just analog media can store the signal in enough detail), or accepts the way digital works, and compares what is comparable (i.e SNR)

The problem with the document, from my point of view is: It says something that is acceptable for its use (recording), with some correct information, but also with other that are mistakes, misunderstandings or erroneous concepts.
I am not implying that the latter are more prominent than the former. Just that they are there.



If i'm reading that right... It is *plain* arrogance.

http://www.ews64.com/mcdecibels.html

The topic was dynamic range and PCM quantizations/representations of dynamic range, not sound pressure levels.

Please See
http://en.wikipedia.org/wiki/Audio_bit_depth
and
http://www.wikirecording.org/Bit_Depth

Perhaps you take issue with my use of the phrase "amplitude levels," which for the purposes of the discussion, I considered synonymous with sample values (levels describing amplitude of a waveform, which may be positive or negative numbers, but are still discrete values representing amplitude).

Regarding the Real-Audio reference, never did I mention the phrase "bit-rate". I said 8-bit. As in 8 bits per sample representing a maximum of 8 significant bits required to quantify dynamic range, *not* 8 bits per second. Bits per sample was the topic. My experience with Real Audio by the time the doc was written was that Real Audio was played back using 8 *significant bits* per PCM sample (regardless of word lengths > 8-bits) once decompressed from Real Audio format and having been fed to a D/A converter. If that has changed since then (I don't mess with Real Audio, so I don't know how far they've pushed its dynamic range losslessly), then any inaccuracies with respect to bits per sample of the Real Audio product would be a function of the information I believed accurate at the time it was written.

One might get a greater appreciation for what it is that makes audiophiles (and "vinyl lovers") prefer 24-bit audio by reading either some sections of my doc (if you dare), or reading more about how PCM vs. DSD represents sound, especially low level signals. Simply put, the attraction to either high resolution 24 PCM digital audio, or analog audio is a more accurate representation of lower level signals--DSD having its own argument in that regard.

Might have been nice to hear some of your suggestions for corrections over the past 7 years, especially if you were around back when the doc was written in 2001. The doc is currently officially unhosted (it was my specific request to 3rd parties hosting a copy of the doc to do so solely with my permission--to date I've granted none), so that point is moot.

I do look forward to reading a publicly hosted pioneering educational document my "critics" here might produce in the future, as I'm always eager to learn more and improve my knowledge on topics that have yet to gain mainstream awareness, and it would appear a few here have much to say.

Bottom line to anyone who doesn't think the FAQ is accurate, or you don't agree semantically with some of my word choices used to describe a complicated topic in a document specifically geared toward newbie high resolution field recordists, don't read it... It's that simple.

For reasons of personal economy, personal feelings of fulfillment of the original goal to educate new recordists on the use of new technologies, and relevance given the availability of capable standalone high-res recording devices, I pulled the hosting of the doc long before I found this thread 2 days ago, even though the thread is a year old. If anyone here had really cared about the accuracy of the document a year ago (or longer), you had a year to contact me and propose your corrections. I received no such propositions, which shows me that there are a few folks that would rather pretend to care about "misinformation being disseminated" rather than actually do something to prove they care (like emailing me about it) beyond classlessly bashing me behind my back on a public board on which I wasn't even a member (much less a reader) until yesterday.

At this point, I consider the doc out of print, anyway, so perhaps my few detractors might consider themselves to have won some kind of bizarre year-long battle I never knew I was in... hey, whatever gets you through the night.

-DH

Doesn't "twice as loud" (your words) imply a perceptional scale? In psychoacoustics it's known that +10dB corresponds to doubling perceived loudness. "loudness" and "amplitude" are terms you seem to have used interchangably in the sections of the FAQ I checked.


It depends on your definition of distortion. David probably refers to other types of possible distortions (i.e. harmonic distortions etc) excluding noise where "noise" means in this context: added noise with characteristics (power & tone) being independant from the signal. I see that you covered the dithering topic...

and yet you came up with comments like

"sterile", "lifeless" are prime examples of audio bull lingo some of us react allergic to. You could have said that at lower signal levels the signal-to-noise-ratio is simply lower.

Note: I'm far from denying your authority in the field you're working in. The comments of yours just suggest that the level of understanding for some of the things involved isn't above the "intuitive level".

Cheers,
SG

Cursed are the peacemakers, for they shall be beaten up on by both sides, but:

@ccryder

this year-old thread was not a gratuitous bashing of your 2001 FAQ; a member asked a question about it, I'm sure in good faith. Alas, he then deleted his original post, so you don't get the context.

