hello,

i was wondering how dramatic the angle has to be for a non-parallel
wall has to be to achieve it's intended effect. i'm guessing an angle
as modest as 15% could be effective, but i'm not sure.

wrote in message ups.com...
hello,

i was wondering how dramatic the angle has to be for a non-parallel
wall has to be to achieve it's intended effect. i'm guessing an angle
as modest as 15% could be effective, but i'm not sure.


Fast help.... F. Alton Everest books.

This thought has often crossed my mind. I have a cunning theory about it...

One theory has it that the room dimensions should vary in third-octave
steps -- that is, 1, 1.26, 1.6. Therefore, it would seem reasonable that the
lengths opposite the non-parallel sides should have the same ratios. For
example, if the front wall of the room is 16 feet wide, the rear wall should
be 20 feet wide.

gaps,

i was wondering how dramatic the angle has to be for a non-parallel wall
has to be to achieve it's intended effect. i'm guessing an angle as modest
as 15% could be effective, but i'm not sure.

The rule of thumb is to have both walls angle by a minimum of one inch per
linear foot. The goal is to have enough of an angle to avoid flutter echo
and mid and high frequencies, without having to apply absorbers or diffusors
to the walls. (Not that there's anything wrong with absorbers!) Note that
this sort of shallow angle has no real affect at low frequencies.

--Ethan

I think it depends largely on why you are angling the walls.

If you are trying to defeat discrete echoes, I think that 15% would be
very effective, but you do have to look at the entire geometry of the
space to know for sure... I was playing around with walls recently and
discovered that I could in fact create echoesG!

If you are trying to alleviate a standing wave problem then I don't
know that there is a rule of thumb, it is going to depend on where in
the spectrum the problem(s) occur.

Mr. Sommerwerk has a VERY intriguing idea, and I'm going to play around
with it a bit and see if it works.

Dang but that is a very clever observation whether it turns out to work
or not!!!

One other point ot keep in mind.. a space with non-parallel walls will
still have standing waves, they will just be WAY tougher to calculate.

Bill

so 1 inch per 12 inches is only 8.3%...cool.


why is 1/3 octave a magic number (other than what a lot of graphic eq's
use)?

for flutter echo avoidance (not standing wave alleviation)...that's
what i'm after.

why is 1/3 octave a magic number...?

My guess -- and it's only a guess -- is that, as a room has three
dimensions, the "correct" choice of dimensions will space the room modes
evenly within the octave (rather than piling them up).

But that "explanation" strikes me as tautological.

for flutter echo avoidance (not standing wave alleviation)...
that's what i'm after.

A few well-placed Fibreglas panels will work wonders...

Mr. Sommerwerck has a VERY intriguing idea, and I'm going
to play around with it a bit and see if it works.

Please keep us posted.

Note that what proposed (which I really didn't recognize when I posted it)
is that it's not non-parallel walls we want so much as a variation in wall
length across each dimension, thus (theoretically) "spreading out" the
modes.

I'm trying to imagine what the room would look like if my suggestion were
applied to all three dimensions. It's a shame Escher is no longer with us...

i have his sound studio construction on a budget book. it's a great
book, but it seems to be like a constant veiled sales pitch for rpg
diffusors. he talks about there being "no question as to the
effectiveness of wall-splaying in the elimination of flutter echo".
but then he goes on to say it's a waste of money and you should just
get some diffusors and absorbers.

well, rpg wants $200 for a few square feet of diffusors. you can erect
an entire non load-bearing splayed wall for less than $200 using
drywall and wood from home depot.

well, rpg wants $200 for a few square feet of diffusors. you can erect
an entire non load-bearing splayed wall for less than $200 using
drywall and wood from home depot.

You can build diffusors yourself if you want to go to the trouble. And, of
course, diffusors have advantages you don't get from non-parallel walls.

In article .com,
wrote:
so 1 inch per 12 inches is only 8.3%...cool.

why is 1/3 octave a magic number (other than what a lot of graphic eq's
use)?

It's one that works.

As I recall from a lecture that Don Davis gave that I attended
back in about 1976 of 1977, at one of his Syn-Aud-Con seminars when
we were building the first 32-track analog studio in the US. [Using
production equipment with 2" tape - not prototype/experimental
such as the 3", 32-track, 22.5 ips MCI beast].

He was originally cutting filters to match the exact resonance
when trying to eliminate feedback in sound-reinforcement
situations. Each room would be measured and a fixed frequency
filter would be made. Then the next point would be measured and
another filter built.

Then in 1967 [I just looked it up] Don was at Altec-Lansing
and invented the first 1/3 octave filters - cut only - that
was called "Accouti-Voice".

I remember having something like that when I was working in
broadcast before I moved into recording.

As to 1/3 octave being magic, we dumped out Urei's for eq and went
with White that had 1/6 octave spacing at the low end. It made
tuning the control room easier. In the end, after bringing in some
charting devices and using warble tones to accurately measure the
response down to a cycle or two, and then measuring things, we
found the console [ about 12 feet long - Sphere Super C - the
prototype ] was reflecting from the compression ceiling. [This was
a long time ago].

So we rebuilt the CR.

Things are so much nicer today compared to 'the olde dayze'.
When parametric EQs came out they seemed like magic.

I've not been in a studio for about 2 years when I observed
a cylinder recording session. And I've not cut anything for
commercial release in about 20 years.

Bill

--
Bill Vermillion - bv @ wjv . com

wrote in message
oups.com

why is 1/3 octave a magic number (other than what a lot
of graphic eq's use)?

1/3 octave roughly corresponds to the ear's critical bands.

Critcal bands relate to masking. The strongest signal in a
critical band tends to "swamp out" the other signals in that
critical band.

IOW, the softer signals anyplace in a given critical band
will be vastly perceptually attenuated by the loudest signal
in that critcial band. But, a signal in the next critical
band will not be masked nearly as much.

It's pretty easy to simulate a group of filters with a range
of bandwidths using a multiband parametric eq. If you start
fooling with filters like these, you'll discover that when
the filters are on 1/3 octave centers, they tend to have
independent audble effects. When they are on 1/6 octave
centers, they are far less independent because they are
changing signals that are in the same critical band.

The begs the question of how close filters have to be spaced
in order to *perfectly* equalize a signal. I understand that
current *revealed truth* is 1/12 octave.

What does perfect equalization mean? A perfect equalizer
would restore a signal that had *any* reasonable kind of
frequency-dependent attenuation and boosting, to one that
was indistinguishable from the original signal.

Needless to say, experiments like this demand blind testing.

There is a widespread belief that you can hear the effects
of *any* equalizer just because it is in the line. And, as a
practical matter this is true, because outside of lab tests,
nobody has ever obtained *perfect* equalization. Trust me,
you'll never obtain it by ear in any reasonable amount of
time.

If you do obtain perfect equalization in a lab, the stack of
equipment will proabably make you want to gag. It seems
improbable that such a mess of stuff could process audio,
and net out to no sonic effect. But, if you do evertything
right, it can work.

"Bill Vermillion" wrote in message


Then in 1967 [I just looked it up] Don was at
Altec-Lansing
and invented the first 1/3 octave filters - cut only -
that
was called "Accouti-Voice".

That would be Acousti-voice.

http://www.realspace.com/SigTech/aes_90sf.html