R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 [2] 3  All   Go Down

Author Topic: A question  (Read 12181 times)

David Satz

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 661
Re: A question
« Reply #15 on: June 23, 2005, 10:44:32 PM »

Terry, what you say is true at extremely high (i.e. literally deafening) sound pressure levels. But if this "air distortion" were significant at ordinary listening levels, we would never hear any sounds in our lives except those that had been affected by it. Also, the farther away any sound source is--the more air it has traveled through--the stronger would be its harmonic content, even as the strength of the tone fades due to distance. And finally, that effect should hold true for microphones as well as for human listeners, since the medium is still air; thus any recording made at a distance, even under anechoic conditions, would show notable amounts of harmonic distortion.

Is this getting absurd enough for you yet? None of these things occurs in reality, so I think that you can reasonably set the entire fantasy/conjecture aside.

--best regards
Logged

Nika Aldrich

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 832
Re: A question
« Reply #16 on: June 23, 2005, 11:32:43 PM »

Dan,

Just for the record, I don't think I've anywhere, in any context seen Apogee "promote" 192kS/s sample rates.  I have seen them make said available to those who require it to meet market demand from record labels, etc.  But I have never seen anything in Apogee's writing regarding supposed benefits of extraneously high sample rate recording.

Nika
Logged
"Digital Audio Explained" now available on sale.

Click above for sample chapter, table of contents, and more.

Lucas van der Mee

  • Newbie
  • *
  • Offline Offline
  • Posts: 12
Re: A question
« Reply #17 on: June 24, 2005, 10:42:28 AM »

And since we are setting things straight again:

I am proud to say, Apogee designs are 100% Lavry-free, they have been for over a decade and …
…business is better than ever!

Lucas van der Mee
Sr Design Engineer
Apogee Electronics
Logged
Lucas van der Mee
Sr. Design Engineer
Apogee Electronics

danlavry

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 997
Re: A question
« Reply #18 on: June 24, 2005, 04:43:37 PM »

[quote title=Terry Demol wrote on Fri, 24 June 2005 02:13
I've been thinking about this over the last few days and
maybe there is something we haven't considered here.

We know for a fact that air itself manifests 2nd harmonic
distortion on any sound wave travelling through it due to
the density difference between the high and low pressure
parts of a wave (compression and rarefaction).

We also know that any medium that imposes 2nd harmonic
distortion on a wave will also impose intermodulation
distortion.

So it appears to me that there will be some intermodulation
occuring by the air "carrier" itself before the sound reaches
our ears.

Does this make sense?
Cheers,
Terry
[/quote]


Terry,

Lets first agree that the ear can hear a certain bandwidth (for example 22KHz corresponding to 44.1KHz sampling or even 48KHz corresponding to 96K sampling).

Whatever we hear in the live performance space WILL include ALL the signals that we want to record and reproduce. Assuming that the air manifests harmonics, intermod or whatever you wish to assume, if it falls within the hearing range, it is already recorded. The mic (covering the audio range) will pick it up.

Adding high frequency capability that causes the same alterations (harmonics, intermod or whatever) on top of material that already contains the audible outcome, means you are doing it twice.

So assuming that such alterations could take place and have a sonic outcome, one is better off to make sure that we DO NOT include the high frequencies. The inclusion of the high frequencies will “double up” the effect, when comparing with the reference material (original performance).

For example, say we have 29KHz and 30KHz tones, and some mechanism in the air to generate a difference of 1KHz. That 1KHz is audible, will be recorded in the performance space and heard on playback. But including the 29KHz and 30KHz in the recording will introduce the ADDITIONAL 1KHz energy due to a new interaction, on top of the already recorded 1KHz. Of course the proper amount of 1KHz is is already in the recorded material, so the additional energy is unwanted.

If what you suggested is correct, it would amount to one more argument AGAINST recording signals outside the hearing range.

Regards
Dan Lavry
www.lavryengineering.com


Logged

kraster

  • Full Member
  • ***
  • Offline Offline
  • Posts: 199
Re: A question
« Reply #19 on: June 24, 2005, 07:20:47 PM »

Can the same argument be used concerning intermodulation distortion resulting from non linearities in loudspeakers? ie. The speaker "hears" or passes frequencies above the range of human hearing and causes beat tones in the audible range as a result of non linearities in the speaker. If somehow these Beat tones are restricted to the upper range of human hearing could it go someway in explaining peoples discernment of an exaggeration of top end when conducting higher sample rate tests? Is there some kind of device in the output path that could restrict the beat tones to a high range? e.g. the crossover.

