Jump to content
IGNORED

Does High Resolution Audio sound better


Recommended Posts

Ricardo, I am interested in where you usually experience live music, because your impression of the sound of recordings (Barry's for example) and equipment (Vandersteen speakers) is so often that they tend to sound more shrill or detailed than live; whereas I could wish for the vocals to be just a bit more forward in Barry's latest recording, and in more than two decades of owning Vandersteens I have never experienced them as shrill or overly detailed.

 

(My wish for a more forward vocal in Winds of Change is likely from hearing most other recordings of vocals in my collection close-miked. Also, I do get to listen to a reasonable amount of live acoustic music.)

One never knows, do one? - Fats Waller

The fairest thing we can experience is the mysterious. It is the fundamental emotion which stands at the cradle of true art and true science. - Einstein

Computer, Audirvana -> optical Ethernet to Fitlet3 -> Fibbr Alpha Optical USB -> iFi NEO iDSD DAC -> Apollon Audio 1ET400A Mini (Purifi based) -> Vandersteen 3A Signature.

Link to comment
While I agree that the late 50s and early 60s did produce very nice classical music recordings, the use of un-EQ'ed mics with peaky treble in a spaced-omni configuration resulted in the infamous "hole in the middle" stereo image and adulterated the sound of vocals and instruments alike, such that of Heifetz's shrill-sounding violin on RCA.

 

On the other hand, mics and electronics of that era produced significant roll-off above 16-18kHz which in my opinion makes recordings sound more "natural" and "realistic" than the modern ones using flat-FR mics, though less "airy" and "detailed".

 

Well as a violinist myself and one who has had the privilege of playing a genuine Joseph Guarneri, what can I say except that when fitted with the particular strings Heifetz favoured (steel "E", plain gut "A" and "D" and wound gut "G") and his predilection for a generous amount of rosin, that is actually the sound those instruments made / make in those circumstances. I say made because unfortunately over the years many have had such extensive repairs and alterations that the height of ribs has been reduced. And that amongst a whole host of other factors were what made those instruments unique.

 

If you are hearing a Heifetz violin sound shrill then I honestly feel you need to be looking at the setup and or component selection or other factors in your system. It should have the Guarneri carrying power and the particular tone associated with the maker but definitely not shrill.

 

I also don't really agree abut the old versus new mics. I have heard plenty of very modern recordings such as Reference Recordings that are the most natural and realistic I have ever heard. And by quite a big margin. Same with the Telarc ones despite the fact that many were only 50 Khz 16 bit masters. Certainly little to no use of ancient microphones in those cases. One factor that made the old classical recordings sound more "realistic" and "natural" to use your terms is the strings used by the players. Synthetic strings came onto the market in the late analogue era (mid 70s) and prior to that the players had little technological latitude and favoured strings such as Eudoxa, Olive, Gold and Kaplan. These made quite a different sound to the modern synthetics and a good straight "audiophile" analogy would be comparing vinyl with valve amplification (the old strings) versus high resolution downloads played on high powered transistorised modern amplification. It is pretty easy to tell the difference between a string section shod with modern synthetics versus the types of strings that without exception would have been used in those classic recordings. I never liked the modern synthetics myself and tended to favour the lower tensioned (very slightly thinner) Olives.

Link to comment

Modern systems with latest hi-res capability able almost non-distorted playback in full audible range.

 

But traditional stereo or multichannel speaker's drivers can't produce "natural" sound - sound of concert hall.

Even was video with stereo track. There was compared some qualitative system vs. live band (fully identical music).

 

Me seems, now here almost achieved threshold.

May be need work more under driver's distortions. But, I think, it don't solve the matter.

 

Now need build audio-holographic system. It complex of recording and playback apparatus.

 

First need omnidirectional speakers.

 

Any audio format (PCM or DSD) and resolution we must consider only in complex with apparatus.

