I've had some succes recording choir and other sources with ORTF(ish)
setups. Lately I've played with AB-stereo and this has turned out less
predictable. There is a too large element of randomness in my results. I
sometimes get a washy, boxy sound. I'm not entirely sure this is just
too much reverb. Could it be some phase related effect? Sometimes this
is helped by moving in, but I've also seen it helped by moving out.
Both cardioid and omni are AKG C480, just different capsules.

Recording choir a cappella in reverberant church, are there any rules of
thumb for AB-stereo[*] Relative ORTF or just rules of thumb for
reasonable starting points? Distance between choir and mics? distance
between mics, height? How compared to ORTF?

best regards
Lars Farm
[*] Yes, I know that I like most am equipped with a set of ears. Those
ears tell me there sometimes is something wrong, but not how to solve
the problem.

--
lars farm // http://www.farm.se
lars is also a mail-account on the server farm.se
aim:

(Lars Farm) wrote in
news:1h4ucha.ms8zvfch1800N%see.bottom.of.page.for. :

I've had some succes recording choir and other sources with ORTF(ish)
setups. Lately I've played with AB-stereo and this has turned out less
predictable. There is a too large element of randomness in my results. I
sometimes get a washy, boxy sound. I'm not entirely sure this is just
too much reverb. Could it be some phase related effect? Sometimes this
is helped by moving in, but I've also seen it helped by moving out.
Both cardioid and omni are AKG C480, just different capsules.

ORTF is IME one of the safest setups you can use. It gives a good stereo
image and "sounds like the room" quicker and easier than other
configurations.

Recording choir a cappella in reverberant church, are there any rules of
thumb for AB-stereo[*] Relative ORTF or just rules of thumb for
reasonable starting points? Distance between choir and mics? distance
between mics, height? How compared to ORTF?

"Ruls of thumb" follow...

AB recording is very sensitive to:
1) distance between mics
2) distance to source
3) size of source
4) room effects
5) Cardioids are twice as hard as omnis to get right.
and all these effects interact.

Your choir probably has strong and weak voices. The more variation between
individual singers, the farther back you must set up to get a blend.

The choir may be in 2, 3, or more rows. The wider the group, the wider you
must set up to hear the entire sound. The deeper (taller) the group, the
higher you set up to catch all the rows.

The wider the mic spread, the farther back you must set up to avoid a hole
in the middle, where left gets left and right gets right and there is not
much common center channel.

The more reverberent the room, the closer you set up. This may conflict
with the above.

Chorus is a bit different from some other sources in the sound is more
omnidirectional (a cardioid pattern?) than other acoustice sources. An
orchestral percussion section projects upwards. Brass and woodwinds go
mostly forward (except the tuba), strings sound best about 45 degress up.
But voices seem to carry in a wider pattern. Given that, you can mic
voices from a high angle close, or a low angle more distant (like points on
a sphere) and get similar tone.

[*] Yes, I know that I like most am equipped with a set of ears. Those
ears tell me there sometimes is something wrong, but not how to solve
the problem.

"Washy boxy" sounds to me like too far apart and/or too far away.

Lars Farm wrote:
I've had some succes recording choir and other sources with ORTF(ish)
setups. Lately I've played with AB-stereo and this has turned out less
predictable. There is a too large element of randomness in my results. I
sometimes get a washy, boxy sound. I'm not entirely sure this is just
too much reverb. Could it be some phase related effect? Sometimes this
is helped by moving in, but I've also seen it helped by moving out.
Both cardioid and omni are AKG C480, just different capsules.

How wide is your A-B set? With these configurations, most of your imaging
comes from amplitude differences between channels... the mikes are spaced
so wide that the phase differences between channels are extreme and your
brain can no longer figure out position based on differential phase. So
the low frequency imaging suffers.


Recording choir a cappella in reverberant church, are there any rules of
thumb for AB-stereo[*] Relative ORTF or just rules of thumb for
reasonable starting points? Distance between choir and mics? distance
between mics, height? How compared to ORTF?

This is so much a religious issue. Most of the folks I know doing A-B
like to get in very close and very wide. Close enough that the voices
are individually distinct. This articulation is considered part of the
sound.

Listen to some of the Mercury Living Presence recordings... these were
recorded with a spaced triad of three mikes with very wide separation.

[*] Yes, I know that I like most am equipped with a set of ears. Those
ears tell me there sometimes is something wrong, but not how to solve
the problem.

Welcome to the club! You are not alone in this.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Thank you Carey and Scott. An example of the boxy, washy quality I don't
like and a couple of others from the same place/organ/player to
illustrate the difference (and contradict myself because here the AB
seems better then the closer Jecklin...(still the unwanted quality is
the one found in the Jecklin sample)).

