Audibility thresholds of amp and DAC measurements

[j_j]

I don't know, it's pretty surmountable according to dr. Griesinger, and he doesn't ignore that work from the 30s.

I'm well aware of what David says. Now apply it to a head that moves, in a space that has a second set of acoustics. Furthermore, I do not see a lot of work that gets over the obliteration of distance cues in the center in a 2 channel system, UNLESS you have that "head locked in a vise" interaural cancelation.

Natural listeners move their heads while listening.

 

[j_j]

I have no idea. I suppose I'm trying to get where he is coming from.

I did injure my head so I'm probably just not thinking correctly. Ignore anything silly I say for a while or I'll just step away from the conversation lol.

Well, the frequency analysis of the ear is not "uniform in frequency" but that does not invalidate anything at all about how the signal going into the ear can be represented.

 

[captainbeefheart]

Well, the frequency analysis of the ear is not "uniform in frequency" but that does not invalidate anything at all about how the signal going into the ear can be represented.

Yes of course, the impedance of the ear is all over the place vs frequency. Headphone designers are no longer following the loudspeaker flat design model where the end result of average effect to the ear is a slight angle sloping downward as frequency increases. They built composite ear models and installed microphones to figure all this out including the standing waves of the ear. The result is quite interesting that headphones in order to sound flat are really anything but flat. Looks more like the smiley face eq, scooped mids and boosted bass and treble, for bass slam they also have a peak around 150Hz, like how distortion harmonics of low frequencies reinforce the fundamental, instead of distorted bass frequencies they just boost this frequency region.


I took what he was saying had nothing to do with the ear and instead saying the Fourier theorem doesn't work for all signals in music through amplification devices. So I was thinking more of a time analysis of the system, where we use Laplace theorem for stability criteria of the system. We use square waves for step response and stability testing amplifiers, I thought this is what he is talking about sound quality that an FFT distortion plot won't show. Basically amplifiers can be stable/unstable systems and that can effect sound quality of course.

 

[dc655321]

So I was thinking more of a time analysis of the system, where we use Laplace theorem for stability criteria of the system. We use square waves for step response and stability testing amplifiers,

You mean Laplace transform?
If so, it may be a good idea to refresh yourself on the relationship between Laplace and Fourier transforms.

 

[j_j]

You mean Laplace transform?
If so, it may be a good idea to refresh yourself on the relationship between Laplace and Fourier transforms.

There are both theorems and the Laplace transform. Both Fourier and Laplace transforms are complex exponentials, of course. Different ones, again, of course.

 

[dc655321]

There are both theorems and the Laplace transform. Both Fourier and Laplace transforms are complex exponentials, of course. Different ones, again, of course.

I’m somewhat familiar with the transforms.

What is the Laplace theorem? Even google is not much help here (unless you mean de Moivre-Laplace Theorem, which seems out of context). Is theorem being used as a synonym for property?

 

[j_j]

I’m somewhat familiar with the transforms.

What is the Laplace theorem? Even google is not much help here (unless you mean de Moivre-Laplace Theorem, which seems out of context). Is theorem being used as a synonym for property?

Laplace did a bunch of stuff, including at least an ostensible theorem showing that Laplace bases were a frame, although not necessarily a tight frame. They didn't use that set of words then,though.

 

[pma]

Nothing can re-create sound we hear in a concert hall from a philharmonic orchestra in our living room. It is technically impossible.

 

[j_j]

Nothing can re-create sound we hear in a concert hall from a philharmonic orchestra in our living room. It is technically impossible.

Um, one can do a very good job of recreating the perception of such. It is probably technically impossible to do properly, but both perceptual spatial synthesis and wave field synthesis can do a decent job.

For the "full thing" one runs into the problem that the measurement in the soundfield will affect the soundfield. It's kind of hard to fix that.

 

[captainbeefheart]

You mean Laplace transform?

Theorems of Laplace Transform: https://engineering.electrical-equipment.org/others/theorems-of-laplace-transform.html

 

[xaviescacs]

Just a thought. People coming from physics, like myself, used to continuous Hilbert spaces for quantum mechanics formulation, can find a discrete setup quite challenging, it requires some careful thought to adapt your mind. The continuous world is conceptually easier IMHO. That might help to explain some confusions in the preceding pages about Fourier analysis in audio while claiming that QED is more difficult and such things. Or maybe not...

 

[dc655321]

Theorems of Laplace Transform: https://engineering.electrical-equipment.org/others/theorems-of-laplace-transform.html

That’s the only link I was able to find yesterday too… I was taught that those were properties, not theorems.