If you look at the discussion as a whole, you'll see that it is real discussion, with criticism of some technical aspects. No one, as I read it, is questioning your professional competence in your field, but there is some questioning of some of the technical explanations in your document.

I don't have any competence to judge any of that, but it's not a case of an unmotivated attack out of the blue.

You need to be more careful with your terminology. Misuse of the term "louder" here can lead the uninformed reader to false impressions later when you talk about sounds that are 48 dB quieter. Instead of thinking that 48 dB is 2^16 times quieter, the reader could think that 48 dB was only 2^8 times quieter, which is a vast difference.

Yes it may be a little harsh to criticize in 2007 & 8 an article that contained the following qualification and request for input/feedback:

While the contents of this document are specifically targeted to the needs and concerns of field recordists, some of the content can be applied to home recording as well. The submission of additions, corrections, and comments, is requested and encouraged ...

I imagine many people would have noticed the odd academic error in passing, but would have been content to read the article broadly; for its practical guidance, in using what was new technology at the time.

Voice of reason, words of truth, on the money.
Thank you for that post. Someone gets it.
If I had wanted to write the bible of digital audio, I would have signed it,
-G_d

Fortunately for many aspiring field recordists in 2001, they didn't have to wait that long for a good read. Regardless of what anyone thinks today of my work, I know the positive effect it had on a community that was struggling to come to terms with available technology vs. the drive to produce a superior recording. Far more positive an effect than anything anyone might say about me on a message board 7 years later. That's all I need, the rest is up to history to judge.... albeit poorly.

Thanks.

-DH

"An armchair quarterback has rarely actually completed a pass himself."

The problem here is that you're using "8-bit" to mean "bad", which is fine in colloquial terms, but really misleading in a technical document where you are talking about bit depth. Particularly when you're using it to refer to an encoding technique other then PCM (such as RA).

To put this another way, would you be comfortable with calling DSD "24-bit"? I wouldn't be, and I would certainly never write that it is.



I really doubt it plays back 8 significant bits even today. 8 significant bits would be very close to lossless bitrates. RA is not a lossless format.



This is doesn't make much sense. The number of significant bits you get out of a codec has nothing to do with its dynamic range. Even those "8-bit" (as you call them) RA files from the 90s had at least 16 bits worth of dynamic range, and perhaps more (I haven't looked at the Cook codec in any detail but its a transform codec so I would expect it to be capable of very large dynamic range).

While it does not call into question either the technical aspects or the possible perceptual benefits of mixing and mastering at greater bit depths, the paper, published late last year, on listening to greater bit depths and sample rate recordings vs listening to the same recordings resampled to 16/44.2kHz casts extreme doubt on any claim that anyone can hear a difference. At the very least it should make any rational person question any listening experience that isn’t blind.

As far as I know, the paper is not available on-line except to subscribers, but extracts have been quoted in many discussions. This local thread also contains links to responses by one of the paper’s authors to some of the “audio is my religion and one can’t question faith” crowd.
http://www.hydrogenaudio.org/forums/index....c=57406&hl=

This article on the article is available and quite easy to read.
http://mixonline.com/recording/mixing/audi...s_new_sampling/

The question of sampling rate and bit depth are a bit unclear to me.

Sampling rate is the easiest to understand, as long as you can’t prove Nyquist wrong, a sampling rate double our hearing threshold is good enough. In real life there is a thing called technology so we have to deal with a couple of problems which don’t exist in the pure mathematical world. As Dunn phrased it quit nicely:

A direct effect of the higher sampling rate is that for an identical filter design the time
displacements will scale inversely with sample rate. Hence an improvement can be
made just from raising the sample rate - even for those who cannot hear above
20kHz.

As far as I knew, there is no theorem proving that a minimum as X bits is needed to reproduce all the details in the signal. We have no Nyquist for the bit depth.

