mike_mccue
The premise of my statement is that the room reflections, the resonance, and impedance of the air within it is not physically changed, but rather it is anticipated and the signal injected into the room is altered in some way which interacts with the existing physicality in such a way that a net effect is achieved.
 
This is, to date, my best understanding of what is going on... I'm eager to learn more.

I'm thinking from the perspective of the listening position. How "correct" the sound arriving there is is what I would focus on.

I started from the premise I read in the original ARC (1) ad copy.
 
The content of their latest ARC2 description has been groomed to seem much more accurate but originally the premise they introduced was that the software corrected the room.
 
Then I thought to myself, Yeah... I wonder if that is actually possible, and eventually I arrived at a new under standing, using what I consider fairly conventional semantics.
 
After having developed that understanding, I have made subsequent statements describing what I thought I understood. In other words, my statement was based on a premise, but my achievement of that understanding was not based on the premise that my statement reflects.
 
 
 
 
 
As much as I like ARC I think having speakers equipped with Audyessy dsp, where each and every single driver can be controlled via an active cross over is the juicy juice. I'm thinking maybe a couple ARC3 equipped iLouds and a iReallyfreakyLoud™ subwoofer all doing the magic would be so good I'd want to fly bapu in for a listen see.

I think what you might still be missing, based on this statement:
 

mike_mccue
In addition to making dramatic EQ adjustments to your speaker output it uses "fuzzy logic" to dynamically combine the collection of impulse response samples you make to compensate for the effect of your room reflections.
 

 
"EQ" and IR compensation are the same thing. I said this in another thread, but an FIR filter just convolves the signal with the filter's IR the same way a convolution reverb does (in theory) with its IR. And frequency and phase together gives you time (f=cycles/second and phase in degrees = cycles/360). So if phase is taken into account, frequency compensation (i.e. "EQ") and time compensation are the exact same thing.

Well Said!
 
It's Audyssey's use of the term "fuzzy logic" and what I've been assuming is some sort of application of dynamic convolution that really fascinates me.
 
Although, honestly, the mathematics of both aspects have always seemed a bit beyond my attention span, I try to appreciate the underlying ideas.
 
Thanks for clarifying.

batsbrew
the mic is cheap...
but IS dialed in for a specific purpose.
 
it was probably easier to write the software using the sonic fingerprint for a known source,
rather than let people arbitrarily pick ANY mic (albiet quite possibly a better one) and the software NOT know how to compensate for all the corresponding readings....
 

ARC's use of a test mic begs comparison to systems that can use any omni mic, but work best when that mic is accompanied by a calibration file that is unique to that mic and reflects the results of a recent test.
 
The use of mics that have calibration data associated with them is fairly routine in test procedures.
 
It may not be as easy to code, but a program such as ARC could use that sort of data if it was programmed to.

Hi drew,
 
 Regarding what you have just said about FIR, what is your take on this:
 
 from: http://www.ikmultimedia.com/products/arc/
 
"Other types of electronic correction systems use some form of equalization that works only on the frequency response of the room but worsen the phase response of what you are hearing."
 
1) a fact
2) a generalization
3) a fact and a generalization

mike_mccue
Hi drew,
 
 Regarding what you have just said about FIR, what is your take on this:
 
 from: http://www.ikmultimedia.com/products/arc/
 
"Other types of electronic correction systems use some form of equalization that works only on the frequency response of the room but worsen the phase response of what you are hearing."
 
1) a fact
2) a generalization
3) a fact and a generalization

If you just change frequency response and completely ignore phase you really have no idea what you will end up with in terms of phase response. This is where the oft heard "you can't correct time problems with EQ" statement comes from.
 

It's Audyssey's use of the term "fuzzy logic" and what I've been assuming is some sort of application of dynamic convolution that really fascinates me.

 
I think all they're talking about is how they combine the impulses from different positions to make a single impulse that works reasonably well over a larger listening area.

Here you go Karyn...turns out they sell the pixie dust separately.

Awe, thanks Dave

See... I knew Dave would have the answer.
 
Bitflipper= nails it every time