AuI ConverteR 48x44 - HD audio converter/optimizer for DAC of high resolution files

ISO, DSF, DFF (1-bit/D64/128/256/512/1024), wav, flac, aiff, alac,  safe CD ripper to PCM/DSF,

Seamless Album Conversion, AIFF, WAV, FLAC, DSF metadata editor, Mac & Windows
Offline conversion save energy and nature

Link to comment
Well as a violinist myself and one who has had the privilege of playing a genuine Joseph Guarneri, what can I say except that when fitted with the particular strings Heifetz favoured (steel "E", plain gut "A" and "D" and wound gut "G") and his predilection for a generous amount of rosin, that is actually the sound those instruments made / make in those circumstances. I say made because unfortunately over the years many have had such extensive repairs and alterations that the height of ribs has been reduced. And that amongst a whole host of other factors were what made those instruments unique.

 

If you are hearing a Heifetz violin sound shrill then I honestly feel you need to be looking at the setup and or component selection or other factors in your system. It should have the Guarneri carrying power and the particular tone associated with the maker but definitely not shrill.

 

As far as I know, Heifetz was recorded closer than what was then standard; this, and the mics with peaky treble are responsible for the brighter tone.

This is well documented.

 

I also don't really agree abut the old versus new mics. I have heard plenty of very modern recordings such as Reference Recordings that are the most natural and realistic I have ever heard. And by quite a big margin. Same with the Telarc ones despite the fact that many were only 50 Khz 16 bit masters. Certainly little to no use of ancient microphones in those cases. One factor that made the old classical recordings sound more "realistic" and "natural" to use your terms is the strings used by the players. Synthetic strings came onto the market in the late analogue era (mid 70s) and prior to that the players had little technological latitude and favoured strings such as Eudoxa, Olive, Gold and Kaplan. These made quite a different sound to the modern synthetics and a good straight "audiophile" analogy would be comparing vinyl with valve amplification (the old strings) versus high resolution downloads played on high powered transistorised modern amplification. It is pretty easy to tell the difference between a string section shod with modern synthetics versus the types of strings that without exception would have been used in those classic recordings. I never liked the modern synthetics myself and tended to favour the lower tensioned (very slightly thinner) Olives.

 

Thanks for the info on violin strings.

 

I only have a single Reference Recordings CD (Stravinsky) and I don't find the sound and perspective of it's "The Rite of Spring" as "natural" or "realistic" as the Dorian (Mata/Dallas) recording of the same piece, perhaps because the Dorian has more hall sound and the RR a bit too much detail to sound real (they were probably spot-mic'ing).

"Science draws the wave, poetry fills it with water" Teixeira de Pascoaes

 

HQPlayer Desktop / Mac mini → Intona 7054 → RME ADI-2 DAC FS (DSD256)

Link to comment
Modern systems with latest hi-res capability able almost non-distorted playback in full audible range.

 

But traditional stereo or multichannel speaker's drivers can't produce "natural" sound - sound of concert hall.

Even was video with stereo track. There was compared some qualitative system vs. live band (fully identical music).

 

Me seems, now here almost achieved threshold.

May be need work more under driver's distortions. But, I think, it don't solve the matter.

 

Now need build audio-holographic system. It complex of recording and playback apparatus.

 

First need omnidirectional speakers.

 

I would much rather have the "right" tonal balance and dynamics (which I guess is so much easier to achieve) than an audio-holographic image.

 

As Wagner once said to his friend Nietzsche at the Festival Theatre in Bayreuth: "remove your spectacles, music is only to be listened to"

"Science draws the wave, poetry fills it with water" Teixeira de Pascoaes

 

HQPlayer Desktop / Mac mini → Intona 7054 → RME ADI-2 DAC FS (DSD256)

Link to comment
Ricardo, I am interested in where you usually experience live music, because your impression of the sound of recordings (Barry's for example) and equipment (Vandersteen speakers) is so often that they tend to sound more shrill or detailed than live; whereas I could wish for the vocals to be just a bit more forward in Barry's latest recording, and in more than two decades of owning Vandersteens I have never experienced them as shrill or overly detailed.

