Jones_r wrote:
"Yes, that's what happens when you try to use a frequency-domain
filter to
fix a time-domain problem in your room. It doesn't work."

Scott, why do you say that ?, digital room correction is all about
fixing time-domain problems.

You cannot fix time-domain problems because they are different in every
location. You can put a test microphone in one place in the room and measure
an impulse response, then move it a few feet and measure a different impulse
response.

This is how the room works, and you cannot use a reverse transform to undo
the impulse response because it is different in every location.

Signals come out of the speakers, and they bounce around the room. Depending
on the frequency (and therefore the wavelength) they bounce around in very
different patterns. They sum in places and they cancel out in places, and
what you hear is the particular sum of different waves which mix together at
the point where your ear is.

In order to compensate effectively for the room response, you would need to
be able to produce a controlled wave front that could be adjusted in pattern
in three dimensions. You can't do this with a speaker, or a pair of speakers.

It fixes frequency + time domain
problems. Of course, like everything in life, it has its limits.

Those limits are very narrow, though and don't include most of the problems
that people encounter in typical home listening situations.

It does include some of the problems that get encountered in large room PA
applications, which is why you'll see things like parametric filters in
common PA use. They are very effective for solving a particular set of
limited problems. But, for example, they can do nothing about flutter echoes
from arched ceilings which are very common problems in PA rigs.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."