HiRes makes no sense?

[kyuu]

It beggars belief that people think that...

... is content that's worthwhile to reproduce.

 

[vintologi]

I could download the two tracks, and analyzed them by MusicScope 2.1.0 and ADOBE Audition 3.0.1.
View attachment 548906

I also did simple listening comparison, not in strict blind ABX test, using my rather heavy duty multichannel multi-amplifier audio setup (ref. #931 and #1,009 on my project thread), of course in exactly the same volume/gain; I heard no difference at all between the two tracks.

Based on the spectrals it doesn't actually look like a suitable test for comparing sampling rates so it's surprising how i found it relatively easy to pick the tracks apart (or at least easier than it should have been). I could try to redo the test with my new speakers but my hearing is worse now so i could potentially fail (i did a quick semi-blind test earlier and found the same thing with one of my speakers replaced with a better one).

Perhaps my hearing simply extended far enough so i was able to hear the tonality impact of the SoX filter used by audacity directly but i sadly don't hear that well anymore so that might not be repeatable if that was the cause.

Note that you are not meant to disclose which tracks are the original quality so if you are going to analyze more of these please do not show which is which.

My current tweeters (TW29BN-B) are definitely more suited for testing these things than your tweeters but i don't know how good the frequency response of my amplifier is (it needs to be replaced but i first need to figure out what to buy).

 

[vintologi]

It beggars belief that people think that...

View attachment 548931

... is content that's worthwhile to reproduce.

24-96 did sound less pleasant than 16-44.1 with that track (statistically significant p = 0.01372).

It's actually noise above 20Khz.

There are tracks with way better spectrals such as

 

[Cote Dazur]

Then why can't I hear the sound reproduced the same way on the Red Book layer of the SACD?

That is the really interesting question. If you take the time to read the many post and threads here at ASR, you will have the answer. It takes efforts be the rewards are great.
In a nut shell, the answer is, becasue your brain wants you to.
Until you prove it to your self, we could be millions telling you here what is true and what is illusion, you will not believe because your audiophile ears know better.

 

[vintologi]

Then why can't I hear the sound reproduced the same way on the Red Book layer of the SACD?

The Red Book layer might have different mastering.

Might be placebo.

Or you are actually hearing a difference due to the format itself.

I suggest further testing to see which was the cause.

 

[vintologi]

Hmm i did a quick test where i compared 24-96 vs 24-54 and i had a hard time identifying a difference to reliably tell them apart. I might try listening some more but i think i want to replace my amplifier for this so i actually reproduce frequencies up to 48 Khz.

MediaFire

MediaFire is a simple to use free service that lets you put all your photos, documents, music, and video in a single place so you can access them anywhere and share them everywhere.

www.mediafire.com

View attachment 548899

I did continue this today after getting some sleep and got to the point where if i was wrong wrong i would just abandon this blind-test and try something else.

So i checked the tracks and turns out i was right. I had correctly identified which track had the higher sampling rate (24-96 sounded better than 24-54). I will however need at least 6 more correct to get to decent statistical significance. This is obviously not going to be easy (FLAC vs ogg vorbis Q100 is easy in comparison)

1.12 GB folder on MEGA

15 files and 4 subfolders

mega.nz

Note that what is A and B was randomized for each test. Do not disclose which is which is you post spectrals.

Ill add 24-44.1 as well for people to test.

I do think you should take all the time you need with these blind-tests and not guess prematurely since a single wrong guess will result in it taking significantly longer to get to statistical significance.

 

[danadam]

I do think you should take all the time you need with these blind-tests and not guess prematurely since a single wrong guess will result in it taking significantly longer to get to statistical significance.

What do you mean "to get to"? AFAIK the number of trials should be established before the test starts. And then you either reach the statistical significance in the test or not.

 

[vintologi]

What do you mean "to get to"? AFAIK the number of trials should be established before the test starts. And then you either reach the statistical significance in the test or not.

You don't need to fix the number of trials at the start. In this case however i don't think it makes sense to do more than 10 trials given that i should replace my amp for this anyway.

I probably end it early if i get more than 1 wrong though (since 8/10 wouldn't be statistically significant anyway).

 

[danadam]

You don't need to fix the number of trials at the start. In this case however i don't think it makes sense to do more than 10 trials given that i should replace my amp for this anyway.

I probably end it early if i get more than 1 wrong though (since 8/10 wouldn't be statistically significant anyway).