A couple of months ago I played a little with radically different setups
when recording a friend at an organ. It's not a cathedral. It's not a
large romantic organ. It's a small stone church (26x12m (85'x40')). The
organ is more of a baroque organ. Somewhat screechy. 16', no 32'.

pictu http://www.farm.se/wps/2005/09/04/orgel-i-skon-igen/
(the text is in Swedish, but the picture is definitely in English... ;-)
(you can actually see pair 1 from the below samples if you look
carefully)

I had tried a Jecklin disc, but got that washy, boxy sound I would like
to avoid. I moved out and wider and got a nicer sound. Perhaps too
reverberant for a contrapuntal Gigue by Bach, but better for Widor.
While I was at it I added an ORTF pair (ca 3m from balcony floor ca 1 m
higher than Jecklin and AB 6m above the floor or 2m above the balcony
floor and 70cm (little more than 2') apart). Since I had tracks and mics
and was experimenting I added a pair of pseudo-PZM (omnis flush to the
outer stone wall) at the sides of the organ. The pedal sections (C, C#
boxes are at the sides of the instrument.

floor plan: http://www.farm.se/bild/orgelskiss.jpg

The Jecklin disc and the AB-pair (pair 1 below) are the same mics and
capsules - AKG C480/CK62. The ORTF-pair on the balcony is Pearl DC-96.

http://www.farm.se/ljud/akf/pair_b_1_jecklin.mp3

AB -- http://www.farm.se/ljud/akf/pair_b_1.mp3
ORTF -- http://www.farm.se/ljud/akf/pair_b_3.mp3
wall -- http://www.farm.se/ljud/akf/pair_b_4.mp3
AB -- http://www.farm.se/ljud/akf/pair_w_1.mp3
ORTF -- http://www.farm.se/ljud/akf/pair_w_3.mp3
wall -- http://www.farm.se/ljud/akf/pair_w_4.mp3

The "pair_w_3" is sort of funny in the way the "bröstverk" (ruckpositiv
in English?) threedimensionally jumps out of the organ facade and sits
itself in your lap while listening...

Lars


--
lars farm // http://www.farm.se
lars is also a mail-account on the server farm.se
aim:

Lars Farm wrote:
Thank you Carey and Scott. An example of the boxy, washy quality I don't
like and a couple of others from the same place/organ/player to
illustrate the difference (and contradict myself because here the AB
seems better then the closer Jecklin...(still the unwanted quality is
the one found in the Jecklin sample)).

Boxy sounds are usually the result of room problems. Small rooms that
are very bright on the top end give you a lot of room reflections right
after the original sound.

In a boxy room, omni configurations will tend to exaggerate the problems.
If the room is long and thin you can often use a Blumlein pair or an
ORTF-like pair of hypercardioids to get rid of the worst of the problems
from the side walls.

If the room is square and small, all you can do is use the ORTF pair,
get up close, and then add fake reverb. When you use the reverb box,
add a considerable pre-delay to the reverb, so the fake reverb picks
up right after the real room sound drops off. This can do a lot to
take the boxiness out when you're dealing with a bad room.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Lars, spaced miking used to be a rather risky technique so far as
stereo imaging was concerned, but in more recent years this has changed
due mainly to research published by Prof. Michael Williams. He found
ways to compute the stereophonic recording angle for spaced microphone
arrays of all kinds, no matter what pattern of microphone (omni, wide
cardioid, cardioid, etc.) you are using, and also other factors which
place limits on the angles and spacings that will work well. And best
of all, his findings work very well in practice. Thus it is no longer
necessary to guess, or go through much trial and error, when setting up
a recording with spaced microphones.

This is a fundamental change in the situation for two-microphone stereo
recording in general. Effective setups can now be planned and carried
out with predictable, reliable results. Many items of "conventional
wisdom" on this topic which have not actually worked any better than
random guessing can finally be discarded--though I suspect it may be
another twenty years before the recording community catches on, people
are so used to fumbling around at this.

If you're an AES member you can download his papers on this topic
(actually anyone can, though it's less expensive for AES members), or
better yet you could try to find a copy of his recent book, _Microphone
Arrays for Stereo and Multichannel Sound Recording_, volume 1 of which
has been published by Editrice Il Rostro in Italy. But it may not be in
widespread distibution--I'm still trying to find out about this.

