Phase issues with Blumlein Capture

[Peacock]

We recently recorded a gospel quartet with blumlein capture and encountered phase issues at mix. We used a stereo AKG 426 mic and everything was done at once with no overdub in blumlein. I do not believe we had a phase invert accidently engaged on the pre.

I think the phase issue maybe from capture outside a 90 degree span. Could that cause phase issues with blumlein?

Here is a link to the recording.

Mastering for the Amateur

[Corran]

Did you have a singer on each side of the lobe of the figure 8?

If you have two singers on each side with similar vocal ranges, then you can get phase issues since the back of the figure 8 will be in opposite phase of the front.

Otherwise you shouldn't have any problem with Blumlien.

[Peacock]

Originally Posted by Corran

Did you have a singer on each side of the lobe of the figure 8?

If you have two singers on each side with similar vocal ranges, then you can get phase issues since the back of the figure 8 will be in opposite phase of the front.

Otherwise you shouldn't have any problem with Blumlien.

No. The quartet has the usual ranges from bass to tenor. I did track them with greater than 90 degree spread. I can't figure where the phase issue occured.

[boojum]

I am new to Blumlein myself. I use it in once instance to drop into the center of a BG jam. And, yes, the guys in the back side of the mic appear on the opposite side of the sound field in the front. Other than that I have had no phase issues.

BTW, I have read on this forum this the effective recording angle is closer to 75 degrees. I have not checked this myself but will today in Eargle's Microphones.

[Corran]

Originally Posted by Peacock

No. The quartet has the usual ranges from bass to tenor. I did track them with greater than 90 degree spread. I can't figure where the phase issue occured.

Perhaps you can post an example? Bass to tenor is not a huge range, likely you have at least some overlap between singers, thus making some phase problems possible.

My first suggestion would be to back up the mic a little to close up the recording angle.

[Audiop]

If the recording angle is greater than 90 degrees the sources outside 90 will be picked up by both capsules but with opposite polarity.

Having two performers centered on each side of one capsule should not cause any problems.

Blumlein = 90 degrees for discrete sounds.

Ambience "out of 90" is ok and can add space in two channel play back.


/Peter

[2pulse]

Blumlein can pick up sources from the front and back quadrants simultaneously without any phase issue. The "opposite polarity" of the back quardrant doesn't matter, since separate sources in front and back shouldn't have a phase relationship. It gets dicey when you capture sources in the left and right quadrants, because it will be picked up by one positive capsule and one negative capsule. The closer the source is to the middle plane, the more completely out of phase it will be between the two channels.

Basically, if you have a group encircling a Blumlein pair, it will not be mono compatible. They need to be kept within the front and back quadrants.

[Tom McC]

Originally Posted by Peacock

I think the phase issue maybe from capture outside a 90 degree span. Could that cause phase issues with blumlein?

Firstly, a most enjoyable song and performance. Congratulations.


Playback of the lead tenor clearly shows a channel phasing problem. You can observe that by (a) uneasy vague imaging, smeared all over the space between the loudspeakers, (b) low correlation in a spectral scope display (e.g. RME's Digicheck), (b) large volume drop when toggled to mono.

Inverting a playback channel, snaps this tenor snap back into sharp center focus...but all four voices still never become well-focused (which should certainly be achievable in a Blumlein array). I found the bass disconcertingly fuzzy to locate. Assuming no mic miswiring or inversion in signal chain, your likely culprit, as others have already pointed, is your placement of singers outside the 90 deg arc.

For illustration, if a singer is wrongly sited 90 deg to the right of straight-ahead, then he will be picked up at equal level by the front lobe of the R-pointing Fig-8, and by the rear lobe (thus phase-inverted!) of the L mic. Thus he would display with the aforementioned centered-smeary-image. (Actually, it's advisable to array the performers in an even narrower arc than 90 deg. For Blumlein, the Sterephonic Recording Angle is about 75 deg...in simplified terms, staying within that arc avoids all periphery singers in the 75 - 90 deg segment having their soundstage image collapse into the nearest playback speaker.)

It would narrow down all conjecture, if you could give more details on the placement of the four performers.

[2pulse]

Keep your sources in the green area.

[Peacock]

Thanks for everyones suggestions

[boojum]

I just read Bartlett's Recording on Location on the subject. An excellent book and my go to book. He says the same thing the little diagram does: outside of that 90 degree span you are asking for trouble as it will introduce phase issues. I will check Eargle's book when I can find it.

[7rojo7]

I'd like everyone to get extremely simple here
These were experiments
well documented
Blumlein was a genius with a pair of mics
It can even be proved mathematically
so what
XY is your starting point
your field of capture is the most important
changing the angle of XY depending on?
this changes every time. (critical distance vs response of proposed array)
anything else is most likely ancillary and negatory to the above
I prefer an omni under an XY pair in the center and half omni's as flankers
I don't trust any encoding/decoding scam as much as I would a discrete array
YMMV

[Tom McC]

Originally Posted by 7rojo7

I'd like everyone to get extremely simple here.

But your recommendation below doesn't look too simple: :-)

Originally Posted by 7rojo7

..an omni under an XY pair in the center and half omni's as flankers.

:-)

Originally Posted by 7rojo7

XY is your starting point.

Blumlein is just a subset of XY

Originally Posted by 7rojo7

I don't trust any encoding/decoding scam as much as I would a discrete array

There's no encoding/decoding involved in conventional Blumlein...unless you elect to rotate the Blumlein array 90 deg counterclockwise and run it as MS.

[Tom McC]

Originally Posted by Tom McC

...unless you elect to rotate the Blumlein array 90 deg counterclockwise and run it as MS.

Oops - should have been rotate 45 deg conterclockwise.