How Science Got Sound Wrong

It is rather cute Teo and Miller how you two bloviate with flowery language that would make it look like you know what you are talking about but to someone who actually understands this stuff you just sound ... Funny. Yes funny.


Teo. Jitter on recording is in the 10’s to 100’s of picoseconds. Ditto on playback. Easily modelled as phase errors or distortion. Distortion 100+ db even on real music signals. I know you don’t want to believe it but audio is easy. We do 100+db on complex signals day in day out in comms.


Not my first rodeo and I know the usual arguments that is why many many posts ago I posted a paper (that included actual experiments) that showed timing accuracy as a fraction of bits with very low signal.to noise ratios, ie like 30db, and the resolution is several orders of magnitude finer than the sampling rate.


Believe vinyl/current analog is superior ... No issues. But don’t hang your hat on something you are going to fail miserably on. I can post many experiments that clearly show high subsample timing accuracy in low SNR environments of which digital audio is not one.


All two channel does is increase the relative jitter mapped as higher effective SNR but since it is already very low the timing resolution is very high. Feel free to argue points you don’t have the background to argue, but at least back that up with some solid work by real experts to support what you are saying. Otherwise it is just hand waving.

I’ll take some of this action. The problem is not on the CD. The standard CD contains data with extremely high sound quality. You just can’t HEAR the sound quality, that’s all. The problem is the playback machine. There are a number of inherent problems with CD playback. Three problems that spring immediately to mind are (1) scattered background laser light, (2) seismic vibration and other mechanical vibration including the CD transport noise and acoustic vibration, and the vibration and flutter of the CD itself while spinning. The same high technology that created the Compact Disc also created the problems - nanotechnology and quantum mechanics.

Yes, I realize atdavid is going to try to tell me streaming solves all the problems with CDs. Sadly, streaming sucks. But it IS convenient, I’ll grant you that.

“You can be a knower or a blover.” - audiophile axiom

- your friend and audio insider,

geoff kait
machina dynamica
advanced audio conceits

So are we hearing distortions in the electronics? Digital filters not removing all distortions and disrupting electronics downstream. There is something going on that people hear. 

@ atdavid

So when I finally had some time ( it’s been a very busy week ) went off and did some sleuthing and reading. Anyways, to cut to the chase....found the following.....

It proves you can infer some time details beyond the sample rate, but this is practicaly a kind of limited exception to the rule implied by the sampling rate -not an escape from accuracy limitations.

It was shown that the phase of a sinusoidal pattern which is assumed as perfect and constant can be resolved to a fraction of the sampling interval.

This subsample accuracy was possible because the pattern recorded is not a discrete event, it’s impression is recorded throughout many consecutive samples and its exact formation is inferable (idealy).

For all discrete or unassumable events, PCM records can only specify time of occurence to within a whole length of the sampling interval. Time resolution can only be improved when a known pattern can be observed throughout multiple samples -which is the case for computing the phase of synthetic frequency components, but not at all when trying to refine the temporal location of unassumed events.

This seems to say that timing beyond the sampling rate is a rather special case that works in situations that have a long term term steady state input, and doesn’t apply to " discrete or unassumable "events. And correct me if I’m wrong but music is most likely something very akin to a discrete or unassumable events, eh.

That proved kinda interesting and sorta relevant so I decided to look at the articles you posted to back-fill your position, and not just go, ooooh, that right there is real honest sound and robust science, and just run away cause people like us are afraid of the dark and real science.

So looked thru those articles and sure enough timing beyond the sample rate is indeed possible.....in gas pipelines, which I assume are fairly steady state noise sources.....and not as one articles clearly states....

Notuseful for real-time applications. The whole signal needs to be known in advance.

Where the real-time applications kinda sorta sounds an awful lot like music....and the whole signal implies long term steady state.

So colour me confused....but it seems the articles you posted may be sound and robust science but the bottom line they don’t really support your contentions....if fact in one case quite the opposite and in the other case concerned with something not at all like music and from a strictly sound and robust science perceptive has very limited relevance .

That being said would be happy to learn more about this so please post other references, though if I were you I may want to actually read the article before posting, you know, just to make sure it says what you hope it does. 

You are confused because you do not understand the science, hence you put it down.

One of the papers Again clearly show orders of magnitude subsample timing at 30db SNR. Never will a digital audio system ever get anywhere near having as little as 30 db SNR, and if you say when the signal level is low, well then in your analog system it is all noise at this point. That paper looks at low SNR situations because the noise can be really bad, Unlike audio! Absolutely nothing I posted says the opposite of what I said and that you would state that suggests you really didn’t try to understand them. One only discusses difficulty in the presence of noise and other sources .... With examples where the SNR is 10dbs of db, not 90+, and they did not bandwidth limit the signal.

You post excerpts, not links to articles, why is that? The parts you link to show a lack of understanding and the same flawed thought process as the author. Audio Is Not a Real Time System. It is a recorded and played back system. There is 0 concept of real time in Audio. There is 0 concept of absolute time. Everything is audio as we are discussing is relative. That is why I clearly and distinctly use the term relative timing in most of my posts. That is the difference between actually understanding a topic and cutting and pasting things that match your world view.

Stop clutching at straws to attempt to justify a point of view that is wrong. I read the article. I understand it. That is why I know it is flawed. You want it to be true, but wanting something true and it being true are not the same.

Everyone who understands signal processing and digitized systems will instantly pick out the flaw in the premise of the article which is exactly what others online have done wrt this article, to the point of contacting editors because it is such a gross misrepresentation of reality.