Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Glenn Bucci]

I am not as smart as you tech guys, but a former engineer from the BBC stated this in regards to recording at 88 and 96 as compared to 44.

"Aside form the changing storage requirements and data through rates, there aer plenty of other changes too.

Metering is significantly more accurate at higher sample rates, as is dynamic processing such as compressors and limiters.

High EQ is more 'analogue like' with higher rates too because the response curves on the HF side don't have to be curtailed due to the brick-wall response filters.

And then there is the anti-alias/reconstruction filtering which, being an octave higher, is significantly less critical, allowing relatively poor designs to sound rather good.

personally, I work at 24/96 most of the time....

hugh

--------------------
Technical Editor, Sound On Sound

Do you guys agree with this?

[trevord]

He lists the benefits of doing higher sample rates.
I think the discussion is more about "can we do higher sample rates good enough to get the benefits"
The way I like to think of it is ...
Everyone know the theoretical benefits of digital when sony started, but would you say the sony pcm on beta demonstrated those benefits.
or
How long was it before we could do 16/44.1 well enough to not annoy the casual listener?

Higher sample rates are nice.. but there is even more digtial crap to get wrong.

[danlavry]

Revelation wrote on Thu, 08 June 2006 01:29

I am not as smart as you tech guys, but a former engineer from the BBC stated this in regards to recording at 88 and 96 as compared to 44. "Aside form the changing storage requirements and data through rates, there aer plenty of other changes too. Metering is significantly more accurate at higher sample rates, as is dynamic processing such as compressors and limiters. High EQ is more 'analogue like' with higher rates too because the response curves on the HF side don't have to be curtailed due to the brick-wall response filters. And then there is the anti-alias/reconstruction filtering which, being an octave higher, is significantly less critical, allowing relatively poor designs to sound rather good. personally, I work at 24/96 most of the time.... hugh -------------------- Technical Editor, Sound On Sound Do you guys agree with this?

Hi hugh

"Metering is significantly more accurate at higher sample rates, as is dynamic processing such as compressors and limiters."

Digital processing and linear processing go hand in hand, but digital and non linear processing have an inherit problem.
Regarding the comments on limiters and compressors, I would expand the comments to include all non linear processing done in the digital domain (such as tube emulation and more).

In analog, the non linearity that produce very high frequencies beyond hearing, can be managed. In theory, what is too high to hear will not be heard. In practice one can filtered it out.  It is a good idea to filter to avoid possible distortions due to inter-modulation in the gear later in the audio path (avoiding unwanted energy from getting back to the hearing range).

But with digital, all high frequency above Nyquist will be aliased back to the hearing region. So non linear processing, generating such high frequency can be a problem. The first impulse is to view the increase in sample rate as a positive step. But if you examine the distortions due to various non linearity (certainly a hard limiter), you find they tend to decay very slowly (with increase frequency) in relationship to the ear’s amplitude response (near logarithmic). For example, a 1% is not much on a linear scale, but for the ear it is at -40dB, still a lot of distortion...  

Avoiding aliasing would require setting Nyquist extremaly high, where the generated high frequencies (due to the non linear process) is at some vey low amplitude (say -100dBFS). The required increase in sample rate would be so very high rates, possibly tens of MHZ. A simple X2 or X8 does not buy you that much.

But I do not have a problem with anyone that wants to up sample LOCALY to very fast rate for a good reason, do what they need to do, then if possible down sample back.  I would direct you to see much better explanation in my paper “Sampling Theory” on my web, under support at www.lavryengineering.com  

One needs to make a distinction between the sample rate of a format and “localized sample rates” such as those we use for AD front end (modulators) and DA back end (interpulators). We do such things (sample way up by X64 or even X1024) with modern DA’s because it solves the requirement for a stiff and costly analog filter. We also do the opposite with AD’s, sampling very fast (X64 to X1024) at the front end, then decimating down.  

"High EQ is more 'analogue like' with higher rates too because the response curves on the HF side don't have to be curtailed due to the brick-wall response filters".

Again, I do not have a problem with  “localized” up-sampling for a particular reason, even to a very high rate (assuming a good implementation of course) followed by down sampling back to the proper format.

"And then there is the anti-alias/reconstruction filtering which, being an octave higher is significantly less critical, allowing relatively poor designs to sound rather good".

