Audible difference 44.1kS/s vs 96kS/s? (files included)

[Audiop]

Three original clips and three downconverted (to 44.1kS/s) and upsampled (back to 96kS/s) versions. Claves, shaker and triangle.

Recorded at 30cm with a single Earthworks QTC1 into Audiofire 4, 24/96. Response of whole recording chain down less than 2dB @ 40kHz.

SRC = iZotope 64bit.

Do you hear a difference blind and with statistical significance?

SRCc.wav
SRCc_001.wav


SRCs.wav
SRCs_001.wav


SRCt.wav
SRCt_001.wav


/Peter

[diogo_c]

[Audiop]

What?


/Peter

[OMNIFEX]

Possibly it would be better if you used some type of music material instead of a sound that resembles something from a Drum Machine.

Cheers!

[Audiop]

Originally Posted by OMNIFEX

Possibly it would be better if you used some type of music material instead of a sound that resembles something from a Drum Machine.

Cheers!

This test is about hearing or not hearing a difference of high frequency content of percussive/transient instruments such as claves, shakers and triangle. The latter two being particularly rich in harmonics and high frequency content.

Nothing could be better for testing triangle than triangle.. I'm sure you can grasp that with a little effort.


/Peter

[The Listener]

I've been looking for your Line Audio preamp clips (still didn't find them, can you help with a link?) and found this interesting test.

My first impression tells me 001 is the unprocessed source file (I mostly listened to the triangle clip), but ABX software would surely beat me, so I can't give any statistically significant opinion.

[Audiop]

Isn't it strange that those that are most verbal in threads about sample rates and wordlength never show up in threads like this?


/Peter

[24-96 Mastering]

Originally Posted by Audiop

Isn't it strange that those that are most verbal in threads about sample rates and wordlength never show up in threads like this?


/Peter

I think most of them are banned...

Or are you referring to SR skeptics as well?

[Audiop]

Aha, that explains it to some degree. However there are some guys that are civil but still have strong opinions so to speak.

Oh well, maybe more members will chime in.


/Peter

[OMNIFEX]

Originally Posted by Audiop

This test is about hearing or not hearing a difference of high frequency content of percussive/transient instruments such as claves, shakers and triangle. The latter two being particularly rich in harmonics and high frequency content.

Nothing could be better for testing triangle than triangle.. I'm sure you can grasp that with a little effort.


/Peter

Fair enough. I will listen to your clips and post my thoughts later on this evening.

Cheers!

[Bob Olhsson]

Originally Posted by Audiop

This test is about hearing or not hearing a difference of high frequency content of percussive/transient instruments such as claves, shakers and triangle. The latter two being particularly rich in harmonics and high frequency content...

FWIW high frequency reproduction is not why one uses high sample rates.

[myrtlebacker]

Is there an ABX Player for Mac OS X ? I actually think I hear a difference but that's just AB without the X

[David Spearritt]

Originally Posted by Bob Olhsson

FWIW high frequency reproduction is not why one uses high sample rates.

Huh?

[AlexK]

Originally Posted by myrtlebacker

Is there an ABX Player for Mac OS X ? I actually think I hear a difference but that's just AB without the X

QuickTime 7 can do it. Open both files, and make sure the 'playback only in the frontmost player' option is ticked in QT Prefs. Push CMD + Return and all the QT players currently open will play, allowing you to switch from one to the other.

It's not sample accurate but it works well enough for me...

EDIT: forgot to add, 001 sounds like it's got more attack on the tringle, definition on the shaker and a snappier clave. But then again, it's possible it's a placebo and I've just made it up in my head. I'm also listening on my (actually pretty good) earphones, and my MacBook Pro's built in output...

[myrtlebacker]

But that's not any better than alternately soloing, which I am doing already. Which is what I meant in shorthand way of sayin AB without the X.

No I am looking for a player, to make a proper blind test.

[OMNIFEX]

Originally Posted by Audiop

Aha, that explains it to some degree. However there are some guys that are civil but still have strong opinions so to speak.

Oh well, maybe more members will chime in.


/Peter

Right!

So I listened to the track entitled “SRCc” & “SRC_001” files and, this is what I am hearing through my loudspeakers.

The attack on SRCc is more prominent in which, appears the SPL level could be higher than SRCc_001 at the given peak. The decay is longer as well. In other words, it does not fade abruptly like SRCc_001.

I am leaning more to wards 1.5 kHz however it may be 1.7 kHz where the accent is noticeable comparing both files. I could even be 2 kHz.

If I were evaluating this through my large system I could probably determine where the transient attack lies. However, the loudspeakers I’m using for the Internet comprise of some small bookshelf speakers (8-inch woofers with 2-inch tweeters). The high frequency response is 20,000 kHz so all is not lost.


Was that the type of evaluation you were looking for?

Cheers!

[myrtlebacker]

Nevermind, I actually had one downloaded called "ABXer", but forgotten about it. Old age.

ABX Test Completed: 2010-01-07 23:58:52 +0100

Number of tests performed: 20
Number of correct answers: 10
Percentage correct: 50%

[Adebar]

Originally Posted by Bob Olhsson

FWIW high frequency reproduction is not why one uses high sample rates.

Good point, that´s also my opinion.

[Audiop]

Originally Posted by Bob Olhsson

FWIW high frequency reproduction is not why one uses high sample rates.

Hi Bob.

I don't know what you want to say but as I wrote earlier:

This test is about hearing or not hearing a difference of high frequency content of percussive/transient instruments such as claves, shakers and triangle. The latter two being particularly rich in harmonics and high frequency content.

Simple as that. It's not really about sample rates but audibility of high frequency content.

