Subject: [linux-audio-user] Re: Room noise + acoustics (was: SNR (not audio))
From: Stephen Burgess (strangelv_AT_lycos.co.uk)
Date: Thu Aug 01 2002 - 01:47:16 EEST
On Wed, Jul 31, 2002 at 10:11:19AM +0100, Steve Harris wrote: (cut)
>I'm tempted to just flood the room with whitenoise and measure with a
>mic at various positions.
Hmmm, I'm afraid room acoustics are a bit more complicated than that.
First, whatever you are going to measure with such a setup would be
applicable to record sustained sounds (think pipe organ), but doesn't
account for the room's behaviour to transient sound (think drums). But
then, what exactly are you going to measure ?
The response of your room excited by the signal coming from your speaker
placed at a position x1,y1,z1, picked by a microphone placed at a
position x2,y2,z2. So this is dependent on the loudspeaker response, the
microphone response, the speaker and the microphone positions, and of
course the room reverberation properties.
If this was all there was to room acoustics, you would place the
performer/ the instrument where your loudspeaker was, the microphone in
its "best" position, apply some equalisation and off you go with big
recording studio sound quality.
Unfortunately, room acoustics are not limited to that aspect, and this
is my second point : what happens when you suddenly interrupt the white
noise ? The sound picked by the microphone doesn't get immediately
interrupted. What characterises that decay is called reverberation. It
is what your brain processes to give you an idea of the size of the room
you're in. It is also what makes a room "pleasant" for listening
to/recording a certain style of music or speech.
It is very hard (if not impossible) to make a small room sound "big".
But this doesn't mean there is nothing you can do to your room. I have
no experience of home-studio recording, but I have the experience of
hifi/monitoring loudspeaker placement and this is what I've learnt :
a. there are "better" positions for your loudspeakers (and I would
assume this is valid for your microphone) to prevent them from exciting
(picking) resonance nodes in your room. You want to apply those rules
(rules of 1/3rd and 1/5ths) because treating the room for bass is
difficult because of the long wavelength involved.
b. you want to treat your room to prevent the most obvious and unplesant
effects like flutter echo caused by parallel reverberant walls. This can
be done by moving objects around and putting carpets or curtains on
walls at appropriate locations. I would also think that for recording
you would want to suppress the first reflection. This can be achieved by
having the instrument and the microphone at set positions in the room,
and having patches of absorbing material on wall/floor/cieling at the
adequate position. This treatment is of course only applicable to
mid-high to high tones.
c. in the case of loudpseaker monitoring it is not recommended to treat
the room too heavily with absorbing material because this makes the room
sound too dull (not enough reverb). I would imagine that for recording,
not heaving enough reverb is not really a problem. But if you are
recording and monitoring in the same room, then you want to think twice.
d. regarding other suggestions I've read on the list : there are indeed
devices you can build into the corner of the rooms to modifiy the bass
response (bass traps). As to the talcum powder is concerned, I'm afraid
this is a case of audiophile snake-oil. The "roughness" in the paint
will only affect very high frequencies (10kHz ?). If you have ever
visited recording or dubbing studios you will have noticed that their
wall treatment is much more radical than that.
Finally I would recommend a book that is not too intimidating with maths
and that I found excellent, it also goes beyond the subject of room
acoustics : Acoustics and Psychoacoustics by David M. Howard & James
Angus, Focal Press 1996 Music technology Series
Sorry this was not strictly speaking a "linux" audio group answer...
-- Contrary to popular belief, Unix is user friendly. It just happens to be very selective about who its friends are. (Kyle Hearn - quotes from alt.sysadmin.recovery)
This archive was generated by hypermail 2b28 : Thu Aug 01 2002 - 01:39:18 EEST