Spaced figure-8s?

[matyas]

There was a thread a few weeks ago in which the subject of spaced figure=8s came up, and since then I've heard it discussed a few times outside of this forum. I'm curious about this technique - when might it be appropriate, and how does it compare to things like ORTF or spaced omnis, or even other figure-8 techniques like Blumlein or M/S? And how would you avoid getting a hole in the center?

[Baithak]

Hi Matyas:
For the past year, I have been miking small ensembles with a pair of Royer 122s spaced about 10 inches apart and about 3 feet from the ensemble and getting excellent results of imaging. I have recorded typical Indian vocal classical music, a harp and a flute with tabla, an acoustic/hawaian guitar, and so on with this method. There is no hole in the middle when the mics are this close (I used a pair of DPA 4003 omnis about 2 feet apart for comparison).

The only disadvantage that I have felt so far is that if the vocalist has a habit of moving their heads too much, the image moves ... If I am recording a new singer or a singer with such a known habit then I stick to the backup recording which I do using a SF-12 in MS.

The advantage of using the SF-12 in MS is that one time I was able to use the M mic of the SF-12 (ignoring the S) for most of the sound and blend-in the spaced pair R-122s for an even more solid center image.

Hope this helps,
Baithak

[staudio]

Originally Posted by Baithak

Hi Matyas:
For the past year, I have been miking small ensembles with a pair of Royer 122s spaced about 10 inches apart and about 3 feet from the ensemble and getting excellent results of imaging. I have recorded typical Indian vocal classical music, a harp and a flute with tabla, an acoustic/hawaian guitar, and so on with this method. There is no hole in the middle when the mics are this close (I used a pair of DPA 4003 omnis about 2 feet apart for comparison).

The only disadvantage that I have felt so far is that if the vocalist has a habit of moving their heads too much, the image moves ... If I am recording a new singer or a singer with such a known habit then I stick to the backup recording which I do using a SF-12 in MS.

The advantage of using the SF-12 in MS is that one time I was able to use the M mic of the SF-12 (ignoring the S) for most of the sound and blend-in the spaced pair R-122s for an even more solid center image.

Hope this helps,
Baithak

Are the 122's angled or parallel with each other when using the close (10") spacing?

[John Willett]

Originally Posted by matyas

There was a thread a few weeks ago in which the subject of spaced figure=8s came up, and since then I've heard it discussed a few times outside of this forum. I'm curious about this technique - when might it be appropriate, and how does it compare to things like ORTF or spaced omnis, or even other figure-8 techniques like Blumlein or M/S? And how would you avoid getting a hole in the center?

This is the Faulkner Array - and is two figure-8s in parallel.

Tony Faulkner invented this when he had to record in a church that had good reverberation, but a horrible flutter echo across the church. This method meant that he could take advantage of the reverberation and cancel out the flutter-echo.

PDF is HERE.

[David Rick]

Although the cannonical "Faulkner array" uses 20 cm spacing, one shouldn't take this as graven in stone. It's just what Tony Faulkner used for his particular circumstance. Simulating this configuration in Image Assistant 2.0 shows that it has an total working angle approaching 180 degrees. That's way too wide for most situations, so be prepared to increase the spacing until it works in your particular venue. You might well end up with something closer to 40 cm.

David L. Rick
Seventh String Recording

[Baithak]

Originally Posted by staudio

Are the 122's angled or parallel with each other when using the close (10") spacing?

The first few times I tried a variety of test setups with angles both acute and obtuse and ultimately realized that the only way to may it work by using the mics parallel to each other (sorry I am not aware of Faulkner's work but will surely check it out).

If the mics are angled towards each other (in an acute angle) in some cases the center image improved a bit ... but in general the parallel config gave the best result.

In the setup I am using, the mics are now about 22cm apart.

Baithak

[aracu]

Originally Posted by Baithak

The advantage of using the SF-12 in MS is that one time I was able to use the M mic of the SF-12 (ignoring the S) for most of the sound and blend-in the spaced pair R-122s for an even more solid center image.