--best regards

David Satz wrote:
spaced miking used to be a rather risky technique so far as
stereo imaging was concerned, but in more recent years this has changed
due mainly to research published by Prof. Michael Williams. He found
ways to compute the stereophonic recording angle for spaced microphone
arrays of all kinds, no matter what pattern of microphone (omni, wide
cardioid, cardioid, etc.) you are using, and also other factors which
place limits on the angles and spacings that will work well. And best
of all, his findings work very well in practice. Thus it is no longer
necessary to guess, or go through much trial and error, when setting up
a recording with spaced microphones.

This is a fundamental change in the situation for two-microphone stereo
recording in general. Effective setups can now be planned and carried
out with predictable, reliable results. Many items of "conventional
wisdom" on this topic which have not actually worked any better than
random guessing can finally be discarded--though I suspect it may be
another twenty years before the recording community catches on, people
are so used to fumbling around at this.

If you're an AES member you can download his papers on this topic
(actually anyone can, though it's less expensive for AES members), or
better yet you could try to find a copy of his recent book, _Microphone
Arrays for Stereo and Multichannel Sound Recording_, volume 1 of which
has been published by Editrice Il Rostro in Italy.

Meanwhile, you can start with "The Stereophonic Zoom" courtesy of Rycote
http://www.rycote.com/products/pdf/The%20Stereophonic%20Zoom.pdf

David Satz wrote:

And best
of all, his findings work very well in practice. Thus it is no longer
necessary to guess, or go through much trial and error, when setting up
a recording with spaced microphones.
[...]
If you're an AES member you can download his papers on this topic

Prolific author... I found the shorter paper that Kurt mentioned.

From a first reading it seems that his system primarily tries to solve
"imaging". Where in the left/right axis we place the soundsources
between the speakers. He compares it to cropping a picture - how much of
the image should be seen. Get this wrong and you won't use the area
between speakers fully or distort the image so that the middle is made
sparser and the edges near the speakers are too dense with sources that
would have fallen outside the speakers. He calls this is a form of
distortion. It seems to me that we still need to solve a couple of other
balancing acts manually (front/back; direct/reverb; etc) -- Interesting
paper. Will need to read it again. Thank you both!

Lars


--
lars farm // http://www.farm.se
lars is also a mail-account on the server farm.se
aim:

I liked the sound of your ORTF recordings the best--thatnks for all the
mp3s and pix--very informative.

Philip Perkins

wrote:

I liked the sound of your ORTF recordings the best--thatnks for all the
mp3s and pix--very informative.

This is "pair 3", the close up cardioid example.

Well, I've broken pretty much all "rules" presented in discussions here.
The mics are not omni. They are cardioid (Pearl DC-96 from about 79-80)
perhaps ca 3m (10-12') out from the facade up on on the balcony behind
the console. Far too close. They're up at the pipes lower third. In
front of the opening in the facade you can see above the console.. This
is also where the smallest division of the organ is ("bröstverk"
whatever that is called in English). This can be clearly heard in the
Widor example that jumps out of the speakers near the end. The Widor
also starts out unbelievably dry. If it would continue longer you'd risk
becoming mummified from dehydration;-) It suits the Bach sample better.
Here it illustrates that an organ needs stereo. The sound jumps from
left to right wildly because that's how the organ is built. C far left,
C# far right, D next to far left, D# next to far right etc... It's the
way is does sound up there with the organist, but it's far from what it
sounds down on the floor where most listeners are.

The experiment taught me some things and posed new questions (partly
presented in this thread). Still confused, but on a higher level...

I think that "pair 4" the pseudo PZM things where a bit of a surprise
too by picking up the lower end much better than the pair I expected to
be reasonably placed (pair 1)

Lars


--
lars farm // http://www.farm.se
lars is also a mail-account on the server farm.se
aim:

Well, there's always X-Y. My fallback setup when nothing else is
working or I don't have any time for experimentation.



Lars Farm wrote:
I've had some succes recording choir and other sources with ORTF(ish)
setups. Lately I've played with AB-stereo and this has turned out less
predictable. There is a too large element of randomness in my results. I
sometimes get a washy, boxy sound. I'm not entirely sure this is just
too much reverb. Could it be some phase related effect? Sometimes this
is helped by moving in, but I've also seen it helped by moving out.
Both cardioid and omni are AKG C480, just different capsules.

Recording choir a cappella in reverberant church, are there any rules of
thumb for AB-stereo[*] Relative ORTF or just rules of thumb for
reasonable starting points? Distance between choir and mics? distance
between mics, height? How compared to ORTF?

best regards
Lars Farm

[*] Yes, I know that I like most am equipped with a set of ears. Those
ears tell me there sometimes is something wrong, but not how to solve
the problem.

--
lars farm // http://www.farm.se
lars is also a mail-account on the server farm.se
aim: