Reg on EQ, improving sound

[RU9]

<regonaudio@aol.com>


Since we have quite a few new people, it occurred to me that it
might be useful to have a quick summary of the overall picture as I
see about sound in rooms and EQ and so on. Details are on
www.regonaudio.com but in quick form, my personal view:

The sound of an audio system is Predominately determined by the time
windowed frequency response. Barring gross distortions(amp clipping
and so on), the overall sound is almost entirely described by
knowinG how much energy in each frequency arrives as a function of
time.

In the lower frequencies(below around 300 Hz) the room contribution
and the direct sound of the speaker are essentially treated by the
ear/brain as a unity. If this region is made RTA (steady state)
flat, it sounds right. Further up, there is a gradual increase in
emphasis on direct arrival,although the room sound still counts.

A good summary of why one might come to believe this picture is
given in www.gradient.fi , the design white paper(note especially
page 3).

Much of what is wrong with audio systems in rooms can be fixed by EQ-
-in the bass essentially ALL that is wrong can be so fixed. Higher
up, life is not so simple since the ear/brain begins to hear the
direct arrival, then the early reflections, then the "room sound" as
a whole in separate and different ways. The best hope is to get rid
of as much of the room sound and the early reflections as possible
in the higher frequencies.

Most speaker/room combinations do not do all that well. But a good
speaker and careful placement can get one surprisingly close. Have a
look at the Harbeth M40 graph in my photo gallery here. One might
like to EQ out a little of the extra "warmth" and I usually add a
subwoofer to pull up the deep bass slightly, but overall, it shows
signs of working! So do some of the other speakers measured on
www.regonaudio.com at least with a little judicious EQ.

So there is the "big picture" as I see it. This message is really
for new people only. Everything in it is old hat to those who have
been around a while!

REG

PS New people: I really really recommend a very careful reading of
the design brief on www.gradient.fi
especially the part (page 3) about how people could essentially not
tell the output of an anechoically-played flat speaker from the
source being played, that the differences were very small. This fact
is a profound truth.