That is puzzling, as we know since the first time Boyk's "there is life above 20 kHz" was cited that "there" exists something.
Spectra of the beloved gamelan music show that as well.
I don't know exactly what you mean by "ultra short duration" but wrt to Tournesol's 24/96 recordings the above example IMO count.
There still is dispute about the audibility for human listeners, but beginning with Oohashi et al.'s work and the quite numerous follow up studies it is a bit hard to argue that no evidence is available.
The last time Nelson Pass in addition mentioned an examples of frequencies reproduced above 20 kHz by seperate speakers and leading to a difference tone perception in human listeners.
You might want to read more carefully.
What was said was that if only a single since wave is present and sampled data is perfect (meaning infinite resolution and accuracy) then 3 points is enough to exactly identify the sine wave (including phase information relative to some reference).
It was a statement about measurement, not reproduction.
Last edited:
Don't think so, at least I wouldn't buy the "completely different things" line...
But, if an audible difference exists, does it make sense to argue that "Redbook" is capable to handle the rise time, but not the above 20 kHz content or vice versa?
You are correct, I pulled out of context in response to someone else pulling out of context...I stand corrected.
My point was 3 points may be enough assuming infinite everything, but that is not what redbook does.
jn
Last edited:
A great example of not understanding the issue. The Haas effect is about masking echoes or time delays of a continuous signal. It has nothing to do with the current discussion.
I use the effect all the time when multiple loudspeakers are used. Any signals within 15 dB of each other should arrive inside a 15 millisecond envelope. That way the perception is all the sound is coming from the earlier arrival source.
Not sure, where and why the 60 kHz suddenly arises, as in case of "Redbook" we are talking about attenuation of everything above 20 kHz by 30 - 100 dB and no actual content at 22.05 kHz and above.
Fastest rise time for a 44.1 kHz system would be (that is what is used for impulse response measurement) ~18 us using artifical construction of the signal.
An exchange seems impossible, let's stop there. I tried, I lost.
In matter of audio, I use my ears. Do-you use a red book ?
Last edited:
Isn't it the maximum slew-rate that is interesting? That would be the derivate for a 22,049 Khz sinus at the 0-crossing - no?
//
I know I'm naive, but I cannot stop hoping the link below would help in clearing the muddy waters and debunk some myths and legends around digital audio processing...
Note, the article illustrates the basics, otherwise it is outdated; for example "AD converter designers can not generate 20 bits at MHz speeds" is wrong, in fact I have one on my bench right now.
http://lavryengineering.com/pdfs/lavry-sampling-theory.pdf
Note, the article illustrates the basics, otherwise it is outdated; for example "AD converter designers can not generate 20 bits at MHz speeds" is wrong, in fact I have one on my bench right now.
http://lavryengineering.com/pdfs/lavry-sampling-theory.pdf
Simon, it seems to me you both are not talking of the same thing.
RNM seems interested at the effect of phase shifts inside a signal on the way we perceive it (group delay). You, and I understand why ;-) on the way our ears can separate multiple sources of the same signal in time (echo, reverberation etc...)
Am-i wrong ?
It seems that today I am making desperate efforts towards a great reconciliation with full of love everywhere. Someone called-me Don Quichote. Scott, I demand a medal.
Last edited:
The link which appears to show spectral content of a cymbal to 60kHz despite lack of cal on the microphone. I am just saying that the rise time of the initial transient does not have to be and should not be expected to need that level of FR. This is all linked to the fact that TT seems to think that transients in music need a bandwidth greater than 20kHz by using only his ears.
So far the adequacy of redbook for replay of music in a domestic environment seems to hold.
Hmmmm. Understanding something you have not heard yet doesnt seem likely to move the needle much. ?? You'll get 'CD is fine and really is for all practical purposes, perfect.'
Just try it and see. its faster in this case. And, gives an accurate answer. At least for that one person.
THx-RNMarsh
No I mean in hardware. can you just double the clock and few other things tweeked and have it?
I have BenchMark ADC and DAC
THx-RNMarsh
Thanks for the link.
And yes, outdated. One specific thing is that at the higher rate there is a loss of accuracy. You specifically point out that it is not true anymore with a converter on your bench.
Of note in the article, on page 25 it shows the filtered 17Khz signal as being close to one half the level when it's converted at 44k. Does anybody have a phase plot associated with that level of LP filtering? Is phase a rock out to 20K, or changing. BTW, it's nice to see that a filter that is shown as quite flat out to 20Khz still drops the 17Khz signal by almost half.
In the end, all three cons listed on page 26 are totally outdated (memory and processing power certainly have advanced since 2004). What is left are only the advantages.
Nice reading however. And thank you for pointing out that you have hardware that dispels the primary concern I would have raised, that of accuracy.
Cheers, jn
Last edited:
No, I read what you write, some of it is quite witty I have to admit, I'm not sure though where the line is between your fantastic utterances and what you believe to be facts, do you know? For instance, you appear to be saying that you can hear better transients due to a higher bandwidth, is that right?
It seems interesting, I have recorded-it and will read-it later.
This said, as i'm not a Digital gear designer, it it just for my culture and curiosity.
No myths or legends, on my side. I plug various digital gears in my systems, and listen. This kind of knowledge just have a benefit for me, to save time not listening at gears that, according to their datasheet, are obviously not performing good enough. Trying to find some shared characters between several gears sharing the same technology.
Myths or legends ? With all these wars of religion, we could start with the ones that cause the most damages. But it is repressed by the rules of the forum. ;-)
Actually, I have already stated that CD is fine for all my purposes..
Others can do the hearing. The anecdotes have led me to evaluate technical aspects, so I do not need to hear anything.
I did not need to hear anything to evaluate the magnetic effects and losses of a shorting ring structure. Didn't affect the outcome, nor did I have any preconceived notions based on hearing something.
I remain a neutral player..
jn
Scott, Let me answer you as honestly as possible.
Yes, it seems to me. In any case, the differences that I noticed between my DACs seem to go in this direction. But nothing excludes in my mind that it is the product of other phenomena that I have not analyzed correctly.
This said, fantastic is not exactly the word I would use to my way of thinking.
I am full of doubts with the most Cartesian approach possible: There is not an object around me which I do not know the principle of operation. Since I was a very little kid, when I had dismantled my parent's radio to see where the gentleman who was speaking inside was hiding, I dismantle and repair everything. Even my car.
Oh, this Wankel engine that i built at the age of 12 with cartoon paper to understand how it was working).
And if I can't, I read everything I can about it (with the same suspicion ;-)
Last edited:
- Status
- Not open for further replies.