R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1] 2  All   Go Down

Author Topic: Baffled omnis processing ala Blumlein 1931  (Read 4716 times)

ted nightshade

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1272
Baffled omnis processing ala Blumlein 1931
« on: April 23, 2004, 02:26:32 pm »

Hello there,
I just got the New Stereo Soundbook and was reading Blumlein's (amazing) British Patent Application 394,325 in the appendix-

He has this circuit for processing baffled omnis, that turns low frequency phase differences into amplitude differences. Apparently the high frequencies are left as is.

Sucker that I am, I'm still trying to figure out a really gratifying way of recording stereo with two pressure omnis that's mono-compatible (pressure omnis 'cause I'm mostly recording outside, wind and that, and hey I like my lows)- this looks to be it?

If the world followed up on this one, I didn't hear about it. Anybody know what happened? Is there a reason why nobody (that I know of) does this?

If you have the New Stereo Soundbook on hand, the idea is discussed in paragraphs 29-31 of the patent application. I could probably transcribe some of it here if need be.

Thanks!
Logged
Ted Nightshade aka Cowan

There's a sex industry too.
Or maybe you prefer home cookin'?

John Klett

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 475
Re: Baffled omnis processing ala Blumlein 1931
« Reply #1 on: April 23, 2004, 04:55:39 pm »

The New Stereo Soundbook
by F. Alton Everest, Ron Streicher
McGraw-Hill/TAB Electronics; (February 1992)

that New Stereo Soundbook?

I don't have it...

here's the patent for those who want to read it

http://www.doramusic.com/patents/394325.htm

That's an eyeful -  hmmm... drawings...

well - the network is passive sum and difference...  he got a lot of mileage out of that...  oh - this is the vertical and lateral disc recording patent.  All mono in phase is lateral and all out of phase information is vertical.  The is how to make stereo record.  There's a cutterhead.  Cool.

I'm not sure what to recommend re your recording situation.  You can do spaced omni's open but mono compatibility might be a problem.  Sounds good in stereo though.  Max Wilcox (classical and jazz producer engineer) likes spaced omni's.

If you have a pair of omni's closer together - like about 5 inches apart or so - and place a baffle between them - a foot square hunk of 1/4" luan plywood with some 1/2" open cell foam on both sides is a decent place to start experimenting - you are sort of making a simple form of dummy head.  

The wavelengths at low frequencies are so long they will ignore the baffle and reach both mics.  For all intents that information is in phase.  As wavelength gets shorter the baffle becomes more significant and the mics essentially go cardioid.  Our ears are omni but all that stuff around them (head, pina etc.) make 'em more directional simply by blocking and diffracting (refracting?).

Anyway - I've done that baffle trick and it makes a rude but useable binaural recording.  You will get some bass buildup when you add the two mics so the two networks back to back with attenuators in between - shown in the drawings in the patent - can be used to drop the sum signal relative to the difference - I have not tried that but it makes sense.

Haeco (I think that is how it's spelled and that company is long gone) made a box that did a stereo sum to mono and gave you quite a bit of control over the bass buildup.  I suppose that an elliptical filter out of a cutting console could do something helpful there too.  I have some prints for one or more of those somewhere.

People use these sum and difference networks all the time to manipulate the proportion of mono and stereo - as a width control...  and for doing MS recording etc....  It's not a lost art

Logged
John Klett / Tech Mecca
http://www.technicalaudio.com

maarvold

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 853
Re: Baffled omnis processing ala Blumlein 1931
« Reply #2 on: April 24, 2004, 02:08:17 am »

At NAB there was a brochure at the Schoeps booth called Newsletter No. 6.  Although it doesn't specifically discuss what you're talking about, they discuss a bunch of mic techniques that might give you insight into what you're looking for and why you're looking for it.  Also a lot of good stuff in John Eargle's "The Microphone Book".  If you email me your address, I'll send you the Schoeps booklet.  

