cjacek;2923179Here you're going by numbers and specs said:
to a "consumer grade audio", is false and misleading as is the rest of your argument.
I'm sorry but that's true. Is the Revox as good as a true mastering-grade deck? If so, I suspect studios would save the money. Didn't somebody say earlier their deck managed 90dB dynamic range? That's 24dB better than the Revox.
Most pertinently, that was on a thread where the OP (a working musician) wanted a $500 total budget solution for producing CDs for audition. How does the Revox possibly fit in that scenario?
In that same thread you also touch upon the notion that there's really no need to record beyond 44.1kHz [in digital] and that going higher may not be of benefit, as far as an audible difference ..... Actually you say, "Sample rate is completely unimportant". If you like your source signal chopped up by inadequate sampling then that's fine but really, you have the wrong crowd here.
Let's discuss sample rate. You quoted earlier than a 10kHz sine wave signal was inadequately represented by a 44.1kHz sample rate
because the waveform looked choppy. Who is using their eyes to judge audio now?
C'mon 1/4 sample rate isn't even a challenging frequency. If you bothered to study, test, or listen at all, you'd realize that a 10kHz sine wave can be
perfectly represented by 44.1kHz. Let me emphasize that again: perfectly. I could send a 10kHz sine wave in and out of my converters 10 times with no degradation
that was not a result of the analog path of the converters. How do I know this? Because I have done it. And it works for 18kHz, 19kHz, with some attenuation there, entirely due to filter behavior as described in the Lavry links above.
This is because the apparent visual distortion entirely consists of frequencies above the Nyquist limit. For a 10kHz signal, the first harmonic is 20kHz. A true 10kHz square wave would have a VERY prominent 20kHz harmonic, and many, many other harmonics above that, all in a row. But there is no 20kHz harmonic in a 10kHz sine wave stored in 44.1kHz format, nor is there any upon D/A conversion.
What about harmonics above the Nyquist frequency? That's an interesting question. The first possible harmonic would be 30kHz. So I decided to see if I could measure it. I took that 44.1kHz file and converted it to 96kHz. I didn't really know what would happen, because it is possible the SRC routine would force the 44.1kHz through an anti-imaging routine before it upsampled. And I believe it did, because the resulting 30kHz harmonic was at -100dBFS. For those keeping score, that's 0.001% THD at frequency that is certainly not audible at -6dBSPL. Most people are (very) lucky to detect 0.05% THD at 85dBSPL on a 1kHz signal. Which is not to say that doesn't matter in the great scheme of things, although by the time you get from mic to speaker with any recording system or none at all, THD will be higher than that. It also matters which harmonics we are talking about; the 3rd harmonic at issue is more audible than the second.
So that is the entire extent of the "distortion" that 44.1kHz creates on a 10kHz signal.
But, as I said earlier, 1/4 sample rate is not a very tough test. So I tried something a bit harder, like 19kHz. I got about the same result with the 38kHz harmonic, but with one item of interest: a nonharmonic distortion at 25kHz and -85dBFS (0.006%). I don't have an easy answer for that; it can't be jitter as this is strictly an off-the-clock procedure.
Still, since 0dBFS signals at 19kHz are, to say the least, extremely unusual, I feel fairly confident that distortion also could not be heard.
I note a disturbing tendency here to completely reject any attempt at scientific measurement whatsoever. That is a travesty. As Lavry demonstrates, any complex waveform can be mathematically described as the sum of constituent sine waves. Therefore, it is appropriate to model real-world signals with sine waves. There is more to testing gear than single sine waves though; more complex combinations are often used to analyze characteristics like intermodulation distortion and transient response. Those test generally aren't published in specs, but designers indeed use them.
But, if you are completely unwilling to yield to science, I am willing to go out to my shed, get one of my highest pitched organ pipes (they are supposed to be installed in my studio, but time is ever short), and repeat the same experiments with the resulting real-world signal. I already know the result . . . because I've done it. But I'll post it for you at different sample rates (all upsampled to 96kHz), and let you pick which is which.
On another topic, I am also increasingly disturbed by frequent claims from tape aficionados to possess not only above-average but ultrasonic hearing. Most audiophiles are male and over 40. Recording types run a wider age range, but lately it's the oldsters claiming they have hearing equal to young children. Medically speaking, that would be very, very unusual, and the fact that such rare random occurrences of exceptional hearing would be highly correlated with a hobby group seems specious.
My cassette deck should sound much worse than CD audio, at least according to the specs but it in fact sounds better, not necessarily as the result of certain pleasing 'tape characteristics' [those too, sure] but rather that it provides a more complete body and coherent picture of the sound, which sampling cannot do....
OK, my cassette decks (Tascam CD-A500 or 424mkII, take your pick) do sound objectively and subjectively worse than CD audio to me. I can prove the objective; you already conceded that. I have no desire to prove my subjective opinion, nor disprove yours, because that is the reduction of your argument: you like tape because it sounds better to you. And I have never found fault with such statements.
.