I record at high sample rates myself because my converter seem to perform better at higher rates. To this day I'm not 100% sure why that is the case and I'm performing tests when I have extra time to sort this and general related questions out.

If a converter sounds better at higher rates, can you keep that quality with an optimum software SRC in post as opposed to record directly at 44.1kS/s or is it about HF content and filter limitations.


/Peter

[Audiop]

Originally Posted by myrtlebacker

Nevermind, I actually had one downloaded called "ABXer", but forgotten about it. Old age.

ABX Test Completed: 2010-01-07 23:58:52 +0100

Number of tests performed: 20
Number of correct answers: 10
Percentage correct: 50%

Thanks for posting. Do you have a link to that software? I have been searching
for ABX for Mac.


/Peter

[Audiop]

Originally Posted by OMNIFEX

Right!

So I listened to the track entitled “SRCc” & “SRC_001” files and, this is what I am hearing through my loudspeakers.

The attack on SRCc is more prominent in which, appears the SPL level could be higher than SRCc_001 at the given peak. The decay is longer as well. In other words, it does not fade abruptly like SRCc_001.

I am leaning more to wards 1.5 kHz however it may be 1.7 kHz where the accent is noticeable comparing both files. I could even be 2 kHz.

If I were evaluating this through my large system I could probably determine where the transient attack lies. However, the loudspeakers I’m using for the Internet comprise of some small bookshelf speakers (8-inch woofers with 2-inch tweeters). The high frequency response is 20,000 kHz so all is not lost.


Was that the type of evaluation you were looking for?

Cheers!

Thanks, yes!

The files are matched tightly in level according to Foobar ABX except for the traingle. The files are not altered in level since the difference should be in the upper range which is filtered out in the converted files.

Do you have the possibility to do a blind test to see if your impressions remains and have statistical significance?


/Peter

[OMNIFEX]

I would imagine the sharp attack (similar to using a gate with a fast release) would give the impression of more SPL with the file offering the higher frequency sample rate on the peak.

I'll have my girlfriend test me to see if my choices differ blindsided.

Cheers!

[ISedlacek]

Originally Posted by Audiop

Three original clips and three downconverted (to 44.1kS/s) and upsampled (back to 96kS/s) versions.

So you recorded 3 clips in 96k, then converted them to 44k and then back to 96k ?

Unless I miss something in your description, the real point that shows the difference is to record the same source once in 96k, second time in 44k (using the same recording chain). Then to convert 96k file into 44k (using high quality SRC - like r8Brain Pro) and to compare the differences in 44k a to see whether recording originally in 96k brought any sound difference comparing to recording directly in 44k. If you recorded only in 96 k (and that too using rather low end quality converters) and then did just some SRC processes, I am afraid it does not say much - maybe something about the isotope SRC algorithm.

[Yetti]

The 001 files all sound a "little" bit better to me..slightly more clear and open.... most noticable on the shaker.

[PatrikT]

Originally Posted by OMNIFEX

I am leaning more to wards 1.5 kHz however it may be 1.7 kHz where the accent is noticeable comparing both files.

Are you sure it isn't 1624 Hz?


Regards
Patrik

[neilwilkes]

Originally Posted by Audiop

I record at high sample rates myself because my converter seem to perform better at higher rates. To this day I'm not 100% sure why that is the case and I'm performing tests when I have extra time to sort this and general related questions out.

If a converter sounds better at higher rates, can you keep that quality with an optimum software SRC in post as opposed to record directly at 44.1kS/s or is it about HF content and filter limitations.


/Peter

It's cheaper to design & build a converter to sound good at 88.2/96 than it is to make it sound equally as good at 48/44.1.
Bob Katz explains it much better than I can in the 2nd edition of his excellent book
Bob Katz - Mastering Audio
Symptoms can include any or all of these:
higher distortion, ripple, ringing (*) and aliasing.
* = this one is very common in poorly designed Linear Phase SRC too

As far as actual HF content goes, can you honestly hear above 22kHz?

[OMNIFEX]

Originally Posted by PatrikT

Are you sure it isn't 1624 Hz?


Regards
Patrik

With the correct Q anything is possible.


Cheers!

[Key]

Originally Posted by neilwilkes

It's cheaper to design & build a converter to sound good at 88.2/96 than it is to make it sound equally as good at 48/44.1.
Bob Katz explains it much better than I can in the 2nd edition of his excellent book
Bob Katz - Mastering Audio
Symptoms can include any or all of these:
higher distortion, ripple, ringing (*) and aliasing.
* = this one is very common in poorly designed Linear Phase SRC too

As far as actual HF content goes, can you honestly hear above 22kHz?

haha no but I just check recently and I can still hear 21kHz.

I downloaded the files but I would actually prefer a whole song to increase the odds of audible artifacts. And it makes the listening a little less like a Chinese water torture test for your ears.

Do all linear phase uneven SRCs have measurable ringing? I think someone told me asynchronous doesn't but anyway is it so much that they are poorly designed or that is just the nature of the beast?

[dcollins]

Originally Posted by Key

Do all linear phase uneven SRCs have measurable ringing?

Sure.

I think someone told me asynchronous doesn't but anyway is it so much that they are poorly designed or that is just the nature of the beast?

That would be wrong. Asynchronous and synchronous refer to the treatment of the clocks and has nothing to do with the filter ripple.


DC

[Key]

Originally Posted by dcollins

That would be wrong. Asynchronous and synchronous refer to the treatment of the clocks and has nothing to do with the filter ripple.
DC

I think I am mixing up or basically don't understand the differences between asynchronous sample rate conversion and a digital to analog to digital conversion with asynchronous clocks on the ADC and DAC.