HiRes makes no sense?

[danadam]

by hearing the ultrasonics directly with music for some cherrypicked tracks

What music has isolated ultrasonic tones, that are intended and not some random constant tone from a monitor, fire alarm or similar?

 

[vintologi]

What music has isolated ultrasonic tones, that are intended and not some random constant tone from a monitor, fire alarm or similar?

It's fairly common for recordings at 24-88.1 and higher to contain ultrasonics from the music itself.

Here is a rare example of ultrasonics extending past 50Khz:

Here there are ultrasonic spikes where you would expect them during guitar plays:

1.12 GB folder on MEGA

15 files and 4 subfolders

mega.nz

I think i should cut out a short section at around 3:24 to see i am able to differentiate 24-96 from 24-54 with that (seemed like that based on a few blind sessions but nothing statistically significant as of now so could be a coincidence).

 

[danadam]

It's fairly common for recordings at 24-88.1 and higher to contain ultrasonics from the music itself.

But that's not what I asked for and not what that research was investigating. They only tested tones in isolation. I don't think you can conclude from that, that the results would be the same if the ultrasonics were accompanied by a harmonically related content at lower frequencies.

And your music examples also show the decrease in level as the frequency increases, which is typical for music. What the total SPL is going to be, if you set the volume so that the ultrasonic content is at 90 dB SPL or more?

 

[SSS]

It's fairly common for recordings at 24-88.1 and higher to contain ultrasonics from the music itself.

Here is a rare example of ultrasonics extending past 50Khz:

View attachment 549314

Here there are ultrasonic spikes where you would expect them during guitar plays:

View attachment 549313

1.12 GB folder on MEGA

15 files and 4 subfolders

mega.nz

I think i should cut out a short section at around 3:24 to see i am able to differentiate 24-96 from 24-54 with that (seemed like that based on a few blind sessions but nothing statistically significant as of now so could be a coincidence).

Interesting files. If the guitar was picked up with a ceramic sensor then frequencies over 20 kHz are possible. And when the mixing desk goes up to 50 or more kHz then it is in the 24/96 wav file. Will try to record my guitar and look for this.

 

[vintologi]

Interesting files. If the guitar was picked up with a ceramic sensor then frequencies over 20 kHz are possible. And when the mixing desk goes up to 50 or more kHz then it is in the 24/96 wav file. Will try to record my guitar and look for this.

I uploaded the first track to this folder:

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

All except the organ one has good spectrals with music tones extending past 50Khz. Found them in a compilation elsewhere.

 

[SSS]

I uploaded the first track to this folder:

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

All except the organ one has good spectrals with music tones extending past 50Khz. Found them in a compilation elsewhere.

Did test my electric guitar with magnetic pickups. Although mag pickup is limited in high frequencies I could measure spikes up to 50 kHz at about -70 dB below the main tone which was around 1.5 kHz. I am sure with ceramic pickup and steel strings the high frequency part would be more. But a other question is whether this has a sound influence for me since I hear only up to 12 kHz due to may progressed age.

 

[PristineSound]

Using headphones, and once you figured out what artifacts to look for, most people can pass an ABX between HiRes vs CD quality. Including me. It's actually not that hard at all, assuming you use a track where artifacts exist between hires and CD quality.

But most people cannot tell hires from CD quality without an ABX, otherwise without comparing. I can't even tell MP3 from HiRes without comparison.

I am fully expecting I get daggers for saying this.

 

[voodooless]

assuming you use a track where artifacts exist between hires and CD quality

But then the conversation just isn’t good, isn’t it? Not the fault of the sampling rate.

 

[antcollinet]

I guess i will be able to get statistical significance for 24-96 vs 24-48 in a track i cherrypick specifically for it at least (i use deadbeef for XY testing).

I would be massively surprised. It would imply you have superhuman hearing.

 

[HalSF]

I enjoy the probably mostly placebo-based pleasures of the umpteen 24/96 etc. files and hi-res streaming on Apple Music and Apple Classical that I listen to. And I have some fantastic-sounding SACDs.

You know what doesn't make sense? Fretting about hi-res vs. Red Book like straining after a gnat, or worrying about the drive space that the bigger files use in 2026, with storage cheap and plentiful, and wi-fi speeds and DACs handling hi-res with ease.

