inverse response speaker flattening

[Magnus Danielson]

From: Martin Czech <czech at Micronas.Com>
Subject: inverse response speaker flattening
Date: Fri, 12 May 2000 07:54:07 +0200 (MET DST)

Martin,

> There is one thing I wonder about:
> 
> You have music from digital sources all the way, till power amplifier. It
> would be no problem to ship a little digital in/out box with a measuring
> microphone, so that anyone can measure the frequency response of his room
> and his speakers, the box takes the inverse to filter the CD data. That
> should give an almost ideal response for that single measuring point in
> space and hopefully around it.

Yes, in theory.

> Of course, once the speakers are flattened, any personal taste could be
> applied as well (bath tub).

Uh, filled or non-filled, if filled, with or without person seated, if seated
in known position or not, if known position, foam or not foam on water, if
foam, at a known bubble size, shape and resistnese over time of listening?

> One problem could be that shitty bass chassis could not take the needed
> power...

All forms of electronic compensation eats on the overhead and noisefloor.

> I have never seen any advertisement for such an apparatus, it seems that
> something like that simply does not exist.

Well, this is not entierly true. I think I recall that Meyer had some simple
system for PA-rigs in the 80thies. Also, I recall that Adamson (another PA
speaker manufactor) was heading this way. Labbgruppen has too made efforts of
this kind with their DSP24 product as a base, it had long been in the cooking
and I never saw the full product (DSP and program) the running. Labbgruppens
efforts is based on research at Chalmers and that fellow moved his work over
to Labbgruppen. At Chalmers he did have a system operating and according to a
friend of mine being in good contact with these fellows it was a really
astonashing sound.

Now, the trouble in trying to do something like this is that you really have
to measure the sound at multiple points in the room, at multiple times with
diffrent arrangements. Also, you are best of starting to compensate for
fundamental properties of the elements, then system before you attempt to
compensate for room responce, since it is more feasable that the complexity of
the problem has been reduced using existing knowledge and it is more probable
that the end result will be good.

While it does sound very interesting to do such a thing, I think one should be
carefull in overdoing it since it can certainly bite back and there are also
the trouble how you are able to measure curtain parameters with accuracy.

> Another question:
> 
> In the 70ties people thought about switching amplifiers.
> I.e. DAC amplifiers. The audio chain would be digital up to
> the speakers. Are there any reasonable advances/designs
> today?

Even more extreme is the Philips patent where each bit controlled individual
parts of the element.

I really think the DAC followed by a linear amplifier has many benefits in
cost vs. preformance and it is not an accident why the market has reached this
solution. It is partly due to accident ofcourse ;)

Cheers,
Magnus