Again, I would encourage you to read my paper Sampling Theory. That old argument about the analog anti aliasing and reconstruction was very important in the days prior to the concepts of over sampling AD’s and up sampling DA’s. The analog filters required for the non oversampling/upsampling 44.1KHz were very stiff indeed. But we have put those analog filter issue behind long ago (a dozen years or so).

There are serious down sides to sampling too fast. There is a big difference between upsampling for easing a computation, or fighting a non linear alaising problem, or making an analog filter circuit work well. In all those case, one is not trying to accomodate a wider bandwidth for the audio. Take a 96KHz sampler (48KHz bandwidth) and upsample it to 12.288MHz, and the audio bandwidth is still at 48KHz. That is different from trying to have a converter operate at 12.288Mhz with 6MHz bandwidth. Making a converter with higher audio bandwidth, way beyond what we hear, just costs in increased distortions and noise.

Finlay, it is not about faster is better. Clearly so. One will be nuts to just go faster and faster. There is no need for 1GHz audio. Of course going too slow will "eat" into the hearing range. The issue is about the OPTIMAL RATE.

At 96KHz we are already over the optimal rate. Going faster is the wrong thing to do.

Reagards
Dan Lavry
http://www.lavryengineering.com

[Jon Hodgson]

Revelation wrote on Thu, 08 June 2006 01:29

I am not as smart as you tech guys, but a former engineer from the BBC stated this in regards to recording at 88 and 96 as compared to 44.

Well bearing in mind that he's saying 96 and not 192, he's right, kindof.

Due to that last downsampling filter in the converter 96 is above the optimal rate for sampling (though not always for processing) whereas 48 is below it. So each has its advantages. The problem is that people assume that if 96 is better than 48, then 192 must be better again.

There is also a question of what happens when you process or generate sounds especially with non-linear processing. Higher localized sample rates can improve this, as Dan rightly points out. However in many cases plugins use internal sample rates which are equal to, or a fixed multiple of, the system sample rate. Doubling your system sample rate will double these, so I can imagine that people might try various experiments, find that certain things do sound better when processed with a higher system sample rate, and then falsely conclude that higher sample rates were better globally.

[danlavry]

Revelation wrote on Thu, 08 June 2006 01:29

I am not as smart as you tech guys, but a former engineer from the BBC stated this in regards to recording at 88 and 96 as compared to 44. "Aside form the changing storage requirements and data through rates, there aer plenty of other changes too. Metering is significantly more accurate at higher sample rates, as is dynamic processing such as compressors and limiters. High EQ is more 'analogue like' with higher rates too because the response curves on the HF side don't have to be curtailed due to the brick-wall response filters. And then there is the anti-alias/reconstruction filtering which, being an octave higher, is significantly less critical, allowing relatively poor designs to sound rather good. personally, I work at 24/96 most of the time.... hugh -------------------- Technical Editor, Sound On Sound Do you guys agree with this?

Hello Hue,

Let me elaborate a little on what I sated earlier:

I agree that 44.1KHz may be a bit tight at the high audio frequency range. And I also agree that having an "octave of margin" is helpful. At 88.2 or 96KHz we certainly have a lot of margin, a bit too much from a converter stand point, but I do not want to argue about 10-20KHz, where there are some benefits to a X2 rate - say you want to make a CD (44.1KHz), then 88.2KHz is a "natural" place to be from a down sample stand point, because it calls for a 2:1 rate reduction, which is a SYNCHRONUS sample rate, certainly easier to avoid the SRC problems associated with running asynchronous clocks. So while I would prefer 60-70KHz for the optimum sample rate, I see some wisdom in 88.2 or 96KHz (at least in terms of "backward compatibility to CD 44.1KH rate or 48KHz).

But having an octave or so margin does not call for another octave, then another and another... One MUST ACCEPT the fact that there is such a thing as OPTIMAL SAMPLE RATE. The industry salesmen trying to promote 192KHz, and now some are at 384KHz have no technical basis for it, not even many years AFTER the introduction of such a concept. They sure tried to explain it.

First by the wrong notion that more dots yield a more accurate wave (the analogy was pixels on a picture). That was based on using "street sense", totally contrary to reality.

Then, by claiming all sorts of stuff about narrow impulse response yielding better time location. That argument totally ignored the direct connection between impulse width and bandwidth of the system INCLUDING THE EAR.