From another post in this forum I learned that flipping even the LSB in 16 bit PCM is clearly audible. I’m inclined to think that a resolution well above the hearing threshold is a bit to ‘rough’
Listening to recordings at a higher resolution (24 bits) should bring an improvement in detail.
Can anybody explains why this doesn’t seems to work in practice?

a) in practice, the average person (and sadly enough, many audio engineers) these days has far more hearing damage than they did 10 years ago, due to noise pollution, poorly mixed/mastered recordings, intentionally distorted musical content, excessive compression, improper EQ (can you say "Smiley face"), and high volume volume listening.
b) In practice, the average person has not developed any ability to appreciate what's missing from lower-resolution audio. They simply don't know what to listen for, and haven't developed an appreciation for what larger wordlengths combined with greater dynamic range bring to the table.
c) The best way to do any kind of test is to use a live source as a control in any experiment. Because of the nature of the potential benefits of greater dynamic range, perceptions of sound reproduction methodologies should not be compared to themselves, but rather to the live source. Most "listening tests," blind or otherwise, are not done this way, and the listener from the start has no reference from which to evaluate the quality representation of low level signals, reverb tails, room reflections, etc. They don't hear the real tails and reflections from the beginning, therefore they have little to compare it to.
d) The experiment part of the test should involve raw, unmastered, unprocessed recordings (with the exception of dither/noise shaping that might be used in the creation of the 16-bit recording).

To make a *loose* analogy, it's like comparing 2 glasses to see which one is more full, without having a sense as to how tall the glasses are. You can compare the amount of liquid in the bottom of the glasses with respect to the bottom of the glasses... but how full the glasses are, and whether the differences between the contents of the glasses are significant, ultimately depend on how tall the glass is. If you don't have the live reference, you cannot as easily perceive the differences (and therefore value) between the live source and the test reproductions.

These are the things greater dynamic range bring to the table. The loss of detail due to poor representation afforded by shorter word lengths occurs in the smoothness of the dynamic changes (waveform amplitude changes) in lower level signals/components of the recording. These are concepts that only a trained ear can discern. It's not hard to train yourself to hear these "features" of a signal, provided that your hearing is not damaged in the manner many peoples' ears are these days. I can instantaneously hear, on my system, the difference between a raw 16-bit recording, and a 24 bit recording recorded at the same levels. It is night and day. Note the word of the use *recording*, and not mixing/mastering/final product.
d) As has been said many times, the greatest value and first bottleneck comes from the initial recording's wordlength and dynamic range afforded by analog circuitry.

When you start with more significant bits, you have the ability to end up with more significant bits in the final product. Merely playing back a PCM stream that uses 24-bit words doesn't imply that the extra bits are significant by itself. Many listening tests involve a normalized 24 bit recording and a normalized 16 bit recording that has been dithered and noise shaped from the same 24 bit source. There is far less audible difference between those two sources, assuming a good dither/noise shaping algorithm is used. However, that test is not the test that validates the value of recording, mixing, and mastering in the 24 bit domain.

In the case of listening back to a raw recording where signal levels were not maximized to take advantage of all of the available dynamic range (i.e. not peaking near 0dBFS) at the time of the actual recording, a 24 bit recording will blow away it's 16bit counterpart in perceived quality for anyone who isn't too deaf to hear reverb tails, overtones and harmonics, and high frequency sounds that usually are not utilizing the full dynamic range available to begin with.

Finally, you used the phrase "above the hearing threshold." The value of 24-bit recording comes into play towards the bottom of the hearing threshold.

Anyone with a *trained ear* who can't hear what raw 24-bit recordings are capable of doesn't either have the proper source material (i.e. source material with at least 19 significant bits of true signal), or they don't have a capable reproduction system.

There's nothing religious about the argument whatsoever, and any pro audio engineer worth their weight in ears will tell you how audibly valuable the 24-bit domain can be. Where people get lost in evaluating these "listening tests" is in understanding what their expectations should be given the test subjects' listening experience, hearing, the quality of the source material, the production methodologies employed, and the reproduction system and environment. People then end up thinking that one test's results regarding a *mastered* recording in some magazine is capable of proving something about all aspects of the 24-bit domain..... The fact that a fully mastered 16-bit recording from a 24bit source can be made to sound as good as its 24bit counterpart is a testament to the fact that there *is* an audible difference between 16-bit and 24-bit *raw source* material.

Folks that question what many audiophiles and pro audio engineers claim to be able to hear generally never make any effort to train their ear to be able to hear them. They'd prefer to jump on the popular bandwagon of folks ready to patently dismiss what isn't blatantly obvious to them at first. In my opinion, the inability to hear the benefits are attributed to either lazyness, deafness, psychological barriers, or lack of proper reproduction environment.