I can't hear 20khz no matter how hard I try but the intermodulation distortion from the speaker (if restricted to the upper range)could fool me into thinking that I'm hearing extra stuff at the top end.

I've heard a lot of people talking about better transient definition in higher sample rates but could the speaker distortion just exaggerate the top end leaving the impression that there is better transient response?

(sorry if this has been covered I'm just curious)

Karl
Logged

bobkatz

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2926
Re: A question
« Reply #20 on: June 25, 2005, 06:48:45 AM »

David Satz wrote on Tue, 21 June 2005 12:37



reasons in particular situations--not because "wider bandwidth sounds better." The latter claim is widely believed by audiophiles, and it's the kind of statement which can't ever be disproved, so they go on believing it. But there hasn't been any proof of it in all these years, either, and one would think that it could rather easily be proved if it were true.

--best regards



It's fortunate that in the digital domain we can construct filters and make listening experiments that put the nail on the coffin of "wide bandwidth is better per se". I'm nearly 100% convinced that "it's the filters, not the bandwidth, that we hear" (when band limiting above about 20 kHz). Ironically, we need extremely high sample rates in order to conduct the experiments to prove that we don't need them  Smile
Logged
There are two kinds of fools,
One says-this is old and therefore good.
The other says-this is new and therefore better."

No trees were killed in the sending of this message. However a large number of
electrons were terribly inconvenienced.

kraster

  • Full Member
  • ***
  • Offline Offline
  • Posts: 199
Re: A question
« Reply #21 on: June 25, 2005, 01:33:14 PM »

I suppose the point I was trying to make above is that in listening tests both professionally conducted and those,in particular, less rigorously controlled there are other salient factors such as loudspeaker intermodulation distortion at higher sampling rates and bad filter design at lower sampling rates that conribute to differences perceived between them. I've no doubt that many people perceive differences between them but ultrasonic perception is pretty far down the list of plausible explanations for the perceived differences.

Karl
Logged

kraster

  • Full Member
  • ***
  • Offline Offline
  • Posts: 199
Re: A question
« Reply #22 on: June 25, 2005, 03:13:41 PM »

Lucas van der Mee wrote on Fri, 24 June 2005 15:42


…business is better than ever!




I think it quite disingenuous to be boasting about your business success in a thread that's debunked some of your misleading (by your own admission) marketing material.
Logged

Lucas van der Mee

  • Newbie
  • *
  • Offline Offline
  • Posts: 12
Re: A question
« Reply #23 on: June 25, 2005, 08:09:18 PM »

Point taken Karl...
I am just very annoyed by Dan's repetitive false claims on my work.
I think I should have written:...and our equipment is better than ever.

Lucas van der Mee
Sr. Design engineer
Apogee Electronics
Logged
Lucas van der Mee
Sr. Design Engineer
Apogee Electronics

David Satz

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 661
Re: A question
« Reply #24 on: June 25, 2005, 08:30:54 PM »

Mr. Van der Mee, up to this point the postings in this thread from people at Apogee have been greatly to Apogee's credit. And now for some reason you evidently would like to change that.

Just speaking as an ordinary user of this forum, if you wish to pursue a dispute with Dan Lavry, I hope that you will do so in a new discussion thread specifically devoted to the facts of that situation. This thread has already achieved what it set out to do. And since the rest of us here haven't got much clue as to why you're running into thermal overload, kindly be specific as to your facts. Or maybe just count to 10 (in decimal, please, not binary) and think it all over.

--best regards
Logged

Terry Demol

  • Full Member
  • ***
  • Offline Offline
  • Posts: 103
Re: A question
« Reply #25 on: June 26, 2005, 08:31:00 AM »

[quote title=danlavry wrote on Fri, 24 June 2005 21:43]
Terry Demol wrote on Fri, 24 June 2005 02:13
I've been thinking about this over the last few days and
maybe there is something we haven't considered here.

We know for a fact that air itself manifests 2nd harmonic
distortion on any sound wave travelling through it due to
the density difference between the high and low pressure
parts of a wave (compression and rarefaction).

We also know that any medium that imposes 2nd harmonic
distortion on a wave will also impose intermodulation
distortion.

So it appears to me that there will be some intermodulation
occuring by the air "carrier" itself before the sound reaches
our ears.

Does this make sense?
Cheers,
Terry
[/quote



Terry,

Lets first agree that the ear can hear a certain bandwidth (for example 22KHz corresponding to 44.1KHz sampling or even 48KHz corresponding to 96K sampling).

Whatever we hear in the live performance space WILL include ALL the signals that we want to record and reproduce. Assuming that the air manifests harmonics, intermod or whatever you wish to assume, if it falls within the hearing range, it is already recorded. The mic (covering the audio range) will pick it up.