 

(My wish for a more forward vocal in Winds of Change is likely from hearing most other recordings of vocals in my collection close-miked. Also, I do get to listen to a reasonable amount of live acoustic music.)

 

Hi Jud,

 

I moved abroad last summer and was finally able to attend a couple of concerts this month: this week I listened to an organ recital at one of the Colleges' chapels and before that I was sitting about 8 meters away from Maria João Pires' Piano at the Sheldonian.

 

As for the Vandersteens, I have just checked the measurements at Stereophile and the Signatures I listened to do have a "hot" tweeter and an uptilting frequency response.

"Science draws the wave, poetry fills it with water" Teixeira de Pascoaes

 

HQPlayer Desktop / Mac mini → Intona 7054 → RME ADI-2 DAC FS (DSD256)

Link to comment
2. They continue the all too widespread confusion of signal-to-noise ratio with dynamic range. While these are *theroetically* the same -- or should be -- in reality they are two very different things. That is to say that low level signals might be well above CD's noise floor but they get increasingly raggedy as the level goes down and the signals are effectively represented by fewer than 16-bits of resolution.
If it is properly dithered the only difference between the upper bits and the lower bits is noise.

If you use TPDF dither without noise shaping, this noise should sound identical to analog "white" noise. (tape hiss, static on a TV screen etc.)

 

If it is improperly dithered, you will end up with a lot of distortion in the lower bits, rather than noise.

The digital representation of the signal in a waveform editor is not representative of what comes out the other end of your DAC.

 

This paper and other sources, I believe Monty as well, suggest high resolution can be deleterious to sound quality. Isn't this contradictory to those who believe there is no sonic difference between standard and high resolution?
This is why I have tried to be careful not to say that there is no difference, but that there should be no improvement.

It's certainly possible for high frequency content to introduce distortion in the audible ranges, whether that's aliasing, IMD, or simply asking more of the speaker drivers.

 

If you actually look at frequency plots of a lot of high-res content, there are spurious tones and other distortions that could cause problems within the audible range. And it demonstrates to me that the high frequency content is clearly not something we detect through means other than hearing, as some have suggested, because those files with constant high-pitched tones would be uncomfortable to listen to in high-res.

 

I largely tend to believe that 16/44 may be accurate or extremely close to accurate in reconstructing the necessary sound wave. On the other hand, if things that occur at >20kHz frequencies can somehow have effects within the audible range, then it would make sense that those "effects" could be both positive and negative. So I agree with you that it is very difficult to square an argument that says hi-res causes bad things with an argument that says hi-res can only affect frequencies we can't hear anyway.
And since we cannot hear those frequencies, it's far more likely that any differences heard are distortions in the equipment, than being a more accurate reproduction.
Link to comment

 

This is why I have tried to be careful not to say that there is no difference, but that there should be no improvement.

It's certainly possible for high frequency content to introduce distortion in the audible ranges, whether that's aliasing, IMD, or simply asking more of the speaker drivers.

 

If you actually look at frequency plots of a lot of high-res content, there are spurious tones and other distortions that could cause problems within the audible range. And it demonstrates to me that the high frequency content is clearly not something we detect through means other than hearing, as some have suggested, because those files with constant high-pitched tones would be uncomfortable to listen to in high-res.

 

And since we cannot hear those frequencies, it's far more likely that any differences heard are distortions in the equipment, than being a more accurate reproduction.

 

I don't follow your logic. I can understand how a specific recording might have spurious tones and other distortions and thus be bad, but I don't see how the conclusions you draw are entailed by the statements you are making.

Custom room treatments for headphone users.

Link to comment

In re-reading this whole thread, I couldn't help but wonder how many of those defending 16/44 as fully sufficient of reproducing everything we can hear were on the opposite side of that issue in comparing CD's to vinyl (i.e. LP's are inherently more accurate than redbook CDs)?