Hope this helps,
Baithak

Interesting and creative use of ribbons. I'd love to hear a sample. I've often wished that my sf24 consisted of two mono mics instead, to experiment with. They have designed such a phantom powered mic but don't sell it.

[MATTI]

Originally Posted by Baithak

The first few times I tried a variety of test setups with angles both acute and obtuse and ultimately realized that the only way to may it work by using the mics parallel to each other (sorry I am not aware of Faulkner's work but will surely check it out).

If the mics are angled towards each other (in an acute angle) in some cases the center image improved a bit ... but in general the parallel config gave the best result.

In the setup I am using, the mics are now about 22cm apart.

Baithak

Yes, that comes to my mind also ( the angles ) debening...

Matti

[Baithak]

Originally Posted by aracu

Interesting and creative use of ribbons. I'd love to hear a sample. I've often wished that my sf24 consisted of two mono mics instead, to experiment with. They have designed such a phantom powered mic but don't sell it.

Sorry "Aracu" I am still living with dialup internet and cannot fulfill your request for a sample. However, as I said earlier, the R-122s are 22cm apart, the SF-12 is dead center between them in MS mode. With four channels being recorded to the 744T, I essentially have the following possibilities in "post":

1. Only R-122 pair,
2. Only SF-12 MS decoded
3. Blend of the M mic of the SF-12 and the R-122 Pair.
4. For mono use, just the M mic of the MS.

Re. height of the mics, my ensembles feature musicians who are seated on the floor; so the mics are about head height of the lead performer tilted downward about 20-30 degrees depending on the distance from the ensemble.

Re. the distance from the performers, besides the balance/blend as mentioned in the Faulkner Array PDF, one has to consider the reverberation of the room and the amount of ambient vs. direct sound you need. For my use in my room, a distance of about 2-4 feet from the lead performer suffices (it is a fairly live, 17 by 27 room with a sloped ceiling 14 feet high going up to 23 feet in one direction).

I hope you try it.

Best wishes,
Baithak

[Corran]

This is a cool technique, I will definitely be trying it next time I have a suitable chance. I actually recorded a bassoon recital last night that this might have been perfect for (large church, lots of reflections, typical).

[matyas]

Good discussion - I'm learning a lot. Thanks, as always, to everyone who's contributed. One thing I've been wondering about is summed up in Sengpiel's discussion:
"What worries me is that I find it hard to imagine how the full stereo stage from left loudspeaker to right loudspeaker can ever be filled using this microphone array. With the Faulkner Array only the arrival time differences can have an effect, given the narrow microphone base-line of 20 cm (8 inches), since level differences with bidirectional microphones will be, at best, marginal. Lateral (from the side) sound sources are always strongly attenuated by the characteristic figure-of-eight pattern."

Obviously, this can be overcome, since Faulkner and others have used it successfully. What would be some situations in which this technique would not work? It also seems like many people use additional mics - can this work with just two mis on a large ensemble? (I do think that it would work nicely on chamber music.)

[John Willett]

20cm is roughly the same spacing as your ears.

And - the technique was specifically designed for a situating where the room had a nasty side flutter echo.

[sonare]

Originally Posted by Corran

I actually recorded a bassoon recital last night that this might have been perfect for (large church, lots of reflections, typical).

But this is NOT what the "Faulkner Array" is for-- he wanted to keep the reverb but eliminate the side-to-side flutter echo.

I would think that a bassoon in a large church would be the perfect situation for ORTF or NOS-- to make the space seem more "bassoon-sized."

Rich

[David Spearritt]

Originally Posted by John Willett

20cm is roughly the same spacing as your ears.

Ear spacing is meaningless when referring to loudspeaker playback. The two recorded channels are loudspeaker signals not ear signals.

[antti]

Nice technique when recording an string ensemble.. especially with a pair of 4038s.