Then came the argument about tight analog filters. That argument does not hold for the last dozen or so years because we over sample the AD's front end, and we up sample the DA's back end, so a 3 pole filter at some high frequency yields the needed performance.  

Then came an argument about latency, which only addresses a small range of applications, and in fact is more about AD architecture. FIR based sigma delta has a relatively long delays, say 1-2msec at 44.1KHz. The proponents of low latency wished for 500usec which is 6 inches of acoustic distance (mic location, speaker location...)

I could go on and on. Needless to say I did take a technical stand against all the false arguments. First at the AES, chairing a tutorial about AD's, then with my paper "Sampling Theory". I will continue to do so, with a new paper about the trade offs between speed (sample rate) and accuracy (distortions and noise).

One octave up may be a "comfortable margin", or a slight over kill, but I can accept it as reasonable. But folks need to realize that going faster, not only increases the data size, not only requires a lot more signal processing, it also reduces the quality of AD and DA conversion. All things held equal, a 88.2KHz-96KHz is better then the sales and marketing driven 192KH.

The comments about a limiter or compressor are in order. I believe I was the first to point out that some of the complaints regarding “digital sound” are based on the fact that non linear processing and digital do not “go hand in hand” because the aliasing problems. I specifically stated that a non linear polynomial  curve y=a0+a1*X+a2*X^2+a3*x^3… requires double the bandwidth for the X^2 term, a triple bandwidth for the X^3 term and so on.  So clearly one must watch what they do when dealing with non linear transfer function in digital.

One way is to do some processing in analog (such as final compression, hard limiting, tube sound…). The other way is to ups sample way up, to the point where the high frequency due to non linearity is at very low amplitude. Only then can you “live with” the aliased energy.

Of course the need for a couple of processes is not enough to call for 10MHz sampling, or whatever it takes to accommodate say a hard limiter.

One more time: There is a big difference between 2 cases:

1.Converting say at 96KHz (48KHz bandwidth), then up-sampling to very high rate, such as 4fs, 8fs…1024fs. In this case, we are using a computational process, and the audio bandwidth stays the same.

2. Having an AD or DA convert at that higher rate such as  4fs, 8fs…1024fs. In this case the “audio” bandwidth is at near half the sample rate. Accommodating such a wide signal range is very costly in terms of performance. The reasons for it are both theoretical (tradeoff between bandwidth and accuracy for sigma delta) and practical (many analog circuits tradeoffs come to mind).

So clearly, if one insists on a digital solution to non linear processing, go for case 1 (the up sampling case), not case 2 which trades of nearly everything – distortions, noise, file size, processing requirement and more!

Regards
Dan Lavry
www.lavryengineering.com  

[danlavry]

I was told the link to "Sampling Theory" is broken. While the link from my "support" is being fixed, here is another link to the paper "Sampling Theory":

 http://lavryengineering.com/forum_images/Sampling_Theory.pdf

Reagrds
Dan Lavry
http://www.lavryengineering.com

[danlavry]

A couple of additional comments:

My significant effort to protect the audio industry from being steered in the wrong direction by audio salesmen has been rather successful. I have received support from technical people who did not want to take a public stand. Some work for companies that make 192KHz IC's or 192KHz gear. Most are are "understandably shy".

Non- technical arguments (subjective, subjective and more subjective, such as "self administrated" listening tests) abound (from people that stand to benefit from the sale of 192KHz.)

Where are the sliders on the ProTools EQ system that boost or cut the 192Khz signals at 96KHz? At 48KHz?
I asked a person employed by Protools demonstarting the system that question. His answer: no one was able to hear it. So I asked if he anyone could hear the 30KHz slider. His answer was, we put it there because a famous recording guy wanted us to put it there.

At the recent  (cigarette smoke-infested) AES Paris, an engineer I know complained to me that my anti 192KHz stand cost his company money. I said,"But you are an EE, a good one in fact. So you should understand." He replied "I do understand you technically, and I know 192KHz is BS but I am also a marketing guy, so if they will buy it, I will sell it."

Will he sell his sister for 800 Euros?

When I designed medical devices you can bet the lawyers would be on a medical device company that pulled anything like this bogus 192KHz con.

Regards
Dan Lavry
www.lavryengineering.com

[Jon Hodgson]

danlavry wrote on Thu, 08 June 2006 22:57

Will he sell his sister for 800 Euros?

If she's cute then let me know his answer  

[crm0922]

The poster "Revelation" is not "Hugh".  

