Shouldn't we upgrade the 20-20 audible range ?!

[krabapple]

He started with a 24/192 file and used a SourceForge Sound eXchange (SoX) SRC with linear phase filters and sloped TPDF dither to generate the 16/44.1
He seemed to be pretty conscientious about his methodology.
Arnie Kruger, no less, looked at his test and said:

Googlebot said he was a Hi Res sceptic.

googlebot attributed the artefact to filtering. Not to hearing above 20 kHz. Which is in fact a reasonable deduction, given the priors. Others posit intermodulation distortion -- a playback artifact. Also reasonable. The least reasonable, therefore requiring the most rigorous testing, is the claim that he heard content above 20kHz.

 

[ahofer]

That sounds like cherry picking.

No it’s the opposite. Don’t be the “man of one study”

Beware The Man Of One Study

Aquinas famously said: beware the man of one book. I would add: beware the man of one study. For example, take medical research. Suppose a certain drug is weakly effective against a certain disease…

slatestarcodex.com

 

[MAB]

Arnie Kruger, no less, looked at his test and said:

You posted only part of his response -- the full response is important, not just the part you cherry-picked:

"I don't see anything obvioiusly wrong with the samples.

How sure are you that the listening environment that produces positive results does not have enough IM to cause problems?

You might want to exclude nonlinear distortion in the playback system as a potential source of problems with the test by doing some listening tests with test files composed of sine waves 4 KHz apart, both > 22 KHz, at several different frequencies. IOW one test would be composed of tones at 24 and 28 KHz, another test composed of signals at 32 and 36 KHz, and another at 38 and 42 KHz. Be careful with these test tones, as it would be easy to turn them up high enough to involve large amounts of power without you hearing that they are loud. If there is excess nonlinear distoriton in your system you'll hear a 4 KHz tone.

Another, potentially safer test signal would be white or pink noise high pass filtered at 22 KHz. If there is excess nonlinear distoriton in your system you'll hear noise in the normal audio band."

And then when asked why the effect isn't audible on a pair of Canton speakers:

"The IM could be in the speakers, or a consequence of amp/speaker interfacing. IM in speakers is not uncommon. Other tests of high sample rates have been invalidated by nonlinearities in the speakers and amplifiers."

This is an extremely plausible explanation for why tests with the Canton speakers showed null. It was not followed up on in the thread.

You keep cherry-picking the parts that support your argument, not just the HA thread.

 

[dlaloum]

A great set of headphones can reveal the incredible amount of inner detail the RB standard can deliver.
OTOH headphones can never deliver the wall to wall soundstaging a properly setup speaker system can.
Damned if ya do, damned if ya don't. LOL

Headphones + HRTF DSP can provide full soundstaging...

 

[RexrothPigeon]

I also don’t understand what you are even arguing about. Is someone out there preventing you from buying or listening to higher bandwidth higher bit depth recordings? Do you want to force everyone to listen to higher bandwidth/bit depth recordings? Or are you just arguing for the sake of argument? Which is functionally trolling even if that is not your intent.

hear, hear

 

[Julf]

I'm going to make a plausible assumption here, and guess that googlebot established his hearing limit using increasing sinusoidal tones. Anyone would, wouldn't they?

Therefore he can hear above 17kHz - but not simple sine waves - only complex signals.

Incidentally, regarding the TWO changes to the samples, this all suggests that the difference is down to the bandwidth rather than the dynamic range.

You are good at making assumptions, and I am sure they are plausible to you. We are wasting our time here - you are convinced of your beliefs, and seems no amount of facts will change that.

 

[welwynnick]

googlebot attributed the artefact to filtering. Not to hearing above 20 kHz.

I read that thread several times, and I didn't see any comments by googlebot where he was explaining away what he heard as being anything other than a genuine result in favour of Hi Res. Were you referring to this comment?

That is exactly the type of pass-band artifact to be expected from low-pass filtering, but until lately I hadn't considered that to be audible, when done right.

The low pass filtering he is referring to is the digital filtering that generated the 44.1k file. He underlines this by saying:

And since that step is necessary it can also not be avoided for Redbook delivery. .......
Since both (speakers) act as a mechanical low pass, I think they just cause the same artifacts in the pass band as the digital low-passing.

Therefore his conclusion is that Hi Res audio does make an audible improvement, even when you can't hear above 17k or 20k.

 

[Julf]

I feel for the poor faithfully departed horse that is being subjected to this cruel treatment.

 

[welwynnick]

I feel for the poor faithfully departed horse that is being subjected to this cruel treatment.

The horse is alive and well, and being gently led towards the water trough. I can't make him drink, though.