Yes, of course.

Maybe there is "air IM" but, as ou say, it doesn't matter. As
long as the ADC converts what the ear would hear, that's all
that counts.  

Quote:



Adding high frequency capability that causes the same alterations (harmonics, intermod or whatever) on top of material that already contains the audible outcome, means you are doing it twice.

So assuming that such alterations could take place and have a sonic outcome, one is better off to make sure that we DO NOT include the high frequencies. The inclusion of the high frequencies will “double up” the effect, when comparing with the reference material (original performance).

For example, say we have 29KHz and 30KHz tones, and some mechanism in the air to generate a difference of 1KHz. That 1KHz is audible, will be recorded in the performance space and heard on playback. But including the 29KHz and 30KHz in the recording will introduce the ADDITIONAL 1KHz energy due to a new interaction, on top of the already recorded 1KHz. Of course the proper amount of 1KHz is is already in the recorded material, so the additional energy is unwanted.




I see your point here, but I'm not sure I agree %100.

It would depend on the ADC's inherent IMD at higher
frequencies. It should be very low for a well designed unit
and as such should not be a significant issue. Possibly some
opamp based IP circuits are more susceptible to higher freq IM
due to their rising distortion Vs rising freq characteristics
but there are others that are very good HF performers.

Also IME low gain apps such as ADC front ends are the easiest
to keep linear at higher frequencies.


Regards,

Terry
Logged

Johnny B

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1134
Re: A question
« Reply #26 on: June 26, 2005, 12:14:34 PM »

[quote title=bobkatz wrote on Sat, 25 June 2005 11:48]
David Satz wrote on Tue, 21 June 2005 12:37


"Wider bandwidth sounds better."


Many "ear people" feel this truly is the case, oddly, it is only with respect to converters that some people argue for a "lesser" or inferior bandwidth.







Logged
"As far as the laws of mathematics refer to reality,
they are not certain; as far as they are certain,
they do not refer to reality."
---Albert Einstein---

I'm also uncertain about everything.

bobkatz

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2926
Re: A question
« Reply #27 on: June 26, 2005, 12:21:42 PM »

[quote title=Johnny B wrote on Sun, 26 June 2005 12:14]
bobkatz wrote on Sat, 25 June 2005 11:48

David Satz wrote on Tue, 21 June 2005 12:37


"Wider bandwidth sounds better."


Many "ear people" feel this truly is the case, oddly, it is only with respect to converters that some people argue for a "lesser" or inferior bandwidth.





To quote Bob Olhsson from my book: "The issues of the audibility of bandwidth and the audibility of artifacts caused by limiting bandwidth must be treated separately. Blurring these issues can only lead to endless arguments."

Remember that it is impossible to create a filter in the analog domain that does not have phase shift, some noise, and some distortion. The more complex the filter in the analog domain, the worse its potential to sound. This is not necessarily the case in the digital domain. Thus, a very good reason why wide-bandwidth analog circuits often sound better... a simple one-pole low-pass filter at 100 kHz is pretty invisible to the ear. But a simple one-pole filter at 20 kHz is not. And constructing a complex, sharp low-pass filter at 20 kHz in the analog domain is almost sure to result in audible artifacts.

I could go on, I could write a book about it. Oh wait, I did write a chapter already Smile. More evidence is in, by the way, on the side of the argument that "it is the filters, not the bandwidth, that we hear", and I plan on putting that in the second edition of "Mastering Audio."
Logged
There are two kinds of fools,
One says-this is old and therefore good.
The other says-this is new and therefore better."

No trees were killed in the sending of this message. However a large number of
electrons were terribly inconvenienced.

danlavry

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 997
Re: A question
« Reply #28 on: June 26, 2005, 05:35:58 PM »

[quote title=bobkatz wrote on Sun, 26 June 2005 17:21]
Johnny B wrote on Sun, 26 June 2005 12:14

bobkatz wrote on Sat, 25 June 2005 11:48

David Satz wrote on Tue, 21 June 2005 12:37


"Wider bandwidth sounds better."


Many "ear people" feel this truly is the case, oddly, it is only with respect to converters that some people argue for a "lesser" or inferior bandwidth.





Remember that it is impossible to create a filter in the analog domain that does not have phase shift, some noise, and some distortion. The more complex the filter in the analog domain, the worse its potential to sound. This is not necessarily the case in the digital domain. Thus, a very good reason why wide-bandwidth analog circuits often sound better... a simple one-pole low-pass filter at 100 kHz is pretty invisible to the ear. But a simple one-pole filter at 20 kHz is not. And constructing a complex, sharp low-pass filter at 20 kHz in the analog domain is almost sure to result in audible artifacts.



