The ASR "objectivists tribe" ;) has to read the Keynotes at the AES160th by Lars Risbo

[krabapple]

What's ultimately needed is speaker and DSP tech that effectively compensates for the room. We're in the early stages of stuff like that with Dirac ART and Atmos and such, but I think we have a long way to go. As long as stereo recordings are the standard there isn't much information to preserve if you are trying to synthesize a sound field Vs. A linear reproduction of an audio signal that happens to radiate in a smooth pattern.

Dirac , Audyssey, etc (at least, below Schroeder) are trying to compensate for the room, at the listening position. People still apply DSP room compensation to Atmos setups, even though adding channels in all three dimensions does tend to lessen the contribution of the room. The main problem is bass, and that's usually offloaded to one or more subwoofers in Atmos configs. So, either ideal subwoofer configs, or some DSP, are needed.
.

 

[Salt]

At the risk of publicly aligning myself with the wrong tribe, I'm gonna give away a "trade secret" here:

Yes that happens all the time when I do sweeps on the raw drivers in preparation for doing the crossover design. It's an ugly little truth which my marketing department never mentions. I make note of where I hear these resonances (or whatever they are) without regard to whether they are visible in the frequency response or impedance curves. Then in the crossover design phase I will often deliberately put a dip there. Obviously this will degrade the measured performance, so I'm prioritizing ears over measurements (without controlled blind testing) where I perceive a discrepancy between the two.

Appreciate Your honest, and agree.

 

[krabapple]

Are you suggesting "intelligent design"??

Or unintelligent, it depends on the designer. ;>

But definitely design. As opposed to chance. If Genelec were to follow the 'method' evolution uses, they'd go bankrupt.

 

[Duke]

Or unintelligent, it depends on the designer. ;>

But definitely design. As opposed to chance. If Genelec were to follow the 'method' evolution uses, they'd go bankrupt.

Yeah, well, argument from analogy is like... um... nevermind.

 

[RexrothPigeon]

Well, mutations are generated randomly, and nature selects......I don't think that's how Genelec's tweeters and waveguide came about

yeah, like social darwinism and every other attempt to steal valor from evolutionary biology and slap it on some unrelated process (cough psychology cough), this one doesn't survive scrutiny

 

[Buckchester]

Are you suggesting "intelligent design"??

But, who designed the designer?

 

[AndyBell]

If blind listening tests are such a great marketing vehicle why no one is using them? Even the Harmann Group did a relatively low amount of promotion with it?

Probably because they don’t give the desired results. They’re used here, however, because we desire results not affirmation

 

[kemmler3D]

Dirac , Audyssey, etc (at least, below Schroeder) are trying to compensate for the room, at the listening position. People still apply DSP room compensation to Atmos setups, even though adding channels in all three dimensions does tend to lessen the contribution of the room. The main problem is bass, and that's usually offloaded to one or more subwoofers in Atmos configs. So, either ideal subwoofer configs, or some DSP, are needed.
.

What I am imagining would go much further and would synthesize a full 3D soundfield as captured (or assembled) in the studio. So, mics would need a different approach, too.

Audio would have to be encoded as a 4D object, (Snapshots of a 3D object through time) not just a pair of 2D objects (snapshots of amplitude over time). Audio files would probably be bigger than video files are today.

In that scenario, you may need something that can compensate for the room full-range, not just in the bass. I'm talking year-2126 technology, I guess.

 

[Salt]

But, who designed the designer?

You may have a look into the Ufo-thread?

 

[Sal1950]

"I've never caught my husband cheating, and he's a good provider and father to our twins, Brayleen and Croyfisch, so of course I trust him when he says he has no idea where those women's panties in his blazer pocket came from"

Is she a relation to Hillary ?

 

[AndyBell]

IMHO this topic is simple : some parameters have not been discovered yet. Other than "something is heard".
Once they will be identified and detected, they should be measured.
This is an universal scientific approach.
(There may be an initial concept, but it is not valid until demonstrated)
“If you can not measure it, you cannot improve it” (Lord Kelvin)

Yes, but it takes properly controlled tests (blind, level matched) before it can be established “something is heard”.

