Topic: New inexpensive ambisonic mic out of India (Brahma)

[Gutbucket]

The mic is called the Brahma. Like other 1st order ambisonic mics (Core Tetramic, Soundfield SPS200) the output is 4-channel A-format which requires decoding on a computer for use.

It was developed by Umashankar Nakul Sood and is available from his Indian company called Oomagamma, either alone or as a kit which includes a modified Zoom H2 with an Amphenol connector on top to attach the mic directly or via extension cable: http://www.oomagamma.com/brahma_kit/brahma_kit.html

They have some other interesting mics as well.. like one called the plate mic, a unidirectional omni phased array.

My take?
Looks like a DIY mic builders dream project that's come to market. In some ways the full kit looks along the lines of Chris Chuch's R-09 mod with his mics sticking out the top as replacements for the internals, but the mics is more involved and is detachable.  I haven't heard samples.. except for some ambisonic research clips from researchers in Italy made underwater! using a special hydrophone version.  It is very simple and inexpensive for an ambisonic recording system, but I have questions about the sound quality, build quality, realiability and especially its suitability for music.

It looks to me like the mic itself uses four elecret capsules arranged in the typical tetrahedral configuration, but in this case they are omnis embeded in the surface of a solid sphere.  Using many, many omni capsules embeded in the surface of a sphere has been one focus in research into higher order ambisionic microphones. Example: the Eigenmike. The problem with using omnis on a small surface is the extreme amounts of differential gain required at low frequencies for high orders. I imagine this is also a problem with a first order implementation, so I wonder how well this works in practice at low frequencies.  Capsule self noise may also be an issue.  Cool to see this offered for ambisonic experimenters, and it seems to have already proven useful for researchers.

Wonder if they'll make a direct to tascam DR2d mod?

[ashevillain]

The Octava 4-D is close to that in price as well: http://www.oktava-online.com/shop/view_prod.php?id=194

[Gutbucket]

Wasn't aware of the Octava 4 capsule mic, thanks.  Looks like a rather large distance between capsules though, which introduces aliasing problems at high frequencies. That's sort of the opposite problem to limited low frequency discrimination from omnis too close together on the sphere.

Jon, thanks for confirmation of my suspicion about the low frequency issue.  No problem for applications where its's cool to have everything go mono below 1-3kHz I suppose.  The ambisonic hydrophone application was to directionally locate and track boat (prop generated) noise polution to study the effect on marine critters.  No problem with the limited spectrum directionality in that situation.  Pretty cool though, they used a number of these, each attached to H2s powered by external batteries and stuffed in small water tight tubes, set the H2 set to intermittanly record once over a certain signal level and left them out there for a week or so.

You might enjoy taking a look at the plate array, an interesting use of omnis on a small surface.  I'd be interested in your thoughts on that one.

[Gutbucket]

Thanks, it was your mics which incorporate discs are what make me make the connection.  I'm very familiar with the 6dB boundary effect and the boundary size / corner frequency relationship.  It's the capsule arrangement on the plate that piqued my intenterest. Your comment on averaging out off-axis phase cancellations makes sense and is probably all there is to it in this case.  The spiral arrangement of the capsules reminded me of similar arrangements in larger multi-element mic arrays, some very large (24 up to several hundred elements) used for visual imaging of auditory sources in industrial applications like automotive NVH analysis, aerospace and such.  Typically plots of color values that denote local SPL or frequency overlayed on static images or video. Complex, non music audio apps.  I'm sure that kind of stuff requires individual ADC channels for each element and heavy computation resources to do what it does.. stuff that certainly does not apply here.  If that mic is doing anything more than simple capsule summing for SNR improvement, it would have to be pretty simple, analog, and fit in the amplifier body.

[Gutbucket]

Here are some photos of the very complex spiral beamforming arrays I mentioned which the simple spiral arangement of capsules on the plate mic reminded me of.
















..and a really big boundary mounted one for imaging aircraft wing wake vorticies-

[lamacchiacosta]

Hello,

after a bit of testing, doing and undoing, I finally found my way to use the Brahma and a workflow with it. I am quite happy with it and is always with me when travelling.
Please read my experience in my paper about this microphone here:

http://www.ecsound.net/pages/projects/tech_notes/engnotes.html

Any feed back is welcome. Please keep in mind that english is not my first language, so I am sure that some of the reading will be hilarious for that. Again, feedbacks are welcome.
Thanks.

[Gutbucket]

Thanks for linking your report here.  You english is not so hilarious at all, but quite good!

The Coresound Tetramic I use is similar, in that the biggest issue in using it is the support equipment (cabling, power supplies, recorder), in particular, accurate gain tracking across all channels of the multichannel recorder.  The Tascam DR-680 with 'gangable' gain trim controls allowing gain adjustment of all four channels in synchrony, using one knob and maintaining accurate interchannel, has helped greatly with that. 

I use the stand-alone VVMic software which was provided with the microphone for conversion from A-format, and it is very simple to use. It supports rates up to 24kHz/96kHz and can either output B-format or virtual microphone channels.  I typically dial in an appropriate virtual stereo pair while listening to the VVMic direct output, save that output to stereo WAV file and import that into other stereo editing software for any further editing.  There is also a VST version for use within editors, but the stand-alone program is convenient and has been all I've needed.