Synology NAS>i7-6700/32GB/NVIDIA QUADRO P4000 Win10>Qobuz+Tidal>Roon>HQPlayer>DSD512> Fiber Switch>Ultrarendu (NAA)>Holo Audio May KTE DAC> Bryston SP3 pre>Levinson No. 432 amps>Magnepan (MG20.1x2, CCR and MMC2x6)

Link to comment
Hi Jud,

 

As for the Vandersteens, I have just checked the measurements at Stereophile and the Signatures I listened to do have a "hot" tweeter and an uptilting frequency response.

 

Not sure which of the graphs you were looking at, but here is what John Atkinson said about his measurements:

 

The 2Ce's upper-frequency output is flat, but with a number of small peaks and dips apparent. The usual metal-dome tweeter resonance occurs at a high 27kHz, well above audibility.

 

Then, with regard to in-room response, first frequency:

The upper midrange and treble in this graph are very smooth, with the slight downward tilt of the curve reflecting both the increased amount of room absorption at high frequencies and the tweeter's increasing directivity.
(Bold mine, underlining standing in for italics by Atkinson.)

 

Finally, regarding in-room time domain response:

In the time domain, despite the Vandersteen's multiway design, its impulse response (fig.7) is as time-coherent as that of the single-driver, crossoverless Fujitsu Ten Eclipse TD712z, reviewed elsewhere in this issue.

One never knows, do one? - Fats Waller

The fairest thing we can experience is the mysterious. It is the fundamental emotion which stands at the cradle of true art and true science. - Einstein

Computer, Audirvana -> optical Ethernet to Fitlet3 -> Fibbr Alpha Optical USB -> iFi NEO iDSD DAC -> Apollon Audio 1ET400A Mini (Purifi based) -> Vandersteen 3A Signature.

Link to comment
As far as I know, Heifetz was recorded closer than what was then standard; this, and the mics with peaky treble are responsible for the brighter tone.

This is well documented.

 

Yes, I've read it all too, but it was just close miking which gave a forward balance of the instrument relative to everything else. You used the word shrill, now you are using the word bright. To me they are two completely different things. As I've mentioned, I have many years experience with the instrument and have played many expensive (and cheap) instruments and have also heard all manner of players and instruments from all perspectives - two feet away, in the same small room, in a concert hall, etc.

 

The claims about the Heifetz sound have not been made by self-professed audiophile violin players that I am aware of. They've been made by players who are by far the worst perpetrators of the close miking syndrome themselves such as Itzhak Perlman or audiophiles who don't have 30 years experience with the instruments and haven't heard enough Del Gesu instruments up close and distant (if any at all) to be even remotely qualified to even comment, let alone judge.

 

I can tell you right now and you can chose to believe it (or most likely not). A genuine Del Gesu up close is going to have dollops of volume and a very penetrating, vibrant, powerful tone. That is precisely what they are renouned for. It will have tremendous texture, clarity, definition and depth, but shrillness is something you get from a cheap Chinese violin, not a classic (or even modern) Italian one.

 

Listen to Pinkus Zukerman close miked giving a demonstration and his violin (also a Del Gesu) and it gives precisely the same type of defined, glossy, even edgy if you like sound. But shrill to me is an offensive sound as I would have thought it is to most people. And if Heifetz sounds shrill then it is an equipment problem, not a problem with the recording. Also bear in mind that yes, the older mikes might have had a presence ranges but you get the same effects by moving from a soft acoustic environment to a harder one as well. And you also get the same effect when you have a largely empty concert hall for recording purposes versus a hall full of thousands of people on a concert night. Furthermore, the frequency response of the older recording machines wasn't flat to begin with, nor were the methods used to derive the tapes used as cutting masters nor the resultant records themselves. Nevertheless, I have all the Heifetz SACD releases from 10 years ago and acquired two of the Analogue Products LPs just last week (Bruch Scottish Fantasy and the Sibelius Concerto). Honestly, shrill is the last word I'd use to describe any of these and if anything at all, I don't really find the perspective to be any different to anything much else out there. Yes, I would have said "shrill" on my poor quality equipment from 15 to 20 years ago, but nothing remotely of the sort these days.