My background and basis for being able to hear what I'm talking about comes purely from listening to raw and/or lightly mastered 24-bit recordings, on good monitors, with a good D/A, in a good room, with little to no EQ, and absolutely no compression. Compare that to the same source's 16-bit counterpart, and it's no contest. Many audiophiles I know don't want to hear any processing at all. They want to hear the pure interaction between mic diaphragm, analog components, cables, and drivers. They want to be able to perceive subtle colorations of components as they add or subtract from the reproduction experience. People who blindly bash audiophiles and audiophile jargon generally don't understand what it is audiophiles are interested in hearing in their music.

-DH

Edit: I started typing this before ccryder's reply, then went away and finished typing it, so I'm merely answering the questions posed by Roseval as thoroughly as I can, not attempting at this time to debunk anyone's laims of roughness (a.k.a. 'digititis') which are indicative more of insufficient dither or poor analogue-to-digital conversion than of deficiencies in 16-bit audio as a playback medium. ccryder actually talks of crude 16-bit recording (possibly on ADCs that don't respect Nyquist and dither properly), not dithered playback



In fact Nyquist alone isn't good enough for sampling rate, because you also need to know the frequency limit of the human auditory system at a reasonable loudness (or alternatively the point at which a low-pass filter becomes inaudible in music, if it's specifically music playback you're interested in).

For bit-depth you also need some knowledge of the human auditory system.

If 0dB SPL is roughly the threshold of hearing at 1kHz and 120 dB SPL is the pain threshold, measurements such as the Fletcher-Munson curves indicate that there's about 120 dB of range from threshold of hearing to pain. Bear in mind that large chainsaws (18" Makita for example) are labelled at typically 113 or 116 dBa, which is very much in this ballpark, and you wouldn't want to listen to those at full-throttle without ear defenders.

Given the time/frequency resolution of the human auditory system, the perceived noise floor of CDDA with flat dither is about -120dB FS per frequency bin at the most sensitive frequencies. (See this old post with graphs and samples)

The maximum sine wave representable is 0dB FS, so there's something close to 120 dB of usable dynamic range.

With noise-shaped dither, the floor is about -135 dB FS at the ear's most sensitive frequencies. The post I linked to above indicates that noise-like signals have a dynamic range of around 112 dB between full scale and where their modulation becomes lost in the shaped noise floor (tested using pink noise).

So, a properly calibrated playback loudness for CDDA or other 16/44.1 PCM will enable us to cover the whole range from the threshold of hearing silence to chainsaw-in-your-hand loudness.

LSB inversion could cause tonal distortions and overtones well above -120 dB per frequency bin, which isn't a fair way to denigrate 16-bit audio. Compare properly dithered 15-bit audio to 16-bit audio and we might have a fair comparison. These forums contain some 24-bit source material (Hollister / Bismarck, I think) to dither down to lower resolution so we could try this on real music and attempt to ABX it. One could use flat dither first, then noise-shaped dither, which would probably allow us to use 2 or 3 bits fewer before it becomes detectable at the full listening volume.

Hello Dan,

I doubt many of those actually interested in audio would argue against recording at 24 bit. But for playback, few consumer/prosumer systems can reproduce 16 bit faithfully, let alone 24 bit.

Also take into account the performance of a DAC operating at 16 bit vs 24 bit. No doubt there can be a sonic difference, but is that difference equal to closer transparency or just coloring of the audio similar to what different speakers do? How would one know for certainty? Is it not possible that those in the recording/mixing profession developed a preference for the way DACs color the output when operating at 24 bits similar to the way they have a preference for certain speakers?

(The above is not meant to sound argumentative, just inquisitive.)

Aside from the fact that it is almost always easy to tell live from a recording, no matter how good the recording, comparison of live vs recording of any particular bit depth is not only unnecessary, it is completely irrelevant to the question of whether there is any audible difference between a 24 bit recording and the same properly converted to 16 bits.

The question is simply can anyone actually tell the difference, merely by listening -- when they have no information except what they hear to tell them which is which. We are not talking about difference in equipment or listening environment, these have to be identical while listening to both versions. We can certainly consider the question of how good the equipment and environment need to be to allow a difference to be perceived, if such perception is possible, but that is not the issue itself.

The claim that only raw 24 bit recordings are clearly different is an interesting one. Just why a difference, audible when recorded at 24 bits, would disappear once the tracks are mixed, etc. (still at 24 bit) is unclear, unless you are talking about the mistreatment most pop and rock music is subjected to on it way from recording to distribution. Quite a few people have pointed out that SACD and DVD-A releases sometimes receive mastering that makes them better, and clearly distinguishable from, the CD version of the same recording. However, as the tests in the paper demonstrated, those differences are not lost or altered when that SACD/DVD-A recording is converted to CD specs.