Mike Aarvold
Logged
Michael Aarvold
Audio Engineer

ted nightshade

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1272
Re: Baffled omnis processing ala Blumlein 1931
« Reply #3 on: April 24, 2004, 12:14:33 pm »

john klett wrote on Fri, 23 April 2004 13:55



If you have a pair of omni's closer together - like about 5 inches apart or so - and place a baffle between them - a foot square hunk of 1/4" luan plywood with some 1/2" open cell foam on both sides is a decent place to start experimenting - you are sort of making a simple form of dummy head.  

The wavelengths at low frequencies are so long they will ignore the baffle and reach both mics.  For all intents that information is in phase.  As wavelength gets shorter the baffle becomes more significant and the mics essentially go cardioid.  Our ears are omni but all that stuff around them (head, pina etc.) make 'em more directional simply by blocking and diffracting (refracting?).


So that low frequency stuff is in phase. That's good, that was one of my concerns. But I hadn't considered I might end up with more of it than I bargained for (never had that trouble with cardioids!)

Quote:


Anyway - I've done that baffle trick and it makes a rude but useable binaural recording.  You will get some bass buildup when you add the two mics so the two networks back to back with attenuators in between - shown in the drawings in the patent - can be used to drop the sum signal relative to the difference - I have not tried that but it makes sense.

Haeco (I think that is how it's spelled and that company is long gone) made a box that did a stereo sum to mono and gave you quite a bit of control over the bass buildup.  I suppose that an elliptical filter out of a cutting console could do something helpful there too.  I have some prints for one or more of those somewhere.

People use these sum and difference networks all the time to manipulate the proportion of mono and stereo - as a width control...  and for doing MS recording etc....  It's not a lost art




So that's what's going on. I'm so technically illiterate, I couldn't figure that out... I've also been trying to figure out what kinda network is involved in the MS thing, looking at the diagrams but it don't mean much to me... So I ought to be trying to find a good design for a sum and difference network. Or a really fantastic sounding prefab... where to look I wonder- is there a liability to making a super simple design? Maybe a new topic on that...

Thanks!
Logged
Ted Nightshade aka Cowan

There's a sex industry too.
Or maybe you prefer home cookin'?

ted nightshade

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1272
Re: Baffled omnis processing ala Blumlein 1931
« Reply #4 on: April 24, 2004, 12:23:32 pm »

maarvold wrote on Fri, 23 April 2004 23:08

At NAB there was a brochure at the Schoeps booth called Newsletter No. 6.  Although it doesn't specifically discuss what you're talking about, they discuss a bunch of mic techniques that might give you insight into what you're looking for and why you're looking for it.  Also a lot of good stuff in John Eargle's "The Microphone Book".  If you email me your address, I'll send you the Schoeps booklet.  

Mike Aarvold


Very kind of you! Sent you private message.
Logged
Ted Nightshade aka Cowan

There's a sex industry too.
Or maybe you prefer home cookin'?

John Klett

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 475
Re: Baffled omnis processing ala Blumlein 1931 - TRY THIS
« Reply #5 on: April 26, 2004, 12:04:31 pm »

Try the following to get a feel for this if you not already done so...

Record the pair baffled as above.

Post process with a small mixing console and generate sum and difference using channels that are in phase and phase flipped.

You'll see in figure three of the patent that there two transformers each with three windings that make a sum and difference.

Do the same thing with four faders and spilt them L, R, L, R and flip phase on R of the second pair and call that "-R".

The first L,R pair sums to a single mono buss to generate an in-phase sum ("S") signal.

The second L,-R pair sums to a single mono to generate an out of phase sum or difference ("D") signal.

You solo each channel and do level match with an oscillator.

You may have figured out that you can do this with three faders but I use four so I can "mess" with it.

On four more faders you do the same thing except you are splitting the sum and difference signals across them S, D, S, -D

You solo each channel and do level match with an oscillator.

NOW the end result is a left and right that should sound the same as what you started with but now you can play with level and equalization of the sum and difference signals before you decode them back to left and right.  You can test this with an oscillator.  If you trim everything perfectly you will end up being able to "pan across" the Left and Right inputs to this whole setup and get the same performance as you would bypassing it.

On a DIY basis you can do this with a number of opamps in a box with connectors and a battery - and record SUM and DIFFERENCE.

Opamp functions

1) one to add LEFT and RIGHT to make SUM

2) one to subtract RIGHT from LEFT to make DIFFERENCE - same as an out of phase sum

These would feed a loop out to external EQ or a recorder so you may want to buffer or use something like the Burr Brown DRV balanced driver output chips to make a balanced out.

The loop returns to calibrated level controls that will allow you to work the relative levels of SUM and DIFFERENCE.  You can do balanced receiver amplifiers on the inputs and loop returns as well...  lot's of ways to go but more amplifiers is more current and that shortens battery life.  I'd throw a hard bypass here as well.

There are two more opamps.

3) one to add SUM and DIFFERENCE to derive LEFT - RIGHT content in DIFFERENCE is out of phase with RIGHT content in SUM so RIGHT is nulled out and LEFT remains)

4) one to subtract DIFFERENCE from SUM to derive RIGHT - now the LEFT content of DIFFERENCE is out of phase with LEFT content in SUM so LEFT is nulled out and RIGHT remains)

By working the amount of difference signal that is recombined with sum you will be making a width control.

If you equalize the difference or sum signals and you are making the width frequency selective though be aware that many EQ's introduce freuquency dependent phase shifts - there are linear phase EQ's.

There are all kinds of applications for this in the studio for special effects and enhancing things.


visible
Logged
John Klett / Tech Mecca
http://www.technicalaudio.com

ted nightshade

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1272
Re: Baffled omnis processing ala Blumlein 1931
« Reply #6 on: April 28, 2004, 12:05:01 am »

Thank you so much John, I have been wondering how exactly I could "mock this up" with a mixer, and here you tell me without I even ask... I guess it really is the logical next step. Alright! This is going to keep me well entertained for a while here... seems like I should be able to use this as a mid-side process as well.

Can't thank you enough!

I've been trying to wrap my head around this, and couldn't quite make the connection... now I can refer to this, and I'm sure I'll manage it.
Logged
Ted Nightshade aka Cowan

There's a sex industry too.
Or maybe you prefer home cookin'?

josh

  • Full Member
  • ***
  • Offline Offline
  • Posts: 191
Re: Baffled omnis processing ala Blumlein 1931
« Reply #7 on: April 28, 2004, 10:29:36 am »

john klett wrote on Fri, 23 April 2004 21:55



If you have a pair of omni's closer together - like about 5 inches apart or so - and place a baffle between them - a foot square hunk of 1/4" luan plywood with some 1/2" open cell foam on both sides is a decent place to start experimenting - you are sort of making a simple form of dummy head.  

The wavelengths at low frequencies are so long they will ignore the baffle and reach both mics.  For all intents that information is in phase.  As wavelength gets shorter the baffle becomes more significant and the mics essentially go cardioid.  Our ears are omni but all that stuff around them (head, pina etc.) make 'em more directional simply by blocking and diffracting (refracting?).



This is roughly the description of Jurg Jecklin's "OSS" technique and the infamous "Jecklin Disk", although the "official" technique is a bit more refined.

The idea behind Jecklin's technique is to record directional sounds in a way that your ear/brain will be able to coherently determine the direction.  For very low frequencies (<80Hz), consider them omnidirectional.  For frequencies above that point, but whose wavelengths are still larger than the width of one's head, your brain uses mostly on phase difference, this means frequencies up to about 400-600 Hz or so, or into the midrange.  For higher frequencies, your brain uses mostly amplitude difference to determine direction.  

The Jecklin Disk is intended to address these issues by making a baffle that's essentially transparent acoustically at frequencies below about 400-600 Hz, and absorbtive at frequencies above that.  This way your brain can use phase differences in the range of frequencies where they are important (both mics hear the same sound, only difference is phase), and amplitude differences where they are important (high end...  sound coming from the right is attenuated in the left mic).

What you said about the mics behaving as if "cardiod" above the baffle's cutoff, that's true in the sense of amplitude axial response, but in terms of phase response, they are still much more linear than a cardiod, and their off-axis response is also very linear.  

With a spaced omni setup, you essentially get only phase/time differences so directionality at high frequencies is compromised.  If you use cardiod mics coincident pair then you only have amplitude difference and you compromise directionality in the lower midrange.  Depending on the source, either of these conditions may be acceptable.  If you are recording an instrument with nothing below 400 Hz then spaced omni will work fine.  If you're recording a tambourine or shaker, then coincident pair will work fine.  However, for acoustic guitar or drum kit, you can't beat the OSS technique.

This all works fine for listening with headphones but things become very problematic when summing to mono.  You are intentionally introducing a phase difference with the OSS-type technique simply by virtue of the spacing of the capsules.  This phase difference is going to cause comb filtering and cancellation, other issues when summed to mono.  The two solutions to this problem are either to use spaced omni with enough space between the mics such that the secondary sound in any mic sounds like ambient sound when summed to mono (Haas effect), or coincident pair (or M-S, same thing) which sums to mono perfectly.  

The other major problem with any stereo recording technique is the problem of playback on stereo loudspeakers, where the variability of listening environments, reflected sound, and crosstalk (the left ear can easily hear the right speaker) means that the only things that are going to translate accurately are mono sounds panned hard left or right so that they really only come from one speaker at a time.  All of these detailed phase techniques, all that, goes out the window.  Wide spaced omni technique translates best to loudspeaker playback imho but also is maybe the worst for mono summing and definitely the worst for headphones.  I am still grappling with this issue myself.  Everything is a compromise.  I guess this is why you must know your target audience, how & where they listen to music, and go for it.  


See ya-

ted nightshade

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1272
Re: Baffled omnis processing ala Blumlein 1931
« Reply #8 on: April 28, 2004, 12:28:28 pm »

Thanks Josh!

Yes, I have been trying to reconcile John's first reply with some information that reads more like your reply.

I have been recording almost exclusively with one main mic, which with the cooperation and hard work of the performers, one of whom is me, has enabled me to get superb spectral balance from ultra-low to ultra-high, precise phase, and generally a very lifelike sound. Stereo definitely seems to be the next frontier for me, but I am unwilling to make significant sacrifices in the spectral balance and the phase picture, and I still want complete mono compatibility. Not an easy thing to accomplish! Plus, the damned sweet spot is so small, and just a millisecond or a foot or two compromises the whole picture...

I've been reading and researching a lot on the subject, and suffice to say, your comments are most relevant and most welcome.

Mono has enabled me to make recordings that sound 3D (by mixing in a little bit of distance mic), balanced, generally very nice, on everything from headphones to boom box to car stereo to nice hi-fi. Most people do not notice that it's in mono- a typical comment from those who do is that it could be "more stereo" or "a wider stereo spread'.

There are basically two kinds of ensembles I'm recording- (1) vibraphone and acoustic guitar and male vocal stuff, which only goes as low as the acoustic guitar, and (2) full range stuff with concert bass drums and keys bass, which goes clear down to 20 hz in a big way.

Right now I'm looking at different ways of using a close mic or pair, coincident or M/S (a natural place to go from one-mic), and a distant pair for dimension and ambience.

The mind reels!
Logged
Ted Nightshade aka Cowan

There's a sex industry too.
Or maybe you prefer home cookin'?

John Klett

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 475
Re: Baffled omnis processing ala Blumlein 1931
« Reply #9 on: April 28, 2004, 12:38:46 pm »

josh wrote on Wed, 28 April 2004 10:29


This is roughly the description of Jurg Jecklin's "OSS" technique and the infamous "Jecklin Disk", although the "official" technique is a bit more refined. ...




That is all correct... the amplitude patern is cardioid but with better phase response...  the baffle I described does nearly vanish a lower frequncies.