He was posting a quote from "Hugh Robjohns", an editor at Sound on Sound.  I assume he was the former BBC engineer referred to.

Just wanted to make that clear.  A well known, respected member of the community was making those points.

He also defended the Roger Nichols article in the SOS forum.  And then backed off slightly I guess.

Chris

[UnderTow]

After thinking about it and talking to some people, Hugh Robjohns backed-off completely and agreed that the article was wrong.

Alistair

[Graham Jordan]

I was just putting my points into the thread on SOS about the Roger Nichols article, and had some good responses from Hugh. He quite clearly realises that the article is "deeply flawed."

Quoting from the SOS forum:

GJordan

Hugh Robjohns GJordan What the article is suggesting is that when you add this noise to a high frequency signal you hear it less than with a low frequency signal. ROger was suggesting that if you count up the number of quantising levels crossed (in terms of 24 bit LSBs), the relative proportion of maximum potential error compared to the overall difference in quantising level for two adjacent samples is far greater for LF signals than for HF signals. The pure logic of what he said seems right to me, but the concept itself is deeply flawed and not applicable to practice. Jackpot! This should be requoted in it's own entry. Very well put - although I might have said "and not applicable to anything in audio (theory or practice, dither or no dither)." Hopefully I've covered some of the reasons why.

[danlavry]

Graham Jordan wrote on Fri, 09 June 2006 19:35

I was just putting my points into the thread on SOS about the Roger Nichols article, and had some good responses from Hugh. He quite clearly realises that the article is "deeply flawed." Quoting from the SOS forum: GJordan Hugh Robjohns GJordan What the article is suggesting is that when you add this noise to a high frequency signal you hear it less than with a low frequency signal. ROger was suggesting that if you count up the number of quantising levels crossed (in terms of 24 bit LSBs), the relative proportion of maximum potential error compared to the overall difference in quantising level for two adjacent samples is far greater for LF signals than for HF signals. The pure logic of what he said seems right to me, but the concept itself is deeply flawed and not applicable to practice. Jackpot! This should be requoted in it's own entry. Very well put - although I might have said "and not applicable to anything in audio (theory or practice, dither or no dither)." Hopefully I've covered some of the reasons why.

With so many "quotes" I can no figure out who said: "The pure logic of what he said seems right to me, but the concept itself is deeply flawed and not applicable to practice".

But whoever said is not being completely straight forward. The statement seems like a political attempt at "damage control". Saying that the "pure logic seems right" is out of place. The pure logic is completely flawed.

The statement should be: "the concept itself is deeply flawed".
 
Adding that "it is not applicable to practice" leaves room to excuse the concept as if it has some non practical value. The fact is, the concept is completely out to lunch, practical or not.

When one partially admits, one partially does not admit. Regaining credibility calls for a clear admission that the concept is wrong, with no qualifiers.

In my opinion, he put out a fight, he is correct he deserves better then a "half ass-ed" response.

Dan Lavry
Lavry Engineering
http://www.lavryengineering.com

[Graham Jordan]

danlavry wrote on Fri, 09 June 2006 12:17

With so many "quotes" I can no figure out who said: "The pure logic of what he said seems right to me, but the concept itself is deeply flawed and not applicable to practice".

That was Hugh, SOS Technical Editor.

[Graham Jordan]

danlavry wrote on Fri, 09 June 2006 12:17

Adding that "it is not applicable to practice" leaves room to excuse the concept as if it has some non practical value. The fact is, the concept is completely out to lunch, practical or not.

Exactly. That why I said that I would have put it: "and not applicable to anything in audio (theory or practice, dither or no dither)." To show that it's not just flawed in a practical sense, but on every level.

I did also manage to looks at a collegue's copy of the SOS issue (although was US/International edition, and the layout was slightly different to that in the PDF, much actually much digger diagrams).

[danlavry]

crm0922 wrote on Fri, 09 June 2006 10:04

The poster "Revelation" is not "Hugh".   He was posting a quote from "Hugh Robjohns", an editor at Sound on Sound.  I assume he was the former BBC engineer referred to. Just wanted to make that clear.  A well known, respected member of the community was making those points. He also defended the Roger Nichols article in the SOS forum.  And then backed off slightly I guess. Chris

The first post was confusing to me. I thought it was posted by the signer.

Regards
Dan Lavry
www.lavryengineering.com