frequency response vs time

[Bending]

I recently found two short articles on AudioXpress https://audioxpress.com/article/hearing-and-audio-part-1-frequency-phase-and-time/ and https://audioxpress.com/article/hearing-and-audio-part-1-frequency-phase-and-time/ by Milind Kunchur (please don't disregard these because you find his cable papers flawed and most of it is not based on his research anyhow). Time resolution is for a long time known to play an essential part of our auditory system especially for sound location. So even one loose high frequency perception over time the time resolution stays in the µsec range. There are some short sample

and

which I find quite interesting demonstrating the audibility of time differences in audio reproduction.

Any standard audio equipment measuring equipment has issues with that, most step response is around a factor 100 less. I wonder how the relevant time resolution of equipment can be measured including in multi driver loudspeakers or anything related to DSP which could cause time shifts (especially if they are frequency related

 

[Talisman]

 

[ZolaIII]

It's time and space in this universe. Measurements are taken FIR based with resolution (1~2M taps) and accustic time reference and to how we psy hear it depending of SPL. What's considered depth of the field is back to front refractions ratio along with improving RT time decay to a goal and what's objective achievable in given space. In sub bass our psy hearing perception makes it lag a lot and in general when ever it does we use boost in frequency domain to compensate for it to given SPL. That's equal loudness compensation (ELC).

 

[Curvature]

He mixes true things established by research, with loose remarks and stupid comments about digital audio and signal transmission.

 

[kemmler3D]

Yeah, I didn't finish reading the article but it's pretty disingenuous or mistaken. I once thought I had discovered the exact same justification for high resolution audio. The sampling rate is lower than human ears' time resolution. Therefore we need high sampling rates.

Turns out, no.

It was patiently explained to me that the relevant time resolution of redbook is not what he claims in this article (I think he made up that formula) but a product of the sampling rate and bit depth, and is much better than microseconds. This makes sense if you think about it. The coincidence of two peaks or sounds in a digital audio file doesn't just depend on the timing of the samples, it also depends on the amplitude of the samples.

The way he's illustrating the supposed problem is also a bit dishonest. Two peaks in subsequent samples represents click sounds occurring 22K times per second. This has no relation to music or any other sound we're familiar with. It especially has no relationship to the observation that timbre is affected by deleting the attack and decay of a musical note. (Which lasts for thousands of samples for any woodwind by the way.)

 

[fpitas]

Here we go again. Strap in and keep your popcorn handy!

 

[Curvature]

Yeah, I didn't finish reading the article but it's pretty disingenuous or mistaken. I once thought I had discovered the exact same justification for high resolution audio. The sampling rate is lower than human ears' time resolution. Therefore we need high sampling rates.

Turns out, no.

It was patiently explained to me that the relevant time resolution of redbook is not what he claims in this article (I think he made up that formula) but a product of the sampling rate and bit depth, and is much better than microseconds. This makes sense if you think about it. The coincidence of two peaks or sounds in a digital audio file doesn't just depend on the timing of the samples, it also depends on the amplitude of the samples.

The way he's illustrating the supposed problem is also a bit dishonest. Two peaks in subsequent samples represents click sounds occurring 22K times per second. This has no relation to music or any other sound we're familiar with. It especially has no relationship to the observation that timbre is affected by deleting the attack and decay of a musical note. (Which lasts for thousands of samples for any woodwind by the way.)

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

 

[Bending]

I am not too interested in warming up the time resolution discussion of digital audio (remember MQA claims). I am thinking more about likely the most significant issues in preserving time resolution:
Recording and mastering. Presently almost all music is recorded through multiple microphones, each with a different distance to the acoustic sources and then mixed often with different delays etc. I find recordings with are done by a single stereo pair mikes and no mixing like MA Recordings have a much better soundstage.

Loudspeakers: most loudspeakers use multiple drivers and therefore require a crossover which have phase issues but also time alignment of drivers is only possible for a single location = sweet spot. Also different driver locations cause different room interactions (not much an issue if the wavelength is large enough). But is even for bass timing essential? Also what is the impact of DSP (which is becoming much more used in active loudspeakers) in time resolution

 

[NTK]

From Dr David Griesinger's (slide 35) : https://www.davidgriesinger.com/wp-content/uploads/2026/07/Banff-intermod-.ppt

 

[john61ct]

So is this topic relevant to

"no point to transporting audio at resolutions higher than 24-bit 48kHz" ?

HiRes FOMO - Am I really missing anything?

Very nice list, thanks for that, looks great for exploring. Just to put things in perspective, I do not plan to involve a PC in the signal chain. Budget wise, my most expensive amp so far is under $300, same with my most expensive speaker. I may easily get up to over 20 output channels, and...

www.audiosciencereview.com

 

[Curvature]

He mixes true things established by research, with loose remarks and stupid comments about digital audio and signal transmission.

I'd like to spell out some implications here.

I think Kunchur is one of the best examples we have of intellectual dishonesty. He is an expert in his field of low temperature physics, a field which has nothing to do with audio, and uses a combination of his reputation and academic skill in presenting arguments to make unsupported claims.

If you have spent any amount of time in serious academic writing and research his published work on audio is instantly, obviously awful.

The exact technique he uses is to place completely unfounded and unresearched claims next to commonplace demonstrations and well-known facts. This is deliberate. As a professor he has written rigorous work, been peer-reviewed, has been the peer doing the reviewing, and likely has caught a fair amount of academic fraud from his students. Outside of that the very work of academic research is carefully reviewing arguments and routinely looking for and assessing support. So he knows what he is doing. He sees the gaps and the unbalanced presentation of claims and facts in his work, and believe me he would never let his students get away with the same. Whatever his motivations are, they cause him to hold himself to a different standard and disregard professional and scientific ethics.

I've met my fair share of professors with tunnel vision, so this isn't all that uncommon. It is frustrating, and his word is used to support the economics of dishonest audio manufacturers and dealers.

 

[fpitas]

his word is used to support the economics of dishonest audio manufacturers and dealers.

There we go. Follow the money...

 

[Curvature]

There we go. Follow the money...

I don't know that he is being compensated. He does, however, strike me as a deluded smart guy who might be trying to justify his immense personal spending.

 

[mlsstl]

That's a common problem with "experts" -- they have a strong tendency to wander outside of their area without realizing they are no longer an expert.

 

[Cbdb2]

His references are a joke. Well they had me laughing. And he seems to think those "spectrum" images replace real fourier analysis. There pretty useless except to hoodwink audiophiles. Lots of hand waving, zero math.

 

[Multicore]

If you have spent any amount of time in serious academic writing and research his published work on audio is instantly, obviously awful.

This