For five years or so, whenever this subject comes up, in this and a number of other audio oriented forums, I have issued the challenge for anyone championing >16 bits and/or >44.1kHz, to provide a sample: something I can convert or resample (properly) and then identify in a correctly done blind ABX test. Only a real music recording is applicable; it isn’t as hard to make up a sample with test tones.

I don’t say that such recordings do not exist, only that I, and at least most of the other people on the planet, have yet to hear one. So far, every time I’ve made this request, the people arguing in favor of greater bit depth and/or sample rate have just gone away. (There is one thread, in this forum, where a poster claimed to be able to ABX just about any change whatsoever to some 24/96 samples, but I don’t believe anyone else heard what he claimed to hear. (There were also some questions raised about his equipment) The thread got so convoluted that I stopped following it, but I don’t think he convinced anyone.) (Also, that was just about sample rates, not bit depths).

Now I acknowledge that my inability to hear a difference, should a test file become available, does not mean no audible difference exists. The sample has to be available for wide ABX testing, to find out if everyone is as deaf as me. Many people who visit this forum are experienced in doing ABX testing.

Three to ten seconds should be adequate, I see no need for an entire track, but will consider any reasoned argument as to why a greater duration might be necessary to reveal a difference. My preference is for 24/44.1 since even a few seconds of audio is tedious to download over my dial-up line, but a restriction to CD spec sampling rate isn’t a requirement. I, and probably most people, can handle 48kHz, 88.2kHz and 96kHz.

Interesting. Do you have any evidence for this? I would have thought that improved standards on noise in the workplace would have decreased hearing loss among the general population. It is well understood that loud sounds can damage hearing, but I am not aware of any evidence that excessive compression (for example) can have the same effect.

It would be great if you could present your evidence.

No. It is well understood that, for properly dithered quantization the quantization error is not correlated with the original signal (see Oppenheim and Schafer, "Discrete Time Signal Processing" or for a good modern treatment, or W. R. Bennet, "Spectra of Quantized Signals", Bell Systems Technical Journal, vol. 27, 1948 for the foundations of the theory). Stating this mathematically, we define the error samples e[n]:

e[n] = Q(x[n]) - x[n]

where Q() is the quantization process and x[n] is the samples of the signal under test. When dither is used (or for self dithering signals, where the analog SNR is below is the quantization SNR) the signal e[n] is not correlated with the signals x[n] and Q(x[n]). So what does this mean in practice? It means that the properly dithered quantization process is an additive noise process - it is equivalent to adding independent noise with a particular spectrum to the original signal.

Another thing is this belief that more quantization causes "roughness" in the output signal. As you are well aware, the reconstruction process does not produced a stepped output - it produces a bandlimited (and hence fairly smooth) analogue output. The output of a good sampling/reconstruction process is not "rough" or "stepped" in any way - if anything it will be less "rough" because of the strict bandlimiting imposed on the process.

This doesn't mean that recording, mixing, mastering an processing at high bit depths are without merit. There are plenty of good technical reasons to argue for this, though, so using audiophile terms like "roughness" and resorting to incorrect interpretations of the process are not necessary.

Of course. Mastering, mixing and recording in high bit depths (24 or more) is a good idea.

Two points to make on this.
#1, if you don't know what you're supposed to be able to hear in that recording, i.e. if you don't have a live source (or preferably a live analog source without the A/D stage) to compare to the two recordings (consider what I'm talking about to be not an ABX comparison, but rather an ABX-C comparison, where C is the live analog source), then it changes your perception of what kinds of differences to expect from the outset. It's been my experience that different kinds of music, with different harmonic and reverberant content will have an effect on how drastic the differences between 16 and 24 bit will end up being. I start with the premise that the most valid test should be to determine which one produces a better sounding facsimile of the original. I maintain that without a reference to a live source... without a glimpse of what should be possible to hear in that passage, your brain may have nothing specific to look for when you're then tasked to look for differences between A&B. However, (for instance) if you know how much reverberant content the original source contained, I propose that would educate your ear for that particular test, and would change outcome of such a test.