I have never really measured this myself but I would think there is phase shift around the frequncy range at which diffraction around the edges of the baffle becomes significant and around the critical frequency of the baffle itself which would that point below which it becomes transparent.  The baffle is damped so this would make that transition smoother... ?

Logged
John Klett / Tech Mecca
http://www.technicalaudio.com

josh

  • Full Member
  • ***
  • Offline Offline
  • Posts: 191
Re: Baffled omnis processing ala Blumlein 1931
« Reply #10 on: April 28, 2004, 02:50:25 pm »

john klett wrote on Wed, 28 April 2004 17:38

josh wrote on Wed, 28 April 2004 10:29


This is roughly the description of Jurg Jecklin's "OSS" technique and the infamous "Jecklin Disk", although the "official" technique is a bit more refined. ...




That is all correct... the amplitude patern is cardioid but with better phase response...  the baffle I described does nearly vanish a lower frequncies.

I have never really measured this myself but I would think there is phase shift around the frequncy range at which diffraction around the edges of the baffle becomes significant and around the critical frequency of the baffle itself which would that point below which it becomes transparent.  The baffle is damped so this would make that transition smoother... ?




I have not measured it either.  The baffle, from a 180-degree perspective (as in dead-on-axis vs dead-off) seems to function basically like a single-pole high pass filter whose corner frequency is dictated by the dimensions of the baffle and the resonant frequency, mass etc., characteristic of the material.  The Jecklin Disks you can get from Josephson (which are very high price considering how easy it is to make one) are (I believe) acrylic about 8-10mm thick and 280mm diameter (about 12") and it results in a pretty low cutoff frequency compared with my Jecklin Disk which is 1/4" plywood.  Of course mine is much lighter weight and that makes it possible to mount it on a boom stand with two mics and it doesn't tip over.

Anyway...  my guess is the difference in mass may have up to a 1/3 to 1/2 octave effect on the corner frequency which is not all that much considering what you're trying to do with it.  Sound localization within one particular person is pretty precise in terms of which frequencies cue what and so forth, but spread across a spectrum of people with different hearing and physical features, it's much more approximate.  Not only that, your ear is very adaptive, and in the absence of a full spectrum of localization information, your ear will cue on whatever is dominant and available...  so having SOME midrange phase difference information that jives with the HF delay characteristic is definitely crude at best, but more than good enough to convince your ear and brain of the direction from which a sound is coming.

The main thing I was pointing out is that a cardiod mic has phase error off-axis which can really screw up stereo imaging, and in this sense, an omni mic with an absorbtive baffle behind it is very much preferable.  This is not to mention the improved off-axis frequency response linearity of most small-diaphragm omnis compared with any cardiod.  IMHO for stereo recording, considering the critical nature of phase in the midband, omnis are the best choice.

The other effects of the baffle are things like reflections off of the edge, reflected sound from the baffle that makes it through the absorber (HF energy coming from the right side, hits the right side of the baffle and reflects back into the mic only less than 1ms behind the original sound) etc.  Depending on what type of acoustic foam material is used, etc., you can have a HF-reflective surface.  This tends to "blur" the HF response a little bit...  you get sort of a HF "echo" in all of the sound that is sort of like the kind of "smeary" attack sound you get multimiking a drum kit due to differing time of arrival of HF energy into multiple mics.  I think the Josephson Jecklin Disk like the one in the photo accompanying the reprint of the tech note on the Josephson site, which appears to have thin closed-cell foam, would be very bad for this.  For my Jecklin Disk I used open cell acoustic foam in 1.25" thickness, which imho works much better, but still there is *some* of this effect.  You have edge diffraction and all of the other acoustical side effects of a baffle in there.  Welcome to physics.  Smile

Suffice to say, even crude stereo recordings with all of these kinds of problems can sound extremely good.  However, to do spaced omni or coincident pair involves a real significant compromise in image quality regardless of other factors, and these little side effects of the baffle in a Jecklin Disk are nowhere nearly as important to the sound as missing phase differences for X-Y or missing amplitude differences for A-B miking.