But that’s not really the point…

 

[captainbeefheart]

That’s the only link I was able to find yesterday too

You edited your post, you originally found de Moivre–Laplace theorem which is basically saying as n grows large, the shape of the discrete distribution converges to the continuous Gaussian curve of the normal distribution.

Laplace transform when we learned it were taught about the theorems and properties to make it more applicable to linear dynamical systems. I apologize if it wasn't clear enough but I thought it was obvious that instead of just naming the transform in general it's more pertinent to discuss the properties or theorems of the transform that apply to what we are discussing.

 

[KSTR]

Nothing can re-create sound we hear in a concert hall from a philharmonic orchestra in our living room. It is technically impossible.

I would beg to differ. When we manage to re-create the original 3D-sound field in large enough space around the listener's head, say within a sphere of 1m diameter, there is nothing in the audible domain that could give tells that this is not the original event.
WFS (Wave Field Synthesis) can do this and has been proven to work experimentally at least.

With current recordings and playback systems you are right, of course. Then again, it has been shown that people might not prefer a playback that is actually a true capture of the live event sound field as other perceptional cues are missing, optical first and foremost. I mean, even if the playback signal is true mono we tend to locate the sound of a guitar player from the side where he is located on the stage we see in live concert or video recording.
Rather, the goal is to create a palpable and plausible illusion of a live event even without those other cues. For this, among other things, stronger separation of all the phantom sources than in real life will be useful and that will probably still be the case even with full-blown WFS.

 

[dc655321]

You edited your post, you originally found de Moivre–Laplace theorem which is basically saying as n grows large, the shape of the discrete distribution converges to the continuous Gaussian curve of the normal distribution.

Laplace transform when we learned it were taught about the theorems and properties to make it more applicable to linear dynamical systems. I apologize if it wasn't clear enough but I thought it was obvious that instead of just naming the transform in general it's more pertinent to discuss the properties or theorems of the transform that apply to what we are discussing.

I edited nothing.
You linked to the only use of the term, "Laplace Theorem" known to Google. Neat.

I would still like to understand how "Laplace Theorems", rather than Fourier Theorems (?), would shed additional light. I don't see what you're trying to say here:

So I was thinking more of a time analysis of the system, where we use Laplace theorem for stability criteria of the system. We use square waves for step response and stability testing amplifiers, I thought this is what he is talking about sound quality that an FFT distortion plot won't show.

EDIT: (see this is an edit! ) As said previously, I have some familiarity with Laplace "theorems" as applied to system stability (i.e. control theory).
I don't see how it's of use here (or supplemental to an FFT), though you seem to have something particular in mind.

 

[j_j]

I would beg to differ. When we manage to re-create the original 3D-sound field in large enough space around the listener's head, say within a sphere of 1m diameter, there is nothing in the audible domain that could give tells that this is not the original event.
WFS (Wave Field Synthesis) can do this and has been proven to work experimentally at least.

Actually, no. No wave field synthesis I've seen or heard had anything near the proper spatial sampling.

 

[captainbeefheart]

What is the Laplace theorem? Even google is not much help here (unless you mean de Moivre-Laplace Theorem, which seems out of context). Is theorem being used as a synonym for property?

Sorry I read your other post and since I didn't see the de Moivre-Laplace Theorem I though you edited it. And yes many people use 'theorems' and 'properties' as synonyms. Americans I find say properties where I hear foreign educations say Theorems. I just looked through wikipedia right now and they also use the two terms.

Chapter 4 "Properties and Theorems"

Laplace transform - Wikipedia

en.wikipedia.org

I would still like to understand how "Laplace Theorems", rather than Fourier Theorems (?), would shed additional light. I don't see what you're trying to say here:

Following the conversation the only person alluding to FFT not satisfying a full analysis of 'music' signals passing through a device. As I mentioned already, I was asking him if he is talking about step response or stability analysis of a system (device under test) where Laplace transform is useful.

Fourier transform is a special case of the Laplace transform so I didn't think it was really this confusing of a concept for possible audibility concerns in equipment. Unless stability isn't a concern for people discussing the quality a closed loop feedback system which most amplifiers are.

 

[srkbear]

Actually, no. No wave field synthesis I've seen or heard had anything near the proper spatial sampling.

You said a few posts up that WFS does a decent job. Help?

 

[j_j]

You said a few posts up that WFS does a decent job. Help?

"a decent job" is not "an accurate job". So far, it's also stuck to the plane.

 

[srkbear]

"a decent job" is not "an accurate job". So far, it's also stuck to the plane.

Understood, thank you.