#2, take 2 recordings, both recorded conservatively so as not to risk ever clipping or hitting 0dBFS for more than 1 sample. Say, -24dBFS is the highest peak on a simultaneouly recorded 24bit and 16bit recording. I don't care how much dither and noise shaping you add, you're not going to end up with more than 12 significant bits on a 16 bit recording, but with good quality analog components & A/D, you will still end up with 16 significant bits of audio on the 24bit version. The difference in listening back is effectively the difference between a 12bit recording and a 16 bit recording (noise floors notwithstanding). Can you hear that difference? Do you understand why you might hear that difference? Now, take an instance where there's one transient spike +12dB higher than the rest of the recorded music, and you happen to know exactly when (in advance) that spike will occur. So knowing that, and not wanting to ride the record levels to as to manually compress dynamic range, you leave the levels set in such a way that the majority of the peaks are at -12dBFS, and you allow room for that one spike to take it a fraction under 0dBFS. The majority of that recording in the 16 bit realm is still getting a "14bit treatment", and even then only at its loudest point that it gets that amount of resoution . And what about that 24-bit version? Still gets a full 16 bits or more of significant dynamic range. Think you can hear that difference? I'll make the bet again and again that any decent audio engineer can tell the difference between a 16bit recording and a 12-14 bit recording. And therein lies the reason a live, unmastered, unadulterated 24bit recording will, more often than not, be obviously better sounding than its 16bit counterpart.

I should say after reading recent posts (and even before reading them), that I agree with about 95% of the sentiment that higher sampling rates are largely indiscernible. I think it is possible that there are special situations when it might be possible to reliably discern the difference between 24/44.1 and 24/96.... but I would probably limit such situations to where a single acoustic instrument containing high frequency harmonic was being captured with *near coincident* stereo mic patterns like ORTF. It would be that type of situation that I would seriously doubt ever entered into the test equation performed by Moran and Co. Asked if I bother with 96kHz, I would tell you that only if I cared very deeply about the recording for my own personal music, or was being paid specifically for a 24/96kHZ, would I use 96kHz to record.
The rest gets the 48kHz treatment. Generally speaking, for mastered, fully produced music, I agree that 96kHz for playback is pretty much a waste of space and CPU power. Not having read Moran & Co.'s paper, but from the articles mentioned above, it would appear to me that the general focus of the test was to prove that higher sampling rates are a waste, and not greater wordlengths. Like I said, most of the time, and for the overwhelming majority of both listeners and listening material, I agree with their findings.

However, wordlength is an entirely different matter.
Bottom line: I believe unmastered 24-bit audio compared with unmastered 16-bit audio from the same source has the potential to indicate vast audible differences in the timbre and decay of both harmonic and fundamental content (as well as reverberant content) over time. Most people don't prefer to listen to unmastered, yet well captured recordings. However, there are some that do, like me, and given what I've mentioned above, for those people, it's night and day. However, crush the crap out of the 24bit recording, and master it to 16 bits with noise shaping and dither, and then compare the two, and you probably won't hear much of a difference.

Back when I did a lot of concert recordings (in those cases, unpaid jobs), there were many times I, being the only one using 24-bit technology at those venues, created a decent sounding recording, while everyone with 16bit technology ended up with noisy rubbish, solely because the majority of the content was far below -0dBFS.

There are a whole host of live, unmastered 24bit recordings on archive.org (of varying qualities based upon mic placement from the source, mic patterns, analog stages, and A/D stages).... some of which I personally recorded, and have been available publicly for the past 6 years or more. Anyone who wanted to do a test for themselves to hear the difference (provided they had playback capability with enough dynamic range) can pull down one of those recordings and do their own experiments. I suspect that most who say they don't hear a difference never tried to listen to this kind of source material.

So for anyone who prefers this kind of listening to the nightmare creations of audio engineers designed specifically to sell more records, there is no argument.
Anyone's perception of the value of 24bit recording *and listening* is simply a matter of perspective and expectations.

I recognize that the majority here doesn't listen that way, or to that kind of music, but I just thought perhaps a glimpse of thinking outside the box might put the argument of 16bit vs 24bit into better perspective, for both recording, and listening. Minimally, any argument in that vain needs to be first qualified by the type of source material and listening attitude of the participants.