IMHO...

I'll still vary my recording technique between X-Y (which I almost never use), A-B (spaced omni), ORTF and OSS (Jecklin Disk) or some other hybrid of all of that, depending on the instrument and other factors.  I believe the OSS/Jecklin Disk technique makes the most consistently usable stereo recording.

josh

  • Full Member
  • ***
  • Offline Offline
  • Posts: 191
Re: Baffled omnis processing ala Blumlein 1931
« Reply #11 on: April 28, 2004, 03:05:12 pm »

ted nightshade wrote on Wed, 28 April 2004 17:28

Thanks Josh!

I have been recording almost exclusively with one main mic, which with the cooperation and hard work of the performers, one of whom is me, has enabled me to get superb spectral balance from ultra-low to ultra-high, precise phase, and generally a very lifelike sound. Stereo definitely seems to be the next frontier for me, but I am unwilling to make significant sacrifices in the spectral balance and the phase picture, and I still want complete mono compatibility. Not an easy thing to accomplish! Plus, the damned sweet spot is so small, and just a millisecond or a foot or two compromises the whole picture...




OK, well you might try some things.  Depending on what you want to DO with the stereo recording, or why you want to do a stereo recording in the first place, you might consider some alternate techniques.

If you've got something working good, then I'd say stick with it, at least as a basis.  Remember though that one mic, panned straight up, played back over a stereo system, is not mono.

Recall the Haas effect which goes something like, if you hear a sound and then hear it again delayed 7-10 ms or more, then your mind basically ignores the second sound as if you didn't even hear it.  Filters it out.  This is the way that "crosstalk" is eliminated in old Dolby Surround matrix encoding...  the rear channels are delayed 10+ ms which means that the front-channel sound that is present in them is ignored by your brain.

So keeping this in mind, if you were to use your regular mono recording technique, and blend in some stereo recording of the same sound that you know is at least 10 ms behind in combined time of arrival plus delay you might intentionally introduce, then you will get the ambient sound you might expect from a "stereo" recording, along with some localization effects, along with the stability of your mono sound.  In this case I'd suggest using a spaced omni setup for your "stereo" recording and putting the mics 10' away from the source (or less, and add delay...  but that can get very problematic).  Space the omni mics about 6-10' apart.   Remember if the mics are 10' apart, then to make them 10' away from the source they only have to be set up in a line about 6' or so in front of the player...

The result is a very exaggerated stereo recording from the ultra-wide spaced omni, but it's way in the diffuse field.  Combined with a mono sound you end up with a pretty natural stereo sound (the mix in of the mono helps tame the very exaggerated stereo recording).  When you sum this to mono, the stereo sound is delayed enough to not cause audible phase anomaly, and also to be ignored or regarded as an ambient reflection due to the Haas effect.  Done right, your mono mixdown should sound pretty dry.

Anyway, that's something to try.  I think you could also build a Jecklin Disk but it's really only useful if you're willing to give up mono compatibility for the most part.  I think coincident pair or Blumlein or some other technique like that might result in a mono compatible stereo recording, but imho the quality of the stereo field is nowhere near as good as with techniques with some space between the mics.

The other thing to remember about Blumlein or coincident pair recording is that your on-axis (centered) sound source is actually at least 45 degrees off-axis to any mic, and you will have the inherent phase and frequency response issues from any directional mic when recording off-axis.  M-S is better in this regard since on-axis is on-axis to one mic diaphragm (the "M"), even though in theory they are mathematically equivalent to coincident pair, that equation assumes perfect microphone linearity which is impossible.  This issue is almost completely moot with omni mics.

Whew.  That's way too much rambling about stereo recording.  The best summary is "try it!"

ted nightshade

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1272
Re: Baffled omnis processing ala Blumlein 1931
« Reply #12 on: April 28, 2004, 10:21:42 pm »

josh wrote on Wed, 28 April 2004 12:05


If you've got something working good, then I'd say stick with it, at least as a basis.  Remember though that one mic, panned straight up, played back over a stereo system, is not mono.