Link to comment
And you also get the same effect when you have a largely empty concert hall for recording purposes versus a hall full of thousands of people on a concert night.

 

Strikes me I haven't thought nearly enough - at all, really - about this.

One never knows, do one? - Fats Waller

The fairest thing we can experience is the mysterious. It is the fundamental emotion which stands at the cradle of true art and true science. - Einstein

Computer, Audirvana -> optical Ethernet to Fitlet3 -> Fibbr Alpha Optical USB -> iFi NEO iDSD DAC -> Apollon Audio 1ET400A Mini (Purifi based) -> Vandersteen 3A Signature.

Link to comment
By "look" too close I mean that I wouldn't probably be listening to the musicians from that position but as you have mentioned mics don't capture sound the way we hear it so I understand that some compromise is mandatory.

 

Hi Ricardo,

 

I wouldn't call it a compromise since nothing is being given up. It is how things work, given an understanding of how a given microphone in a given setup is going to "hear".

 

I did mention fake "detail" but not fake "space" because the high frequencies don't sound "natural" from what I remember hearing live; I have tried your samples with different speakers and headphones and that impression is common in all of them.

 

Keep in mind that you were not present at any of the recording sessions. It is also possible you just prefer (or are used to?) recordings that don't have the same extended response. The mics I use are quite flat in both amplitude and phase and they are of much wider bandwidth than most other mics.

 

I would like to try recording music sometime: what (very) affordable (preferably used) mics and pre-amps do you suggest?

 

Of course, as with anything in audio, if you ask three folks, you will get at least four different answers. ;-}

 

The least expensive microphones I would recommend for achieving excellence in a recording start to approach $1000 for a pair (Earthworks TC20). Same with the preamp/converters I would recommend (Metric Halo ULN-2 or 2882). So for something costing less, my best suggestion would be to look for small diaphragm omnidirectional condenser microphones.

 

By the way, do you perform any EQ'ing to your recordings?

 

Cheers,

Ricardo

 

Soundkeeper Recordings do not use EQ or any other processing. I find the sound from the mic array, mic cables, mic preamp and A-D converters I'm using is faithful to what I hear at the sessions and does not require any additional processing.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

Link to comment
You mentioned that

 

"For things like room size cues, which I would agree tend to "live" at the low end frequency-wise and the low levels amplitude-wise, I think the word length is where many of the prime benefits lie"

 

and that

 

"one of the great benefits of having a system that can reproduce the bottom octave (and having recordings where this range is captured) is how much more in evidence information about the room is"

I wish to test this by listening to your 16/44.1 and 24/96 samples low-passed.

 

Cheers,

Ricardo

 

Do it any way you like, of course.

However, I don't think such a "test" would reflect anything of great value, unless you normally listen with the top 7 1/2 octaves filtered.

 

Also, if you say you don't hear a difference between high res and Redbook when listening to the full bandwidth, I would doubt adding a filter (which will color both files) will make any differences more audible.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

Link to comment
...

(My wish for a more forward vocal in Winds of Change is likely from hearing most other recordings of vocals in my collection close-miked. Also, I do get to listen to a reasonable amount of live acoustic music.)

 

Hi Jud,

 

Are there any microphones in use in that live acoustic music? Or do you mean completely without reinforcement?

 

It really depends on the player(s) and vocalist(s) too of course.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

Link to comment
If it is properly dithered the only difference between the upper bits and the lower bits is noise....

 

Hi Skeptic,

 

That's the theory. I understand.