 

[test1223]

You have left the scientific sector! Godspeed!

I am not that familiar with the current state of the argument. So I might be wrong here?

This is not how engineering works. You do not use the most tedious, expensive and slow measuring method (blind listening tests), if anything else is available.
And therefore in engineering there are calipers, clocks and scales. And in modern times lasers, ADCs and analysers of all kind. This is what you use to check for differences.
Only if you are chasing something new and unknown and you do not know how to do it with your tools, you have to resort to the basic stuff of observation: in audio that means listening with your ears, and if is is supposed to work at least to some degree, it has to be controlled (aka blind and level matched and ...)
But this is more science than engineering and so it is another sector. But it is the sector that Lars puts himself in, and quite outspokenly so.

Of cause there are measurements and the current understanding of the interaction of measurements and the perception of these involved. We discussed about (blind) listening tests and of cause the focus of my post is about that aspect. Listening test are important since most of the time you trade one distortion with another. Listening tests help a lot to make a lot of these decisions. Since a lot of times there is no metric which can make a better decision.

Showing a measurement that illustrates the audible gain does not give away their IP. It proves it.

In many cases a unique or rarely used measurement technique with the finding that a problem certainly exists is one of the key findings, so I don't agree here.

 

[Salt]

Is she a relation to Hillary ?

Quote:..."and he's..”

 

[MaxwellsEq]

That is noise. Measurable in vision and sound.

Yes.

And there are different types of distortion

Noise is NOT distortion.

They are totally different characteristics caused by radically different physical mechanisms.

 

[Mr. Widget]

What's ultimately needed is speaker and DSP tech that effectively compensates for the room. We're in the early stages of stuff like that with Dirac ART and Atmos and such, but I think we have a long way to go. As long as stereo recordings are the standard there isn't much information to preserve if you are trying to synthesize a sound field Vs. A linear reproduction of an audio signal that happens to radiate in a smooth pattern.

Maybe more than a long way to go?

Dirac , Audyssey, etc (at least, below Schroeder) are trying to compensate for the room, at the listening position. People still apply DSP room compensation to Atmos setups, even though adding channels in all three dimensions does tend to lessen the contribution of the room. The main problem is bass, and that's usually offloaded to one or more subwoofers in Atmos configs. So, either ideal subwoofer configs, or some DSP, are needed.

I am currently dealing with a new listening room where the sound field is a reverberant field that makes a mess of stereo localization as well as hearing details in the recording. The DSP controlled bass from my D&D 8Cs easily handles the room's acoustics, but no amount of DSP is going to fix the reverberant nature of the room. Can I treat the room and fix most or all of these issues? Sure, but the room's primary purpose is my living room and filling it with diffusion and absorptive panels is not what I want in this room.

 

[Sal1950]

Noise is NOT distortion.

Not as in the measurable harmonic type.
But if audible and not a part of the original source signal, then it is a distortion of it's reproduction.

 

[Sal1950]

In that scenario, you may need something that can compensate for the room full-range, not just in the bass. I'm talking year-2126 technology, I guess.

Sounds good, but I fear I may be long gone.

 

[Duke]

yeah, like social darwinism and every other attempt to steal valor from evolutionary biology and slap it on some unrelated process (cough psychology cough), this one doesn't survive scrutiny

That's a valid point; my analogy leaves out (presumably) intelligent design being a part of the loudspeaker design process. Analogies are often imprecise. Anyway, I've edited my post to include the qualifier "to a first approximation...".

 

[Mort]

Yeah, well, argument from analogy is like... um... nevermind.

At least you didn't compare to a car or a watch.

 

[krabapple]

Maybe more than a long way to go?

I am currently dealing with a new listening room where the sound field is a reverberant field that makes a mess of stereo localization as well as hearing details in the recording. The DSP controlled bass from my D&D 8Cs easily handles the room's acoustics, but no amount of DSP is going to fix the reverberant nature of the room. Can I treat the room and fix most or all of these issues? Sure, but the room's primary purpose is my living room and filling it with diffusion and absorptive panels is not what I want in this room.

You canna change the laws of physics. The problem you have is not that there is insufficient DSP to solve it.