But since we can’t hear anything over 20kHz or so, none of this matters, and hi-res is just another source of angst.
1 Like
First let me say I agree with most of your sentiment here - so much so that I don’t bother with anything more than 16/44.1 anymore. BUT if you are buying snake oil you want to make sure you are getting the snake oil you paid for!
1 Like
True. That’s why I said selling up-scaled tracks would be fraud. The least you can do is sell authentic ultrasonic processing artifacts.
1 Like
Exactly right! This : “authentic ultrasonic processing artifacts.” was legitimately hilarious.
not that this hasn’t been discussed a billion times elsewhere, but people do feel SACD’s sound better than CDs…
If that’s honestly, reliably, statistically true, do we think it’s because of increased resolution/dynamic range? Or is it a PCM vs DSD thing?
Yes, it was discussed a billion times, but again, unless those statistics were collected during double-blind tests, the only conclusion is that statistically, people expect SACD to sound better. What’s funny is that SACD is DSD64, which people feel it’s crappy when compared to DSD256 or more. One can only wonder how many veils can possibly be removed… But I won’t go down that path here.
2 Likes
People who can afford the infamous "sufficiently resolving equipment’ generally are old enough that they do not hear anything anywhere near 20KHz.
Having said that, some high rez recordings might sound better because they have been mastered better. Of course, these are different masters and the comparison is pretty much useless, but it works for the “we don’t need no blind tests” crowd fine.
1 Like
My lack of SRE is the main reason that my ears don’t bleed when I listen to MP3s. 
OTOH, the standard audiology tests based on continuous tones may not be that informative about overall perceptual performance with respect to music. This overall argument is silly from both sides: the “objectivists” seem to assume that standard frequency-domain measurements are all there’s to know, the “subjectivists” fail at basic experimental design. In the meanwhile, the most basic part of it all, encouraging the production of quality recorded music no matter what sample rate, is ignored in a race to the bottom.
I wasn’t talking talking audiology tests, I was talking about double-blind ABX testing, which use music tracks of various genres, not test tones. Based on those, most people, including people with musical training, can’t tell properly encoded MP3 from lossless CD or hi-res.
What was the original question again?
I know some of the people who invented perceptual audio coding, we were colleagues at Bell Labs. They ran lots of those tests back then. Not every subject, not every track, but enough cases to get a statistical difference. Heck, over a decade ago I inadvertently had the same tracks of a Bill Frisell album back to back in AAC and lossless, my wife walked in and immediately said “this version sounds better,” which was the lossless one.
1 Like
You guys are funny.
There is a very good reason why high resolution recordings can sound better than Red Book and there is a reason that DSD128 and DSD256 can sound better than DSD64. It all has to do with the reconstruction filters and noise shaping. Using the higher sampling rates allows the filters to move the artifacts and noise outside of the frequency range humans can hear. The higher the sampling rate the higher you can move the noise. With Red Book and DSD64 not upsampled and noise shaped, that noise is the audible range.
Now, if you are never using a NOS DAC and the DAC you do use upsamples AND has good reconstruction filters and noise shapers, you don’t have to worry too much about Red Book and DSD64.
1 Like
DSD64 has a Nyquist limit of 1.4MHz (with an M), so there’s plenty of room to push quantization noise out. But I won’t go there again, we’d be going in circles, as usual.
Thread question:
What’s the diff between a 192/24 track in tidal/qobuz vs a 44/16 track upsampled to 192/24 in Roon?
My attempt at an answer:
A track from Qobuz or Tidal at 192/24, if recorded and mastered as such, will have more information (not necessarily audible) but gives what is technically audible more room to breathe.
A track at 44.1/16 upsampled to 192/24, in my eyes/brain makes little sense as you can’t add information that isn’t there, but could help a DAC keep all the original 44.1/16 data without imperfections.
Upsampling does have benefits though, such as applying DSP.
My views only and may technically be incorrect 
1 Like
44.1KHz done right sounds amazing. But it does have digital artifacts in the audible spectrum at fractions of 44.1KHz, such as 22.05KHz and 11.025KHz. By upsampling from 44.1KHz to 88.2KHz or 176.4KHz you move those digital artifacts above the audible spectrum and you get a subtle improvement in smoothness and a bit more harmonic coherency.
Torben
1 Like
11.025 is also a fraction of 88.2 and 176.4, isn’t it? There’s no reason to have artifacts there with proper ADC and decimation filters at any sample rate.
If the DAC uses statistical error correction, than insane up sampling can sound better and if the DAC does not use statistical error correction, than insane up sampling can sound worse.
Torben
I have found that the care taken in the recording and mastering will make more of a difference than the resolution it was recorded in.
5 Likes
A quote from the NativeDSD web site:
“For DSD64, the uncorrelated modulation noise is about -110dB at 20KHz, rising to about -50dB at 100KHz . For DSD512, the modulation noise is about -110dB at 160KHz, and -50dB at 800KHz.”
1 Like