To my ears, it just doesn't work that way in practice. Doesn't matter what kind of dither or if there is no dither. Doesn't matter what converters are used for A-D or D-A. It is what I have experienced, regardless of the gear, in every comparison of 16-bit vs. 24-bit I've ever done.

 

Personally, my opinion is that we need a new theory.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

Link to comment
Not sure which of the graphs you were looking at, but here is what John Atkinson said about his measurements:

 

"The 2Ce's upper-frequency output is flat, but with a number of small peaks and dips apparent. The usual metal-dome tweeter resonance occurs at a high 27kHz, well above audibility."

 

Then, with regard to in-room response, first frequency:

 

"The upper midrange and treble in this graph are very smooth, with the slight downward tilt of the curve reflecting both the increased amount of room absorption at high frequencies and the tweeter's increasing directivity."

(Bold mine, underlining standing in for italics by Atkinson.)

 

Finally, regarding in-room time domain response:

 

"In the time domain, despite the Vandersteen's multiway design, its impulse response (fig.7) is as time-coherent as that of the single-driver, crossoverless Fujitsu Ten Eclipse TD712z, reviewed elsewhere in this issue."

 

 

To be honest I don't pay much attention to J.A.'s comments, his analisis of his measurements is overly subjective for my taste (i.e. he ignores or dismisses the mediocre performance of a few Wilson speakers on several measured parameters, and his comments are often tinted by his sonic preferences).

 

One must also take into account is that J.A. averages his (pseudo-)anechoic FR plot from 11 measurements taken over an (in my view excessively) wide 2.40 x 1.00 metre grid and his measurement technique produces an artifact that manifests itself in the form of a ~6dB bump at around 100Hz.

For the on-axis response I usually look at the waterfall but I much prefer the measurements published by Soundstage.

 

Taking this into account, on Fig.4 the response is tilting upwards over three distinct steps (roughly 50-700Hz, 800-4.000Hz, 5.000-18.000Hz) and the same can be seen on Fig.10.

For a really flat J.A.-style (averaged) frequency response plot look here (Fig.3).

 

His in-room response measurements are also averaged so I guess that speakers with wider dispersion in the treble will show a flatter response in this region (and should be kept well away from the side walls).

 

As for the the time-domain, from what I have learnt there really isn't a consensus on the audibility of time alignment so I would say that it's measurement is academic.

"Science draws the wave, poetry fills it with water" Teixeira de Pascoaes

 

HQPlayer Desktop / Mac mini → Intona 7054 → RME ADI-2 DAC FS (DSD256)

Link to comment
Yes, I've read it all too, but it was just close miking which gave a forward balance of the instrument relative to everything else. You used the word shrill, now you are using the word bright. To me they are two completely different things. As I've mentioned, I have many years experience with the instrument and have played many expensive (and cheap) instruments and have also heard all manner of players and instruments from all perspectives - two feet away, in the same small room, in a concert hall, etc.

 

The claims about the Heifetz sound have not been made by self-professed audiophile violin players that I am aware of. They've been made by players who are by far the worst perpetrators of the close miking syndrome themselves such as Itzhak Perlman or audiophiles who don't have 30 years experience with the instruments and haven't heard enough Del Gesu instruments up close and distant (if any at all) to be even remotely qualified to even comment, let alone judge.

 

I can tell you right now and you can chose to believe it (or most likely not). A genuine Del Gesu up close is going to have dollops of volume and a very penetrating, vibrant, powerful tone. That is precisely what they are renouned for. It will have tremendous texture, clarity, definition and depth, but shrillness is something you get from a cheap Chinese violin, not a classic (or even modern) Italian one.