[/b] I'd be interested in some of the theory behind this, but having sampled these goods many times and many times compared to one-speaker mono, I think I have a pretty good idea of what happens...

Quote:


Recall the Haas effect ... if you were to use your regular mono recording technique, and blend in some stereo recording of the same sound that you know is at least 10 ms behind in combined time of arrival plus delay you might intentionally introduce, then you will get the ambient sound you might expect from a "stereo" recording, along with some localization effects, along with the stability of your mono sound.  In this case I'd suggest using a spaced omni setup for your "stereo" recording and putting the mics 10' away from the source (or less, and add delay...  but that can get very problematic).  Space the omni mics about 6-10' apart.   Remember if the mics are 10' apart, then to make them 10' away from the source they only have to be set up in a line about 6' or so in front of the player...

The result is a very exaggerated stereo recording from the ultra-wide spaced omni, but it's way in the diffuse field.  Combined with a mono sound you end up with a pretty natural stereo sound (the mix in of the mono helps tame the very exaggerated stereo recording).  When you sum this to mono, the stereo sound is delayed enough to not cause audible phase anomaly, and also to be ignored or regarded as an ambient reflection due to the Haas effect.  Done right, your mono mixdown should sound pretty dry.

Anyway, that's something to try.  I think you could also build a Jecklin Disk but it's really only useful if you're willing to give up mono compatibility for the most part.  I think coincident pair or Blumlein or some other technique like that might result in a mono compatible stereo recording, but imho the quality of the stereo field is nowhere near as good as with techniques with some space between the mics.



Thanks much for the "rambling"... The approach you describe is one I have tried some variations on, and I like the idea. I have had remarkable results from stereo distance mics with a mono main mic. Typically I use very, very little of the distance mics, but get a nice dimensional effect that way. After some reading I am thinking the "incoherence" of the stereo distance mics has a definite advantage over a mono distance mic, and my work has told me something like that, although it's more definite when I have stereo main mics too. But then the main mics don't sound as good... I am also considering using a Jecklin disc for the stereo distance mics, so I am still much interested in your Jecklin disc construction experience. Thanks for sharing some of that.

I suppose with the distance omnis 10 feet apart, there's not a lot of point in baffling them, at least not with a head-sized baffle, but in the spirit of experimentation, I'm liable to try it with some big old baffle and see.

Today I did some first experiments with Mid-Side, using my usual mono main mic, and adding the fig. 8 side mic. Should be a lot of fun to feed it into a sum and difference type mock-up- thanks again for that, John!

I found this bit on M/S today: http://www.uneeda-audio.com/ms-mat.htm I'd like to find a super-puriste design sans op-amps, passive, and use one of my nice mono tube mixers to bring it up to level.
Logged
Ted Nightshade aka Cowan

There's a sex industry too.
Or maybe you prefer home cookin'?

josh

  • Full Member
  • ***
  • Offline Offline
  • Posts: 191
Re: Baffled omnis processing ala Blumlein 1931
« Reply #13 on: April 29, 2004, 10:04:03 am »

ted nightshade wrote on Thu, 29 April 2004 03:21

josh wrote on Wed, 28 April 2004 12:05


If you've got something working good, then I'd say stick with it, at least as a basis.  Remember though that one mic, panned straight up, played back over a stereo system, is not mono.


[/b] I'd be interested in some of the theory behind this, but having sampled these goods many times and many times compared to one-speaker mono, I think I have a pretty good idea of what happens...



What I meant was that while it may sum to mono and be playable without problems on a one-speaker playback system, it only "sounds" mono when you hear it in a stereo playback system because your brain is really ignoring a lot of information.  The thing is that the ambient sound and reflected sound from the left speaker is going to not jive with that from the right speaker and certainly when you turn your head or whatever, you can easily localize that it's not a mono, centered source, but it's two identical sources coming from different places in the room.

