Can 176.4kHz be beneficial after all? (humans' hearing latency limits)

[PHD]

I found this and other sources that claim that the human ear could be trained to detect audible time differences even below 30us

Binaural_Hearing

So, let's say some people could detect down to 15us time differences between two sounds, accounting for the Nyquist-Shannon sampling theorem, this translates to 2/15e-6 = 133.34kHz. Therefore, the next standard audio sampling frequency would be 2*CDDA = 4*44.1= 176.4kHz. Also, some humans could hear a dynamic range of up to 120dB, which could also justify 24bit resolutions (effective dynamic range, accounting for a 20dB noise floor).

Does this prove the possible benefits of high-res audio?

I wish the Redbook standard had been set at 24bit/176.4kHz instead of 16bit/44.1kHz. All they had to do was delay the CD format release by 10 years until DVD media became available, which could easily hold full albums at that rate.

 

[DeLub]

Does this prove the possible benefits of high-res audio?

No.

 

[JayGilb]

The page is referencing sound localization, not high frequency hearing limits in human beings.

 

[Blumlein 88]

What you have missed is timing of the PCM stream is not limited to the time between samples. Subsample timing accuracy is much lower. Even redbook CD at 44.1 khz can accurately time between the two channels below 100 picosecond accuracy. Orders of magnitude below 15 usec.

Time resolution of digital audio

A frequent claim by detractors of digital audio is that the time resolution is equal to the sampling interval, 22.7 μs for the CD format. This is incorrect. Although there is a limit, it is much sm…

troll-audio.com

Maybe watch this too.

 

[Ze Frog]

I've never found CD's to be lacking anything myself, certainly nothing related to the format at least. Personally I think the whole Hi Res thing is just money making thing for the industry. I dabbled with streaming a while back, was no real difference besides maybe the version of the studio recording used for the stream or CD. You can get various albums on CD that are the same but be slightly differing in quality depending on release year etc. Streaming is for two things, firstly convenience, secondly to sucker in people to pay more for a higher tier in the belief they will gain some kind of improvement.

Maybe my ears just aren't super duper, but that's my take on it.

 

[DVDdoug]

So, let's say some people could detect down to 15us time differences between two sounds,

"Some people" make lots of claims.

Also, some humans could hear a dynamic range of up to 120dB, which could also justify 24bit resolutions (effective dynamic range, accounting for a 20dB noise floor).

Perhaps. But most people can't hear the difference between high-resolution audio and a copy down-sampled to "CD quality" in a proper, scientific, blind, ABX Test. Often, people can't hear the difference between high-resolution and a good quality MP3 (in a blind listening test).

The acoustic noise floor is never near 0dB SPL (unless you're in an anechoic chamber) and there's really nothing "musically interesting" around 0dB. And if you're listening to an average level of 120dB you're going to get some temporary (hopefully temporary) hearing loss. ...Your hearing might not be affected with a short-term 120dB peak.

[/quote]
All they had to do was delay the CD format release by 10 years until DVD media became available, which could easily hold full albums at that rate.
[/quote]
I wouldn't have wanted to wait 10 years to move-on from viny!!! A lot of money would have been lost by music companies, musicians, and hardware manufacturers. Would you expect them to wait-around for 10 years? DVD-Audio and SACD were introduced, and now Blu-Ray, but consumers aren't that interested... It seems that CD and iTunes AAC (and now streaming, even lossy streaming) is good enough for the market.

 

[NTK]

Please also see this thread started by JJ to refute this myth/misunderstanding.

Time resolution of Redbook (16/44) PCM

Ok, that old BS about "only one sample time resolution" came up again elsewhere. I made the following demonstration to put the screws to that. I generated a 10kHz sine wave at 44.1 kHz sampling rate. The matlab file is attached. I generated 128 discrete phases uniformly split over one cycle...

www.audiosciencereview.com

 

[PHD]

I'm aware that the page discusses sound localization and binaural hearing. I'm also aware that even at 44.1kHz, it is still possible to reproduce left and right waveforms that may be time-shifted by less than the sampling period (post-DAC and filter).

But, could a limited sampling rate capture time differences that are shorter than the sampling period? I don't see how is that possible unless everyone here agrees that audio recordings and audio processing should be at higher than Redbook CD specs...

 

[Blumlein 88]

I'm aware that the page discusses sound localization and binaural hearing. I'm also aware that even at 44.1kHz, it is still possible to reproduce left and right waveforms that may be time-shifted by less than the sampling period (post-DAC and filter).

But, could a limited sampling rate capture time differences that are shorter than the sampling period? I don't see how is that possible unless everyone here agrees that audio recordings and audio processing should be at higher than Redbook CD specs...

I don't know the best way to make it clear, but yes. Even with a single mono channel, differences in when a sound shifts in time are captured with accurate timing well, well below one sample period.

Maybe think about this. If you sample a sound and get one set of sample values for a few seconds. If you then sample a sound which you have shifted by much less than one sample it will result in the sample values all being different. If it were a 1 khz tone, it will reproduce a clean 1 khz tone either way, but the sample values for each and every sample will be different due to the time shift even if the time shift was less than 1 microsecond.

 

[danadam]

I'm also aware that even at 44.1kHz, it is still possible to reproduce left and right waveforms that may be time-shifted by less than the sampling period (post-DAC and filter).

But, could a limited sampling rate capture time differences that are shorter than the sampling period?

You'll have to explain what you mean by that, because as it is, it sounds contradictory to me: you are aware it is possible but you are asking if it can.

 

[danadam]

So, let's say some people could detect down to 15us time differences between two sounds, accounting for the Nyquist-Shannon sampling theorem, this translates to 2/15e-6 = 133.34kHz.

To me it doesn't make sense at all to translate inter-aural time differences into frequency. You are not the first one to do this but I don't know where people get this idea from.

 

[Multicore]

High sample rates can afford the advantage of lower latency in real-time systems but here at ASR we are principally concerned with distribution formats for and playback of finished audio products in which those small latency differences afford no advantage.