Focus on 24/192 Misguided?.....

Al, What sounds inconceivable to me is that 24/192 recording supposed to gain sound quality by downsampling it to 16/44 to be upsampled again, perhaps to the same 24/192. Am I reading it right? Is downsampling + upsampling somehow improving sound by replacing real samples with artificial interpolated samples and recreating same harmful 192kHz?

Hi Kijanki,

The only reference to downsampling + upsampling that I recall seeing was in the paragraph headed "clipping" in the lower third of the page, and in footnote 21. He was saying that by taking 192 kHz source material, downsampling it, and then upsampling back to 192 kHz, a sonic comparison could be made between the two 192 kHz signals that would be indicative of the adequacy of the lower sample rate.

Not sure if that is what you are referring to. But in any event the methodology he is describing doesn't make sense to me, because the comparison would not reflect the effects of the sharper anti-aliasing and reconstruction filters that would be required for recording and playback at the lower sample rate.

Best regards,
-- Al

Al, On one hand he has whole chapter titled "192kHz considered harmful" describing harm that 192kHz can do to amplifiers and speakers to later say this about oversampling:

"This means we can use low rate 44.1kHz or 48kHz audio with all the fidelity benefits of 192kHz or higher sampling (smooth frequency response, low aliasing) and none of the drawbacks (ultrasonics that cause intermodulation distortion, wasted space)."

According to above when DAC is exposed to 192kHz sample rate from CD it is harmful to all analog circuitry afterwards (including power amp and speaker) but when DAC upsamples redbook CD (24/192 downmixed to 16/44), the same 24/192 becomes benign. In either case DAC outputs samples at 192kHz rate but in one case it is less harmful to analog circuitry?

Ok, I see what you are referring to. Note this statement:

Oversampling is simple and clever. You may recall from my A Digital Media Primer for Geeks that high sampling rates provide a great deal more space between the highest frequency audio we care about (20kHz) and the Nyquist frequency (half the sampling rate). This allows for simpler, smoother, more reliable analog anti-aliasing filters, and thus higher fidelity. This extra space between is [sic] 20kHz and the Nyquist frequency is essentially just spectral padding for the analog filter.

Because digital filters have few of the practical limitations of an analog filter, we can complete the anti-aliasing process with greater efficiency and precision digitally. The very high rate raw digital signal passes through a digital anti-aliasing filter, which has no trouble fitting a transition band into a tight space. After this further digital anti-aliasing, the extra padding samples are simply thrown away. Oversampled playback approximately works in reverse.

So what distinguishes the two situations you are referring to is that the 192 kHz hi rez format will presumably include a significant amount of ultrasonic audio information, which is what he is saying might have harmful consequences as a result of intermodulation effects in downstream components, while the oversampled redbook data will not include that information, and therefore those effects will not occur.

Best regards,
-- Al

04-20-12: Kijanki
Al, On one hand he has whole chapter titled "192kHz considered harmful" describing harm that 192kHz can do to amplifiers and speakers to later say this about oversampling:......

Kijanki, all that the author is saying is that when sampling at higher freq like 96KHz or 192KHz, you get intermodulation products that fold down into the 20-20K audio band due to typical preamp, power amp bandwidth limitations of not being able to reproduce higher freq products distortion-free i.e. due to the non-linearities of the electronics. And, systems having smaller bandwidths have the situation worse in that the probability that they'll amplify the high freq signals is much higher. So, the point is that if you do not sample at 96K or 192K you won't have these higher freq intermod products, they won't fold down into 20-20K & your preamp/power amp will not amplify them due to its non-linearity.

it's clear to see that if a 96K or 192K sampled signal is downsampled to 48KHz then the anti-aliasing filter will cut off all these high freq intermod products. So, according to the author, since this signal is free of any ultrasonic content, it's safer to playback with the idea that distortion products due to ultrasonics are not being played back.

I do not think that it's unreasonable to say that ultrasonics created due to higher freq sampling can create in-band intermod products that can be amplified by the non-linearities of the playback electronics & that they are harmful to the playback listening pleasure.

I don't think that the author should have labeled the paragraph as "192KHz considered harmful". People like Kijanki have read this literally thinking that the very act of sampling at 192KHz is harmful. No, I don't think that the very act is harmful; it's those ultrasonics folded down & amplified that are harmful.....

Suppose we want to compare the fidelity of 48kHz sampling to a 192kHz source sample. A typical way is to downsample from 192kHz to 48kHz, upsample it back to 192kHz, and then compare it to the original 192kHz sample in an ABX test [21].

Al, Kijanki: I *think* that I might know what the author is intending to say here: To do an A/B comparison, the author would like to level the playing field. Thus, he does not want to use the original 192KHz signal as-is. What he wants to do is downsample on-the-fly the 192KHz signal to 48KHz & create signal A. Then, upsample this 48KHz signal on-the-fly back upto 192KHz & create signal B. Now, the playing field is level because the same machine downsampled & upsampled the signal & the same filters have screwed up the A & B signals. The signal X is the original 192KHz signal. If you were to use the original 192KHz which was created on some different machine against the 48KHz created on your CDP, you would have the effect of 2 different digital filters & you could not do a true A/B comparison. Does this make sense guys?