-DH

That explanation would seem to imply to me that there is diminishing perceptibile difference in the slopes of different fade-in and fade-out curves for lower level signals. It also implies that the reconstruction of low amplitude waveforms can create "something" out of virtually nothing, and that "something," when normalized using pure bit shifting, is just as accurate as the reconstruction of high amplitude waveforms, with the same slope error. I beg to differ on at least the latter account. It is my experience that lower level harmonic and reverberant content "drops off the map" faster with 16 bit quantization, and that the timbre of such content changes over time as it decays in a different manner than a 24bit recording.

As for my comments about increasing deafness and loss of hearing, my perception of that situation may be limited to personal experience in increases in noise exposure in the US city I live near. Loud subways and trains, more people using mass transit, squealing breaks, loud engines, extended cell phone use, increased noise attributable to increases in population density, etc. Then there's the simple notion that more people are doing extended listening of highly compressed audio, and audio pros well understand the concept of ear fatigue which is a function of SPL's over time. It might be purely an American thing, I don't know. But either you believe that average deafness is decreasing, average deafness has not changed, or average deafness is increasing (without regard to deafness awareness having been increased recently). I have no reason to believe that in the area where I live at least, it is anything other than the latter.
Though I've heard this sentiment often enough from the media, I'll continue to look for specific research to support my assertion. But to me in the meantime, it seems to follow purely logically that as living environments become louder (workplace environmental measures notwithstanding, the value of which I believe to be diminished by the fact that limits for loudness exposure as decreed by the US government are not quite as adequate for long term hearing protection as they would have you believe), the rate of hearing loss doesn't stand much of a chance of abating.

-DH

No offense, but this only goes to show that there really is a lack of understanding on your side. Dithering turns the quantizer into a friendly noise source -- just like cabbagerat said. I'm not counting anymore how many times this topic has come up here. Dithering seems to be one of the most misunderstood things...

"smoothing out artefacts" is not the way I'd put it no matter what kind of audience I'm dealing with.

Edit: Just to be clear: I think nobody here is arguing against recording in 24 bit. There might be problems with certain ADCs recording at 16 bit etc. But I'm not talking about possible "implementation issues". I'm talking about what you can expect from 24 bit -> 16 bit conversion.

Cheers,
SG

I'm not quite clear whether you're talking about 24 bit recording, or a playback medium that's 24 bits compared to CD 16.

Everyone here with an opinion seems to agree that for RECORDING, 24 bits is the way to go, but the consensus is that no-one has demonstrated the difference between 24 and 16 for PLAYBACK. In your example, @ccyrder, I assume, in my ignorance, that it would not be hard to produce a CD with the 16-bit depth the 24-bit recording produced.

I really enjoy watching people hit each other over the head with science I scarcely understand, and mathematics which is, alas, incomprehensible to me. But it's more fun if everyone's in the same cage.

I never said a word about dither in my last post.
Don't know why you insist on bringing that up.
There was discussion of reconstruction of the waveform in the D/A stage, and a suggestion that smoothing of slopes as part of the D/A stage creates an equal result with specific respect to perceived timbre and decay of signal content at all recorded signal levels, but no mention of dither.

Furthermore I don't see how one can talk about dither like that, when it raises the spread spectrum noise floor, and without noise shaping, doesn't have the same effect of increasing perceived resolution as it would with noise shaping. And once you get into noise shaping, you get into the argument about what specific effect different noise shaping algorithms have on different sources. Can of worms, I ain't going there, other than to say that different noise shaping algorithms obviously impart different colorations with are audible. Which again comes back to supporting my assertion that there is easily a potential difference in sonic character of raw 24bit recordings vs. raw 16bit recordings, especially given the need for noise shaping on the latter.

Either you captured accurate low level content, or you didn't capture accurate low level content because the signal information didn't fit into the word, and you try to approximate the missing LSB data by adding dither and noise shaping at quantization time (yes, I know dither in the 16bit domain causing algorithmic toggling of the LSB's is based upon low level content beyond the dynamic range afforded by 16-bit) . But no matter what, low level signals dithered (with or without noise shaping) to 16 bit during recording and then normalizing through bit shifting will not in my opinion result in the same analog waveform with the same qualities of decay and timbre as those same low level signals recorded at the same level but quantized with 20+ bits (dithered or not), normalized non-destructively through bit shifting. This is especially true where the peak levels of the original recording are far away from 0dBFS.

Additionally, I would imagine (guessing here) the effectiveness of such reconstruction would be contingent upon or impeded by the quality of an individual's D/A.

The argument is really academic, since the perceptive effect of dither and noise shaping for the listener is in some way related to the overall original recorded signal levels with respect to available dynamic range.