 

[Julf]

The horse is alive and well, and being gently led towards the water trough. I can't make him drink, though.

Tell me, is this horse in the room with us right now?

 

[solderdude]

I read that thread several times, and I didn't see any comments by googlebot where he was explaining away what he heard as being anything other than a genuine result in favour of Hi Res. Were you referring to this comment?

The low pass filtering he is referring to is the digital filtering that generated the 44.1k file. He underlines this by saying:

Therefore his conclusion is that Hi Res audio does make an audible improvement, even when you can't hear above 17k or 20k.

His conclusion indeed.
Completely ignoring Arnie's remark at the possible cause for what he heard and did not do the proposed tests which might shed some light.
It is understandable he did not or did and did not post results for obvious reasons.
Just like you did not measure the actual output signal of your amp going into your drivers with your test signal.

Again... here too the benefit of Hi-Res was NOT proven as what was heard was not confirmed to be something else than just frequency extension.
Just like your test was not proof of your theory as you did not show the actual signal you listened to and draw conclusions on.

 

[pkane]

googlebot attributed the artefact to filtering. Not to hearing above 20 kHz. Which is in fact a reasonable deduction, given the priors. Others posit intermodulation distortion -- a playback artifact. Also reasonable. The least reasonable, therefore requiring the most rigorous testing, is the claim that he heard content above 20kHz.

IMD needs to be ruled out, but also there are pretty obvious artifacts below 20kHz in the two samples in that thread. Here's the difference between them in the upper frequencies. An averaged FFT plot can hide a lot of the details, but these become more apparent on a spectrogram. You can see that differences of up to 3dB extend down to below 16kHz, while 1-2dB differences go down to well below 10kHz in some spots:

 

[solderdude]

And the next question would be if those differences come from the selected downsampling/filter ?
This seems to point to detectable differences below 20kHz and has nothing to do with 20kHz + assuming the plot is correct.
In that case ... there goes the theory that > 20kHz is the reason for a found difference.... bummer.

 

[welwynnick]

IMD needs to be ruled out, but also there are pretty obvious artifacts below 20kHz in the two samples in that thread. Here's the difference between them in the upper frequencies. An averaged FFT plot can hide a lot of the details, but these become more apparent on a spectrogram. You can see that differences of up to 3dB extend down to below 16kHz, while 1-2dB differences go down to well below 10kHz in some spots:

Thanks, that's really useful. I'm surprised that a decent SRC should make so much difference.
Can you clarify what samples were being compared? 192k vs 44k, or 96k vs 44k? Thanks.

This seems to point to detectable differences below 20kHz and has nothing to do with 20kHz + assuming the plot is correct.

Agreed.

 

[Multicore]

I think anyone who wants to pursue DC to daylight audio reproduction should go for it. Given the givens there's little chance such activities will adversely affect the consumer audio market I'm interested in. Ok maybe a little bit seeing as hi-res audio streaming is a thing. But relative to Spotify and Apple it's a niche thing.

 

[pkane]

Thanks, that's really useful. Can you clarify what samples were being compared? 192k vs 44k, or 96k vs 44k? Thanks.

Agreed.

96k vs 44k

 

[kevinh]

My hearing goes out to ~ 13-14k , glad I don't have to worry aabout this stuff. Funny thing is I still enjoy music.

 

[krabapple]

I read that thread several times, and I didn't see any comments by googlebot where he was explaining away what he heard as being anything other than a genuine result in favour of Hi Res. Were you referring to this comment?

googlebot made a reasonable inference at the top of his HA thread:

I do not think, that I actually hear the presence or absence of stationary HF content. But what I am hearing is louder and somewhat smeared sounding transients in the Redbook version. That is exactly the type of pass-band artifact to be expected from low-pass filtering, but until lately I hadn't considered that to be audible, when done right. And since that step is necessary it can also not be avoided for Redbook delivery.

His Cantons , he says, are mimicking that filtering. He simply discounts that his multistep conversion process of his contrived 192/24 sample to 44/16 could have played a role, since it was 'done right'.

And now pkane is showing us differences in the audible range between googlebot's 96/24 and 44/16 samples.

Could anyone access the original 192/24 synthetic sample for further comparison?

 

[krabapple]

I asked welwynnick . but as yet have got no answer: who is 'setting the limits of audibility at 20Hz-20kHz? Who are you actually fighting?

 

[Mart68]

I asked welwynnick . but as yet have got no answer: who is 'setting the limits of audibility at 20Hz-20kHz? Who are you actually fighting?

I think he's fighting the people like me who reckon CD does it all and there's no point in higher bandwidth for playback.