Hi Bob,

As long as you are at it, lets keep the eye on the ball and see when and if we have to deal with what:

With today's technology, you DO NOT have to deal with 20KHz analog filter.

On the AD side, the conversion is done at a high oversampling rate, and filtering (decimation) to 20KHz (for 44.1KHz CD) is in the DIGITAL domain. As you stated, it can be done without any phase shift.  

On the DA side, the data, even with CD's (as low as 44.1KHz rate) is up sampled to a much higher rate in the DIGITAL DOMAIN and it can be done without any phase shift.  

So all objections that are based on a 20KHz analog filters have no leg to stand on. Talking about it is, in fact, raising issues that were gone 15 years ago...

Regards
Dan Lavry
www.lavryengineering.com
Logged

danlavry

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 997
Re: A question
« Reply #29 on: June 26, 2005, 05:50:33 PM »

[quote title=Terry Demol wrote on Sun, 26 June 2005 13:31

I see your point here, but I'm not sure I agree %100.

It would depend on the ADC's inherent IMD at higher
frequencies. It should be very low for a well designed unit
and as such should not be a significant issue. Possibly some
opamp based IP circuits are more susceptible to higher freq IM
due to their rising distortion Vs rising freq characteristics
but there are others that are very good HF performers.

Also IME low gain apps such as ADC front ends are the easiest
to keep linear at higher frequencies.

Regards,
Terry
[/quote]


Terry,

I believe you were talking about some ultrasonic impact that happens in the air. David Satz. pointed out that such interaction only takes place at very high pressure levels, and to the best of my knowledge he is correct. But I decided to go with the assumption that some degree of what you said may possibly be correct, and the simple logic I used suggests that it is UNDESIRABLE to include the ultrasonic.

Now you are talking about AD non linearity at ultrasonic frequencies, which could completely changes the conversation. But "surprisingly", my answer is the same:

Any non linearity at high frequency, be it a converter, the air itself, the speakers, is undesirable. The presence of such non linearity will ADD signals that were not present prior to recording. The signals we want are already picked by the mic.

So anyone that wishes to have their gear extended to higher frequencies, better be sure that the linearity holds up over the range of operation. The arguments are often: "well, the device is less linear way up there but we do not hear that high", and of course such arguments are flawed, because non linearity may make signals we do hear at lower frequencies.

I read the Boyk paper a number of times with great interest, and I appreciate his work and his contributions. He measured all sorts of musical instruments in very high frequencies, and I do not dispute his findings at all. What I disagree with are the conclusions regarding what needs to be done in view of his findings. Some people, including some well known ear people understandably came to the very simplistic conclusion that we need to record it all, and that there is no harm in doing so. I disagree with that conclusion:

We want to record what we hear at the performance place, and nothing more. Any mechanism AT THE PERFORMANCE, by which high frequency energy will "fold back" to the audible range will be picked up and recorded. We need to stay true to that recorded material. We can do so by LIMITING the mic to the bandwidth of the ear, and by NOT ALLOWING the material we do not hear into the electronics. If we allow high frequencies that we did not hear at the performance into the electronics audio chain, we in fact become "sitting ducks" to any non linearity we may encounter.

Of course, I am not suggesting that all audio gear be limited to 20KHz. The frequency cutoff is a COMPROMISE between various factors (including "safety margins"). I am suggesting that it is good to eliminate signals above what we can hear, when all considerations allow so. I am saying that the arguments suggesting an advantage in recording higher frequencies than we can hear are "180 degrees out of phase". ..Or put another way the opposite of what they should be.

We are lucky that mic and speaker makers did not push the bandwidth in the manner that the converter makers did. With 20KHz mic, a high frequency non linearity is not a problem because there are no high frequency "tones" (energy) to "fold back".
Again, the high frequency (beyond our hearing) is either "trouble" or "potential trouble", making one more argument against 192KHz sampling rate. I have a few more up my sleeve, for another time.

A little off subject:

The MP3's are getting better, because they are trying to be "almost as good as possible". To do so, one needs to have a relatively clear grasp of what is the maximum possible required bandwidth and dynamic range. Only then does one start applying principles of psychoacoustics for data compression.

Meanwhile, the "leadership" of pro audio (many of whom are very lacking in technical  knowledge) have pushed the industry into a far from optimal place! Stay tuned for the next installment of this insanity - people with lesser technical leading the industry towards 384KHz. Are the blind leading the ignorant, or are the ignorant leading the blind?

Regards
Dan Lavry
www.lavryengineering.com

Logged
Pages: 1 [2] 3  All   Go Up