HiRes FOMO - Am I really missing anything?

[klettermann]

In things audio, I'm usually dominated by my inner-skeptic - rarely upgrade gear, no fancy cables, etc etc. Streaming is 90% of my listening. The rest is old CDs and DAT tapes. All of that goes through a MiniDSP SHD Studio and then on to Yggy LiM. Whatever you feed the MiniDSP gets up-or-down sampled to 24/96khz.

Lately I’ve been reading more on some forums that emphasize high-rate upsampling (192 kHz+, DSD, etc.). At the same time there seems to be increasing availability of “hi-res” content on streaming services. This has triggered a strong case of sample-rate FOMO. The obvious upgrade path would be replacing the MiniDSP with something that preserves 192 kHz end-to-end. At the same time, I don’t see a clear mechanism by which moving from 96 kHz to 192 kHz would produce any audible difference at all. For that matter, 16/44.1 sounds just fine, maybe because I'm not a bat.

So the question is straightforward: Is there any credible evidence, controlled listening or otherwise, that a 96 kHz processing ceiling is audibly limiting? Or is this psychological / marketing-driven? I guess a secondary questuon is the effect of multiple resamplings going through the chain? I've always assumed not, but..... ??

Comments greatly appreciated. Thanks and cheers,

 

[NTK]

If you trust J J.

 

[somebodyelse]

The tests show that people generally can't tell the difference, with a couple of big caveats: it needs to be the same mastering, with the sample rate and bit depth converted using a resampler that doesn't have significant artefacts. Neither are necessarily true with commercially available content (streaming media, downloads or physical media), just as different releases of a CD may not have the same mastering. Unfortunately you have to check on a case by case basis to find the best version.

Edit: https://src.infinitewave.ca used to have detailed measurements of different available resamplers, but it seems not to be responding at the moment.

You should be able to carry out ABX tests yourself if you have any doubt.

 

[Pluto]

Is there any credible evidence, controlled listening or otherwise, that a 96 kHz processing ceiling is audibly limiting?

No.

The higher you take the sampling frequency, the more the Uncertainty Principle works against you on several different fronts. An optimum sampling rate would probably be around 60kHz. When you provide for an audio bandwidth of nearly 50kHz, more than half of the audio “pipe” is used to carry unproductive noise. The number of records around that have any useful musical content above [let's say] 20kHz is minute.

 

[halverhahn]

Btw: there are virtually no speakers or headphones designed to reproduce frequencies above 20 kHz accurately and linearly.

Due to non-linearities in playback devices, high sample rates and unfiltered HF contend e.g. from SACDs (by design the Players should filter>50kHz [see Scarlet Book]) can lead to distortion in the audible range.

Check out: https://people.xiph.org/~xiphmont/demo/neil-young.html

 

[krabapple]

In things audio, I'm usually dominated by my inner-skeptic - rarely upgrade gear, no fancy cables, etc etc. Streaming is 90% of my listening. The rest is old CDs and DAT tapes. All of that goes through a MiniDSP SHD Studio and then on to Yggy LiM. Whatever you feed the MiniDSP gets up-or-down sampled to 24/96khz.

Lately I’ve been reading more on some forums that emphasize high-rate upsampling (192 kHz+, DSD, etc.).

Advice: Stop doing that.

 

[jcarys]

t the same time, I don’t see a clear mechanism by which moving from 96 kHz to 192 kHz would produce any audible difference at all.

Because there is none. It is impossible to create more or new data by doubling the sampling rate. Converting from PCM to DSD or vice versa can only introduce errors. You can test right now with your own equipment. Sample an original recording at 44.1 or 48kHz versus your 96k upsample. If everything is working correctly, you shouldn't be able to hear any difference. In the old days (last century) you could avoid a brick wall filter frequency by upsampling first, but that limitation is now gone.

 

[klettermann]

Thank you my friends. You have released me and I am now free! And the xiph link is superb. live long and prosper.
Chers,

 

[halverhahn]

Thank you my friends. You have released me and I am now free! And the xiph link is superb. live long and prosper.
Chers,

There is another video from xiph relating to the article mentioned above.

 

[antcollinet]

In things audio, I'm usually dominated by my inner-skeptic - rarely upgrade gear, no fancy cables, etc etc. Streaming is 90% of my listening. The rest is old CDs and DAT tapes. All of that goes through a MiniDSP SHD Studio and then on to Yggy LiM. Whatever you feed the MiniDSP gets up-or-down sampled to 24/96khz.

Lately I’ve been reading more on some forums that emphasize high-rate upsampling (192 kHz+, DSD, etc.). At the same time there seems to be increasing availability of “hi-res” content on streaming services. This has triggered a strong case of sample-rate FOMO. The obvious upgrade path would be replacing the MiniDSP with something that preserves 192 kHz end-to-end. At the same time, I don’t see a clear mechanism by which moving from 96 kHz to 192 kHz would produce any audible difference at all. For that matter, 16/44.1 sounds just fine, maybe because I'm not a bat.

So the question is straightforward: Is there any credible evidence, controlled listening or otherwise, that a 96 kHz processing ceiling is audibly limiting? Or is this psychological / marketing-driven? I guess a secondary questuon is the effect of multiple resamplings going through the chain? I've always assumed not, but..... ??

Comments greatly appreciated. Thanks and cheers,

Here is quite a convincing take from the same guy who made the video on the previous post.

24/192 Music Downloads are Very Silly Indeed

people.xiph.org

I think it is fair to say that even if there is a an audible difference between 16/44.1 and (say) 24/96K, it is so tiny, you have to strain your hearing to detect it with fast switching between the two. It follows then that it cannot have an impact on musical enjoyment.


EDIT : Damnit - ninjad again - this time by @halverhahn

 

[levimax]

Thank you my friends. You have released me and I am now free! And the xiph link is superb. live long and prosper.
Chers,

The other thing you can try is to take one of your favorite recordings that you can verify was actually recorded and mastered at 192 Khz or higher and downsample (with dither) it to 44.1 Khz and then use something like Foobar 2000's ABX tool (level matched and blind) and see if you can tell any difference between the original and the down sampled version. For me the actual physical listening experience makes the point even better than reading about it. Being set free as you describe it is exactly how I felt after buying many high priced hi-rez downloads to make sure I wasn't missing out on something.