Honestly, I have nothing further to add. I perceive distinct audible differences in the character of low level decay, timbre, and harmonic content between most raw 16bit recordings vs 24bit raw recordings I concurrently make (from the same A/D converter with dual outputs), with or without dither and noise shaping. The only time the differences aren't drastic to me is when average recorded peak RMS levels are above about -18dBFS, or when the signal has undergone several processing steps generating mantissas. Been listening to, recording, mixing, and mastering 24bit audio recordings extensively for 9+ years now, I know what I hear, and I know under what conditions I'm able to hear it. That's it. If you don't like the way I describe it, feel free to disagree.

-DH

Last post on this. For anyone to say that there is consensus that no-one has demonstrated the difference between 24 and 16 for playback is to patently discount the experiences of folk who would prefer to listen to their recordings in their original state--Raw and unprocessed.

The need for people to argue the point about there being no perceivable difference indicates to me that those very same people probably have never made a high-fidelity 24bit recording themselves.

Beyond that, yes, a 16bit CD mastered from a mastered 24bit source can be made to sound indiscernable from the mastered 24bit version. Without mastering, the difference is night and day most of the time for the trained ear.

-DH

Expectation and belief can, and have repeatedly been demonstrated to do more than just color perception. They can produce perception of something that actually does not exist in the sample and block perception of something that does exist, leaving the subject fully confident of his/her correct discrimination of the completely wrong thing. The point of blind testing is to remove that bias.

There isn’t any need to have another reference, live or any other state, to make the determination of whether X is A or X is B. If you can not tell just by listening to them, you have demonstrated that you cannot tell any difference between A and B. There are no other constraints during testing that can lead you to a false choice or prevent you from hearing something that is there -- if it really is there. It does not matter what your belief about it is, nor what all you former experience may have led you to expect. This is not to say that experience with listening to different things might not improve discrimination, but the test is about what one can hear right now.

Many people, probably many tens of millions by now, have found that audio they once found distinctly different (e.g a PCM recording and an mp3 version thereof [in some cases]) somehow lost all audible differences once they accepted the challenge to listen to them in an ABX test.

Because that is what would prevent introducing any of the alterations you mentioned except for noise. Of course it totally makes sense to record low level signals with 24 bit simply because of the lower noise floors that are achievable with 24 bit. Anyhow, there's no reason for why "low amplitude harmonics" should "drop off the map" when converted to 16 bit even without prior shifting of the peaks closer 0dBFS because a proper conversion is only expected to add noise. Don't get me wrong: I'm not arguing pro or against 16 bits. I'm just taking issue with the things you say about quantization, dithering and their effects. If you experience any other sorts of artefacts besides noise you're either doing something wrong or imagining things. As long as the noise floor is low enough (ie below the threshold of hearing) you're fine. You don't need to expect any other kinds of artefacts. To reiterate what David said: The parameters sampling frequency and bit depth just have an effect on the achievable bandwidth and noise floor, nothing else.


Of course it's a bad idea. Nodoby here is advocating these kinds of processes. I'd simply record in 24 bit to be on the safe side. I think we established this already.

Cheers,
SG

For me it's more like 'I could believe Woodinville and I could believe guys like dpierce. But I long ago stopped believing those recording/mastering engineers who base everything on sighted comparisons and "I hear it and I'm a pro therefore it's real' belief systems, even when their anecdotal experience flatly contradicts known physics, psychoacoustics and engineering. I've seen too much utter nonsense come from their pens.

Btw, DH cites Bob Katz, who, if you read his book 'Mastering Audio', actually highly (and correctly) qualifies
his beliefs about higher-than-redbook sample rates and bit depths. He differs from many pontifical MEs in that he has participated in, and endorses, blind tests and is interested in the TRUE CAUSES of differences. For SR the 'benefit' comes from filter implementation (rather than any contra-Nyquist arguments, so popular in audio-land), and for bit depth it comes down to avoiding audible rounding error during production (or, in live situations, allowing more headroom for unexpected peaks). In other words, 16/44.1 CAN be perfectly adequate for a recording and playback under proper conditions.

A difference in low-level resolution of native DSD vs trancoded-to-redbook was indeed noted by Meyer and Moran in their test, where it could only be made audible during soft passages, and only by jacking up playback to levels that were ear-splitting during normal passages. Which is pretty much what one would predict. The difference would NOT be expected to be audible at normal listening levels, and wasn't.