 VVAudio http://www.vvaudio.com/ also supports the Brama, so you might look into their ambisonic tools for easiser A-fomat conversion.

[Gutbucket]

I find ambisonic recording useful primarily for "improved coincident stereo" rather than optimal surround recording.  The ability to point virtual microphones in any direction with any 1st order directional pattern is unbelievably useful, yet one is always constrained to coincident microphone techniques, unless introducing additional microphones.

The primary value is the very compact and portable recording system and that virtual microphone adjustability after the recording has been made.  I don't find ambisonics as useful for optimal surround recording of music however.  Coincident first order directional microphone patterns are simply not tight enough to be useful for more than 3, maybe 4 channels total.  Superior surround recording of music requires space between the microphones to make up for that in my opinion.

lamacchiacosta, I'm curious about your bluround multichannel microphone technique I found mentioned on your website.  Can you provide some detail on what that entails and how you set it up?  Here's a link to another discussion thread on this site specifically about surround recording of music which you may find interesting, and which may be a better place for that discussion- http://taperssection.com/index.php?topic=170019.0

[lamacchiacosta]

Hi Gutbucket,

thanks for reading it and for the feedback.
Yes I found out about the new version of VVMic just when I was about to finish writing my paper. I will probably integrate that in my workflow later on. I am waiting to have the funds to buy that, then I will add that to the article.

I developed bluround about a decade ago when I was quite frustrated about the not so precise "source positioning" playback of all the multichannel techniques I tried. Also I found that pretty much all the multichannel techniques are missing definition when a source is coming from the center back and needed to find a way to have hat point locked and very clear.
I needed a three points technique to get positioning clear and I got inspired by the Decca tree, ORTF and MS technique.
It can use three or more microphones to record "planar" multichannel and 360º if needed.
I never written a proper paper about it for many reasons (moved to a different country, long testing, waiting for more people interested in use it for their recording).
Let me try to explain how to set it up:
You build a ORTF with two cardioid microphones facing forward. This will be your front end. Add a MS facing backward keeping the same distance (17cm) from the front end. This will then be your rear end. This equilateral triangle is fundamental to keep clarity of the source positioning. You can change the distance for the microphones if you like but is important to keep always the same triangle. You can also move the front end and rear end, this will work with a bit of compromise of the clarity.
These 4 tracks are the base of your surround environment and should work straight away. In your DAW assign the front end to the front speakers (in this case Left and Right) and the MS to the rear end (Rear Left and Rear Right) using a MS decoder.
If you record yourself walking around this configuration, when play that back you should be able to spot your movements exactly as they were recorded.
If you then need more channels you need to draw an "ideal circle" containing the triangle vertexes. You then add the needed mics. For instance if you need a center front you will add the center mic half way from the LR ones and on the circumference of the "ideal circle".
You do the same to add two more mics for the 7.1 multichannel configuration. The same to add Z axis for three dimentional recordings, where you can add a mic on top an done below if needed.

[Len Moskowitz (Core Sound)]

I find ambisonic recording useful primarily for "improved coincident stereo" rather than optimal surround recording.  The ability to point virtual microphones in any direction with any 1st order directional pattern is unbelievably useful, yet one is always constrained to coincident microphone techniques, unless introducing additional microphones

One of the interesting qualities of a single high quality ambisonic mic is that it has the potential to be the best figure-8 mic and Blumlein array (a pair crossed figure-8 mics) in the world - you needn't use it for surround recording. To my knowledge, there is no better figure-8 mic or Blumlein array than a single TetraMic. Its patterns hold up better across the frequency range and it goes deeper in the bass.

And as Gutbucket points out, you easily escape the constraint of coincident mic techniques simply by using more than one ambisonic mic.

Two TetraMics feeding a Zoom F8 makes a very compact and relatively inexpensive surround recording system that includes spaced mic techniques.

[lamacchiacosta]

Two TetraMics feeding a Zoom F8 makes a very compact and relatively inexpensive surround recording system that includes spaced mic techniques.

Slightly off topic.
Have you found a way, on the F8, to control multiple recorded tracks with just one trip pot to achieve the best recording control over an Ambisonics microphone?

[Len Moskowitz (Core Sound)]

Have you found a way, on the F8, to control multiple recorded tracks with just one trip pot to achieve the best recording control over an Ambisonics microphone?

I asked a Zoom rep at the recent AES show in NYC if there was any way to gang multiple channels so that one knob would control the gains for all. He said no.

[John Willett]

Here are some photos of the very complex spiral beamforming arrays I mentioned which the simple spiral arangement of capsules on the plate mic reminded me of.

Seeing as we are talking about measurement microphone arrays - how about this one:-

[DSatz]

As a bonus, the array in John's photo offers a "Mesmeric" mode in which it spins--first slowly, then gradually accelerating--while its capsules light up in various colors. This can be used to gain the band's cooperation for any recording projects that you may have in mind.

[Gutbucket]

..with the unintended consequence of simultaneously entrancing all spun-drunk ravers within eye-shot to the "pretty spinning lights on the pole".