You might want to take a look at this excellent post by Robert E Greene (of TAS) from his forum, below:
"It looks to me as though we need a bit of review
on the whole subject of EQ. Of course, I am all for
using it. The failure to do so has held back audio a lot. But
as this failure ends, and it is ending--EQ devices are proliferating
enormously-- one needs to give some serious thought
to what one can do with it and what one ought to do with it.
There are really two issues, or more precisely a continuum
between two poles. One is broad-band balance, The other
is micro-detail, errors associated to narrow band effects, high
Q resonances in particular, Both things are important. But
they are quite different.
Broad-band shifts in frequency response are highly audible.
They are also easy to fix with EQ. Here, agreement is
wide-spread among audio pros. People in the pro world
take it for granted that one can deal with broad imbalances
without doing any damage otherwise. This is what the Schitt
devices do , and the Cello devices before , and what
"tone controls" in general do. Such things need to be done
carefully, but they are not all that hard to do. Everyone
takes this for granted that is thinking straight. After all,
your speakers contain such EQ at speaker level , where
it is harder to do than at lower levels. I think one can take
as given that some sort of smooth correction of broad band imbalances
is a good thing to do.
One really convenient way to do this is 1/3 octave slider EQ devices
or their digital analogues. The choice of 1/3 octave has to do
with the critical band width of the ear. In practice it is detailed
enough to smooth out broad band effects and not so narrow
as to (used within reason) make for solutions worse than the problem.
Even broader band devices like the Schitt EQs are quicker
and easy to adjust , to the point where one can adjust them
recording by recording. Very useful. And one gets to where
one can do it fast and easily to good effect.
In another category entirely are the higher Q (narrower band)
effects, arising for example from resonances in speakers.
These are harder to deal with. Ideally, they would not be there.
In principle they can be dealt with --all linear effects can be dealt
with by DSP. But in practice it is good if they are simply not there,
are not present in the speaker to begin with. Designers may
want to use DSP to kill resonances in their speakers but this
is a little tough for consumers. In outrageous cases., one can
help things, es ofpecially with parametric EQ, whether digital or analog.
But some caveats are needed. One can make a real mess of things
by doing a lot of narrow band EQ attempting to smooth things out..
And automated systems for this often do not work at all well, either--
they look good sometimes in certain types of measurements
if one measures superficially but the actual
sonic effect is highly dubious.
And crazy ideas arise. One of them is the idea of bands with abrupt cutoffs.
This is almost sure to be a disaster.
The idea of what should happen is that one should get a
resonance free system , whether by nature and design or
by designer-done DSP (eg in digital speakers) and then
work with the broad band tonal balance matters with comparatively
low Q corrections.
The one area where somewhat abrupt things need to happen
is in the bass and low mids, where the room is knocking things
around a good bit. Here is the one region where one really
needs to get in there and "give em hell". Automatic systems
to do this can work well. But again one should keep in mind
that you do not really need to get things micro-corrected
and indeed are probably better off not to do it.
As EQ becomes acceptable again, people are getting a little
crazy about trying to correct spatially unstable, micro variations
which are not really correctable. This EQ stuff can be made to
work, but one has to think. First stage is to get the system itself,
the speaker in the room, to work well. And then fix up the borad band'
balance as needed.
Almost all systems need some fixing up. But one has to
think about what is seriously audible, what is not and what
damage one can do trying to fix things that would be better
left alone. And don't worry about equipment choices until
you have decided what you are trying to do. Audiophile
tend to reduce problems to the question of what equipment
they should buy, In this EQ world, the first question is what do I want to
do, Electronic devices tend to work well-- but you have to
think first about what device addresses the issue you want to address.
REG
--
Robert E Greene
_._,_._,_
_._,_._,_