I don't seek ou hi-res, but I exert zero energy eschewing it when it comes along for the low, low price of absolutely free.

 

[antcollinet]

Edit: Test aborted after one wrong guess.

That is the very definition of cherry picking. If you keep on restarting the test every time you get one wrong, then eventually you get ten in succession right just by blind chance. That 0.05 p value says it will happen once in every 20 trials (roughly)

 

[vintologi]

That is the very definition of cherry picking. If you keep on restarting the test every time you get one wrong, then eventually you get ten in succession right just by blind chance. That 0.05 p value says it will happen once in every 20 trials (roughly)

P = 0.05 isn't particularly good to begin with.

I did try some other stuff but i never got to the point where i made a guess due to having trouble hearing a difference (i did hear differences but it was very inconsistent even when playing the same version more than once).

I think i want a better amplifier that extends past 50Khz before i continue.

 

[levimax]

Using headphones, and once you figured out what artifacts to look for, most people can pass an ABX between HiRes vs CD quality. Including me. It's actually not that hard at all, assuming you use a track where artifacts exist between hires and CD quality.

Can you describe the artfacts you listen for? I have tried taking a known good quality hi-res track and then down sampling it to 44.1 /16 and then try to ABX between the 2 and have never been able to hear any difference except if I use headphones and turn up the volume load enough on a section before the music starts playing and listen for hiss which is louder on the 44.1 track .... but that is cheating

 

[LTig]

P = 0.05 isn't particularly good to begin with.

I did try some other stuff but i never got to the point where i made a guess due to having trouble hearing a difference (i did hear differences but it was very inconsistent even when playing the same version more than once).

I think i want a better amplifier that extends past 50Khz before i continue.

... and a similar wideband transducer...

 

[PristineSound]

Can you describe the artfacts you listen for? I have tried taking a known good quality hi-res track and then down sampling it to 44.1 /16 and then try to ABX between the 2 and have never been able to hear any difference except if I use headphones and turn up the volume load enough on a section before the music starts playing and listen for hiss which is louder on the 44.1 track .... but that is cheating

It's been a while, but decay of vocal and echo is one thing I look for. One sounds a bit more "hollow" if I recalled.

Do this. Take a 24 bit track down sample it to CD quality, use foobar ABX and sweep it second by second, you will likely find something, but some tracks are harder than others.

 

[antcollinet]

It's been a while, but decay of vocal and echo is one thing I look for. One sounds a bit more "hollow" if I recalled.

Do this. Take a 24 bit track down sample it to CD quality, use foobar ABX and sweep it second by second, you will likely find something, but some tracks are harder than others.

You should also re-upsample the cd quality to 24 bit again to match the comparator file - or you might be hearing differences in file processing in the DAC rather than differences in the format itself. Needless to say the down conversion should be using appropriate fast/sharp filtering.

 

[vintologi]

... and a similar wideband transducer...

I might already have the best tweeter on the market for this.

Main issue is my lack of training. I definitely have natural talent for these things (enough to differentiate between say flac and ogg vorbis Q100 without any training) but i might still need training to say hear the difference in impulse response from the higher sampling rate (if that is audible to begin with that is).

 

[SSS]

I might already have the best tweeter on the market for this.

View attachment 553254

Main issue is my lack of training. I definitely have natural talent for these things (enough to differentiate between say flac and ogg vorbis Q100 without any training) but i might still need training to say hear the difference in impulse response from the higher sampling rate (if that is audible to begin with that is).

You are lucky to be in a young age to hear the differences. At my 70+ age I can't anymore.

 

[Proud Audiophile]

I thought that the relevant science says that a person can not distinguish between hi res and red book? I can't. Mastering quality massively overwhelms any difference between formats as best i can tell. Is that the science? So why are you scientists not concluding that your belief in perceived differences are the result of bias? Am i wrong? Do carefully controlled studies demonstrate an auditable difference between the formats or not?

 

[fpitas]

Do carefully controlled studies demonstrate an auditable difference between the formats or not?

Not to my knowledge, no. Higher res is appropriate for studio manipulation. But it will never go away as long as someone can make money on it.