Sounds like you trying many tests until, possibly by chance, you get 9/10. But that's cherry picking.

 

[vintologi]

Sounds like you trying many tests until, possibly by chance, you get 9/10. But that's cherry picking.

That's not the case.

This is my first proper blind test with my current speakers and i don't plan on doing more until i have replaced the amplifier.

So far 2 correct and 0 wrong.

Every attempt at a blind test will be documented from now.

I did some blind-testing with my old speakers but less than 10 proper tests in total. Of these i got to a statistically significant result 2 times

Test0: i preferred 192 KS/s to 48 KS/s 9 of 10 times (p = 0.02148)
Test1: i preferred 16-44.1 to 24-96 30/43 times (p = 0.01372)

Edit: Test aborted after one wrong guess.

There is a potentially revealing segment at 3:24 but i found it to be rather inconsistent (it did initially point me in the right direction but the rest of the listening indicated to me otherwise). Problem is that there could simply be audible effect from IMD due to the ultrasonics for the tracks being quite intense).

1.12 GB folder on MEGA

15 files and 4 subfolders

mega.nz

Notice that that here there isn't much to see at 3:24 but the spectral analysis did show ultrasonics there

 

[vintologi]

Something to keep in mind is that some people might simply be able to hear a difference between say 24-96 and 24-44.1 due to having very good hearing to the point where they hear the tonality effect of the filter.

Even with audacity resampling (down to 44.1 and up again to 96) there was still effect above 17Khz

And of course in the real world the filters can affect lower frequency a lot more than above due to being less sharp (some audiofiles/audiofools actually think that is better).

1.12 GB folder on MEGA

15 files and 4 subfolders

mega.nz

 

[RexrothPigeon]

Something to keep in mind is that some people might simply be able to hear a difference between say 24-96 and 24-44.1 due to having very good hearing to the point where they hear the tonality effect of the filter.

the word might is doing some real heavy lifting here

 

[vintologi]

the word might is doing some real heavy lifting here

I remember amir talking about passing an ABX test with regard to a DAC by listening for the effect of the filter but that was a very old model.

If we take a look at this data it seems plausible that some people would be able to pass ABX test by hearing the ultrasonics directly with music for some cherrypicked tracks (maybe 1 in 1000 people can do that)

Hearing thresholds for pure tones above 16kHz

Hearing thresholds for pure tones between 16 and 30kHz were measured by an adaptive method. The maximum presentation level at the entrance of the outer ear was

pubs.aip.org

But the real thing i would like to know is if ultrasonics that are otherwise inaudible can make an audible difference when it's mixed in with other frequencies (such as when listening to acoustic guitar music). I do have good enough tweeters to test that but i might need to replace my amplifier (i had trouble differentiating 24-96 from 24-54 due to lacking ears/amplifier).

Something i might try is to place 2 speakers close to each other and then only use one for the ultrasonics. That's fairly easy to implement if the signal is in mono.

 

[RexrothPigeon]

by hearing the ultrasonics directly

Possibly some attentive nine-year-olds, yeah

 

[kyuu]

Possibly some attentive nine-year-olds, yeah

Even if those attentive nine-year-olds could detect a difference, there would still be the question of if they preferred the super high-frequency content. I suspect that sounds that high in frequency are not terribly pleasant to behold, but as someone whose hearing cuts off at around 15kHz these days what do I know.

 

[RexrothPigeon]

there would still be the question of if they preferred the super high-frequency content.

I for one would question the judgment of anyone who'd care what kind of sounds nine-year-olds prefer

 

[kyuu]

I for one would question the judgment of anyone who'd care what kind of sounds nine-year-olds prefer

Well if they are the only ones who can detect it there's no one else to ask

 

[RexrothPigeon]

Well if they are the only ones who can detect it there's no one else to ask

It sounds to me like we are once again triangulating on "those frequencies heard only by bats and some kids don't matter"

 

[kyuu]

It sounds to me like we are once again triangulating on "those frequencies heard only by bats and some kids don't matter"

Hey, bats, dogs, and toddlers have opinions too! Who are we to deny them the dulcet tones of 25kHz noise when we play our tunes?!

 

[RexrothPigeon]

Hey, bats, dogs, and toddlers have opinions too! Who are we to deny them the dulcet tones of 25kHz noise when we play our tunes?!

well, none of them are ever in my house whether or not I'm listening to music, but I won't judge if you let them in yours