[John Willett]

Originally Posted by David Spearritt

Ear spacing is meaningless when referring to loudspeaker playback.

Not necessarily - as the signals still arrive at your two ears.

[joelpatterson]

I wonder... there must be an obscure Indian philosophy that posits an ear in the middle of your forehead....

[Corran]

Originally Posted by sonare

But this is NOT what the "Faulkner Array" is for-- he wanted to keep the reverb but eliminate the side-to-side flutter echo.

I would think that a bassoon in a large church would be the perfect situation for ORTF or NOS-- to make the space seem more "bassoon-sized."

Rich

Right, though it seems to me this technique will eliminated some reverb coming from the sides. I actually used an ORTF pair and an A-B pair, and mixed accordingly. The ORTF pair was too up-front and the omni pair too verby. It came out really well mixed, but I would have loved to try this just to hear the differences, you know? Another stereo technique is just another tool to try in different situations.

[mljung]

Nice thread.

I have always thought that spaced Fig8 was interesting, but I have made too few experiments myself. Once did it for a String Quartet and as far as I recall it came out fine, but they were combined with two DPA 4006 [flanked, as in a decca-tree] so it was the mix between the two sets that made the sound.
High humidity in the room "drowned" one of the Fig8 microphones during the performance so it buzzed instead. I have to experiment with this kind of setup again [this time using my Milab DC-196, which has a pretty good Fig8 pattern]

[Baithak]

Originally Posted by matyas

-snip- One thing I've been wondering about is summed up in Sengpiel's discussion:
"What worries me is that I find it hard to imagine how the full stereo stage from left loudspeaker to right loudspeaker can ever be filled using this microphone array. With the Faulkner Array only the arrival time differences can have an effect, given the narrow microphone base-line of 20 cm (8 inches), since level differences with bidirectional microphones will be, at best, marginal. Lateral (from the side) sound sources are always strongly attenuated by the characteristic figure-of-eight pattern."
-snip-

The level differences between the two mics matter depends on the distance of the ensemble to the mics. At distances of less than say 8 feet, the angle of incoming sound from the two extremes of the soundfield will be significant enough to have level differences as well as arrival differences. As you can see from Royer's chart for the 122 for angles greater than 30 degrees of dead center (either side) the mic attenuates the signal. Higher the frequency, higher the attenuation.

Regards,
Baithak

[audibell]

Originally Posted by joelpatterson

I wonder... there must be an obscure Indian philosophy that posits an ear in the middle of your forehead....

Yes, that's right, it's called Coincidentyclops, a persistent theoretical condition which assumes all sounds are best matrixed within the center of the skull.

[Legacy]

FWIW, I once tried a widely spaced pair of fig.8s directly in place of my usual omnis for fun on a concert.

Aside from the tasty coloration of the particular ribbon mics, the dominant difference was how it presented front to back stage depth. The Fig8s had much less front to back depth (a little unnatural for my taste) but it was interesting to hear how much closer to me the inner strings and winds sounded.

In some very reverberant spaces I've had issues with the exaggerated front to back depth spaced omnis can give and was quite surprised how differently it was captured without the side information. Wondering in my general lack of understanding if i can chalk it up to the side info.

-Silas

[7rojo7]

Coincidentyclops. brilliant

Figure 8s seem to be the opposite of what I look for in a far field sit

Though, I like this with 8s or hypercards if i have to use them

L mic points R and R mic points L, imagine a flashlight or even 2, the danger is losing the center back.

If I try spacing 2 8s, they've always been about 3 meters back and 4 meters wide and I miss the center front, closer and the whole center goes.

If the performance is good, you're lucky

google the 'Hamasaki square' for arrays that use multiple figure 8 patterns


It would be nice to have a mic that captures the middle with the character of a ribbon, the lows with the power of an omni and the highs with the directionality of a cardiod. But having all these mics separately helps too, get out the laser rule and the slide rule

[MATTI]

8 is the tighest polar pattern you get without intrference tube with their problems

Matti