^^^ Nice description.
Some additional points: 1) This is not for those who prefer non-coincident methods (spaced, or near coincident). This is a personal decision, but I prefer to get stereo information from both intensity and phase information, hence the desire for non-coincident mics. 2) The quality of the recording depends on the capsules. So, if you love the sound of Sheops or whatever, you may not be satisified with the capsules used in the Soundfield mic.
Thanks. You're absolutely right about these being coincident only, but of course if you have the channels it can be combined with other mics. I like to run spaced omnis with it, partly for my own comparison, partly because I also have favored the open, atmospheric 'sound' of spaced mics, and partly for the ability to play with combining the sources or routing them to different channels for surround at some point.
Yes, the ultimate quality of the recording depends on the capsules, but this is a slightly different and unusual case compared to most microphones. In this case the quality of the capsules determines an upper limit to the capabilities of the mic, especially in noise level, clipping level, and distortion, and also in the hard to quantify subjective measures, but at least for the Tetramic, the unique measurement and post correction process that is specific to each individual capsule in each specific microphone, attempts to optimize the performance and minimize variations between capsules within that ultimate envelope of capsule potential. In doing so the system somewhat compensates for production variations and for general response irregularities of the specific capsules used.
The system corrects for response irregularities which vary from the ideal response in a way somewhat analogous to playback 'room correction' devices that measure the response of a speaker, then create inverse filters that correct for phase and eq response irregularities in the sweet spot. It can only correct so much though, within the ultimate limits of the capsule and the S/N ratio of the system, and can't correct for some things at all such as the dynamic range limits. Actually the Tetramic corrective system has a somewhat easier job to do than playback 'room correction' devices, because it is simply correcting and matching the anechoic on-axis response of each capsule. Speaker/room correction devices correct for the sweet spot while sacrificing the response elsewhere in the room and people like to move around.
I don't have any first hand experience with speaker/room correction devices, but I did get to demo a room/stereo/surround virtualization technique for headphone monitoring a few years ago that actually makes for a far better analogy here, not only because you can't move out of the haedphone sweet spot.. I’ve posted about it here elsewhere, but briefly, it was also was a measurement based correction system that required calibration to be useful. Calibration was done by placing miniature omni microphones in each ear canal and running sweep tones though all the various speakers in turn for a few different head rotation positions. The response of the headphones coupled with my ears was also measured. The dataset for that combined speaker/room/head-rotation/individual-ear-response and headphone/ear-response data was collected, manipulated and then convolved with whatever audio program I cared to hear. The result was incredible- my brain was completely convinced that I was hearing actual speakers in the original room, even though I was listening through headphones. Of course the ability of the system to make those corrections was ultimately constrained by reproduction limits of the headphones. The system I tested used entry level Stax headphones and amp which the developer considered more than sufficient in response (combined with a seat shaking device). The point I’m belaboring is that music through this sounded nothing like the Stax phones themselves. It sounded instead like the speakers and room where the measurements were made… as long as the phones were ‘good enough’.
The question then becomes, “what headphone (or what microphone capsule) is
good enough to support the correction requirements?”, rather than, “what is the particular sonic signature of this headphone or microphone capsule and do I like it?”
The Tetramic is designed to a price point and the specifications of the capsule used constrains the system, but unlike standard mics, the system attempts to remove the unique sonic signature of the capsules themselves.
Questions: What are the "first order" patterns? For example, can I get a highly directional pattern oriented in any direction or is it limited to hyper-sub- cardioid through figure-8 shapes? How much actual distortion of the patterns do I get due to non-coincident placement of the capsules?
All 1st order patterns can be described as the sum of an omni and figure-eight response mixed with various levels: that means any pattern along the continuum from omnidirectinal, subcardioid, cardioid, hypercardioid and figure-8. The theory relies on all the capsules being perfectly coincident. The virtual patterns tend to be far more accurate than most all standard microphones at low and mid frequencies, but start to breakdown at high frequencies and get very odd shaped as the wavelengths of the frequency in question approaches the capsule spacing. There are mathematical techniques that attempt to correct for the effects of the capsule spacing, but closer is better. Luckily the capsules are typically close enough to push irregularities above about 10kHz or so (I think) into the region where the brain is less critical about imaging.
In my opinion these mics are probably best suited to ambient recording or film work. Places where you may want to adjust the pattern in post-processing. Even then, a good sound guy with a shotgun mic may outperform them anyway.
Yes, film/TV dialog guys will continue to use shotguns on booms. ‘Shotgun’ interference mics are a different animal that have greater directionality at mid and high frequencies than any 1st order pattern. I’d guess many ambient guys want the open, ambient sound of spaced techniques for many things, but I know film sound guys collect location sounds and effects with Soundfields.
However, the post-processing pattern adjustment is what makes these things particularly attractive for the type of recording we do where we must set-up quickly, often in less than ideal situations, with limited ability to monitor and re-adjust things before recording.
If a coincident technique is appropriate, an ambisonic mic leaves the most options open.