With a well-done stereo recording, you can get a much more stable and effective centered image.  But this by definition won't sum to mono without big problems.

Quote:




Thanks much for the "rambling"... The approach you describe is one I have tried some variations on, and I like the idea.

<snip>

I am also considering using a Jecklin disc for the stereo distance mics, so I am still much interested in your Jecklin disc construction experience. Thanks for sharing some of that.



Well, the point of spacing the mics apart is so even the lowest frequency you record will have at least one wavelength between the mics, such that there's no phase problem when you sum it to mono.  If the mics are less than one wavelength apart, then you introduce phase problems when you sum them.  This is the big drawback of the Jecklin Disk.  FWIW speed of sound is about 1150 fps depending on temp and humidity, which means your 10-foot wavelength is at 115 Hz.  So frequencies below that could still have phase problems even at 10' spacing.  As you can see, to avoid phase problems with spaced microphone technique, you have to get a BIG distance between mics, or just determine what's appropriate for the instruments you record.  The fundamental of the open A string on the guitar is 110 Hz and this is near the bottom end of most baritone vocalists' performances (even though most baritones can sing a note at least 1 whole step below that).

Quote:


I suppose with the distance omnis 10 feet apart, there's not a lot of point in baffling them, at least not with a head-sized baffle, but in the spirit of experimentation, I'm liable to try it with some big old baffle and see.



Yeah the head-size baffle will have almost no effect at all.  For a baffle to have the effect of blocking crosstalk between the mics like a Jecklin Disk, the radius of the baffle has to be equal to or larger than the distance from the mic diaphragm to the baffle.  In this case, you'd do just as well to place a baffle near each mic (use two).  This will just increase the stereo separation.  That's a good thing!  Also using baffles and absorbers around the mics like this helps you control which ambient sound you record.  If you're going for mostly directional cues and not really reverberant field, then putting absorbers on the outside of the mics would help reduce the reverberant field they pick up (as opposed to between them).  FWIW I have a piece of carpet tacked to the wall in my studio where I typically put my ambient mics so I don't get as much early reflection into the mics which can really upset the imaging.

This is a very interesting topic to me, and I have yet to really deal with the problem of mono compatibility of my recordings.  I guess I should.  

ted nightshade

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1272
Re: Baffled omnis processing ala Blumlein 1931
« Reply #14 on: April 29, 2004, 11:09:01 am »

I'm grateful that you're interested, Josh!

Perhaps what you are saying about the two-speaker "mono" situation explains why it sounds so stereo sometimes... I suppose it must be our brains that sort out that confusion in logical ways, like having the chimes or vibes go left to right, instead of randomly all over... handy things, those brains...


My mid/side experiment yesterday, my first, really came off in an encouraging way- great to be able to tack on a little stereo field to my established one-mic techniques! Seems like it could be the best of both worlds, but me being me, I still want to get some distance mics involved for that great depth dimension effect.

One further idiosyncrasy to my situation- I'm recording outdoors, and my prime go-to mics, tube LDCs, are super vulnerable to air movement and wind. A lot of times I can get away with it, but there's always the possibility of a great take getting blown away on the wind unexpectedly... so I'm dreaming that I can come up with a distance omnis technique that can serve as a main pair backup in case the close pair get swamped by the breeze.

I realize I'm getting greedy here- I want it all! And I realize at some point I have to prioritize and choose. But not until I have to!

and...

...So for full-bandwidth material, those omnis would have to be 30' apart! wow...

Then there's this LittleLabs IBP... I've had bizarre things happen when trying to reconcile spaced omnis in phase this way, but perhaps with the omnis far enough apart, all I have to get is the lows, and the highs can take care of themselves. That should be worth a try, anyway. Really interesting little box, and the upcoming continuous phase adjust on the Manley mic-utility box should be worth a spin too, for sure...

Logged
Ted Nightshade aka Cowan

There's a sex industry too.
Or maybe you prefer home cookin'?
Pages: [1] 2  All   Go Up