 

Listen to Pinkus Zukerman close miked giving a demonstration and his violin (also a Del Gesu) and it gives precisely the same type of defined, glossy, even edgy if you like sound. But shrill to me is an offensive sound as I would have thought it is to most people. And if Heifetz sounds shrill then it is an equipment problem, not a problem with the recording. Also bear in mind that yes, the older mikes might have had a presence ranges but you get the same effects by moving from a soft acoustic environment to a harder one as well. And you also get the same effect when you have a largely empty concert hall for recording purposes versus a hall full of thousands of people on a concert night. Furthermore, the frequency response of the older recording machines wasn't flat to begin with, nor were the methods used to derive the tapes used as cutting masters nor the resultant records themselves. Nevertheless, I have all the Heifetz SACD releases from 10 years ago and acquired two of the Analogue Products LPs just last week (Bruch Scottish Fantasy and the Sibelius Concerto). Honestly, shrill is the last word I'd use to describe any of these and if anything at all, I don't really find the perspective to be any different to anything much else out there. Yes, I would have said "shrill" on my poor quality equipment from 15 to 20 years ago, but nothing remotely of the sort these days.

 

I admit that "shrill" was a wrong choice, "bright" is probably closer to what I am hearing but as you say might be partly due to the violin he is playing.

"Science draws the wave, poetry fills it with water" Teixeira de Pascoaes

 

HQPlayer Desktop / Mac mini → Intona 7054 → RME ADI-2 DAC FS (DSD256)

Link to comment
I like when you wear your engineer hat :)

 

I think we all have to be aware that there can be commercial pressure on standards.

 

In some instances a lot of people are hanging out hoping for a change in the standard that will enable them to sell a new generation of equipment.

 

This is one reason there has been some industry support for attempts to show that hi-res works. Unfortunately (for the industry) the results have been equivocal at best, so this is coming to be seen as throwing good money after bad by more level-headed managements. Probably the most recent crowd-funded attempt will be the last nail in the coffin.

 

I came across an interesting example of this at a lecture on medical effects of RF radiation. This was at the time when concerns first really came to be heard about mobiles causing brain tumours. It was also at the time when we were getting more information out of the former USSR. Anyway the interesting thing was that they had much stricter regulations than the West in terms of intensity of radiation and duration, and the obvious inference was that these standards had been established in the absence of commercial pressure from telecoms companies.

Mike zerO Romeo Oscar November

http://wakibaki.com

Link to comment
If it is properly dithered the only difference between the upper bits and the lower bits is noise.

If you use TPDF dither without noise shaping, this noise should sound identical to analog "white" noise. (tape hiss, static on a TV screen etc.)

 

This is not exactly true. Rectangular dither causes the average quantization error to be uncorrelated from the musical signal. This removes first order errors, commonly called "distortion". TPDF dither, in addition to removing these first order errors also removes second order errors, i.e. the second moment of the quantization error is uncorrelated from the music. These second order errors are commonly called "noise modulation". However, higher order moments of the quantization noise remain correlated with the musical signal. Therefore the resulting dithered signal is not statistically identical to the original higher resolution signal plus the addition of dither noise. It has been claimed that third and higher order correlations are inaudible, but this is just another psycho-acoustical assertion and has nothing to do with signal processing dither.

 

It is true that subtractive dither eliminates all correlation between quantization errors and the musical signal. A recording that has had dither noise added prior to quantization and subtracted back out during playback will be equivalent to the original full resolution signal plus added dither noise. In addition, the noise floor will be reduced by 6 dB. Subtractive dither was invented in the 1970's for video but has seldom been used with audio, because it is difficult to implement from a systems perspective. I believe it was used in HDCD.

 

There is an excellent PhD thesis that explains the theory of dither in great detail.

 

http://www.iet.ntnu.no/courses/fe8114/files/Wannamaker_phd_2003.pdf

 

There are lots of false claims bandied about in discussions of digital audio. TPDF dither being equvalent to noise is one of them. Close, but no cigar. Sixteen bit audio, made with TPDF dither is good enough for some people but other people prefer 24 bits. My personal preference is that I want all the bits that the people making the recording had without any of these having to be thrown on the floor.

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...