It's now 2025, 40 years later. Even as amateur non-professional location recordists, we have the ability to quite easily and inexpensively do things that were challenging for professionals when that article was written. Its no big deal for us to record, easily and inexpensively, up to 8 fully sync'd channels all with full bandwidth and high dynamic range if we like. Recording four channels, even six, is common. Still, nothing in that article is incorrect. I concluded long ago there are essentially three basic categories of microphone technique that apply to live concert recording: 1) pseudo binaural (head-worn, mostly stealth); 2) Optimized 2-channel configurations (Improved PAS ranging from coincident through near-spaced arrays); 3) Multichannel recording arrays specifically designed to combine the best attributes of each of the recording approaches Lipshitz describes, while minimalizing their drawbacks. Regardless of approach, we all seek to achieve a reasonable combination of accurate direct-sound imaging, good spatial impression, natural audience reaction and good room impression, in repeatable and reasonably achievable ways.
We are fortunate to stand on the shoulders of giants.
I had been waiting to add a bit more of my perspective. With the sentence I bolded it begins. We are incredibly more flexible than our forebearers in that we have almost infinite channels at our disposal, cost and gear availability considered. The improvements in microphonics, especially the ambisonic types have given us even more flexibility.
However, the Giants referenced did most of the basic legwork for us. Blumlein is certainly a great person to start with if you wish to dig deep into his/their research. When we studied psychoacoustics, Blumlein is one of the first scientists referenced as he made the earliest attempts to incorporate the recording techniques with consideration of the playback/listening technology. In his time, alongside EMI, Amar Bose, Henry Klipsch, and The Bell labs guys were all studying, experimenting, and documenting the results. Lipshitz was a "second generation" of these earlier pioneers, and many engineers such as Bob Carver, and Bose specialized in playback designs of amplifiers and loudspeakers, of which Lipshitz was experimenting with and documenting those results. He was a math guy who liked to present his theories in less technical ways, as in the article referenced. He definitely was making a lot of field recordings at the time.
1] Coincident techniques such as what Lipshitz is talking about in this paper are compared with the typical multi-mic arrays used in studio or live recording of symphonic orchestras or big band jazz from the 1950's and early 1960's. Those ear;y multi-mic recordings could have had three or four up to 16 rmicrophones all arranged in a line at stage lip or in the orchestra pit with "spot" mics typically placed in the orchestra, near the smaller woodwinds, the brass, and strings. We made recordings like this for the University, typically only using eight microphones, four in a line and four spots. I *think* the main point to it all was to use up the 8 or 16 (later 24) channels of tape the modern large studio machines used. The end results depended A LOT on the expertise of the engineers. Shifting to two or four channel field placed and location optimized microphones is the point to what this paper discusses and what was being done in the field at the time (Late 1970's, early 1980's).
2] Early "Stereo microphones" were constrained by the size of capsules, magnets, and circuitry. As time progressed, and smaller elements were introduced, the electret condenser and such, the types of coincident microphone matching and maufacturing became easier to do. Note that at the end of this paper he gushes about early ambisonics using a BORROWED Soundfield 1 to make his experiments. I recall doing loudspeaker measurements at the anechoic chamber at University int he early 1980's. We rarely had the latest and greatest even at a private University as the studio and radio station budgets were only so much. Our mentor, Mark Fitzgerald, who ran the studios for three years, was always borrowing and squeezing out gear from the manufacturers. So Lipshitz was also constrained by budgeting, of course. So after studying the results, to their ears, of these Soundfield recordings, he was very high on using those mics to obtain excellent coincident samples (which can be altered later with software) He was certainly chasing the best sound capture for that era. I have recorded with several stereo mics, notably the AKG c422, c426 and c34 family. IN the early 1980's we also had a Sony stereo mic (ecm 902?)
3] I found his statement that to preserve accurate imaging only three speakers, not four, should be used for optimal playback to be almost funny in 2025. One heavy area of research in the mid 1980's was multi-channel sound, surround sound and the beginnings of the immersive arrays such as the Cirque du Soleil's 16 channel system. I think the industry quickly moved to four, five and seven point surround sound within the next few years, certainly by the early 1990's. (My area of research and the US Patent we were awarded was in this multi-channel playback field) My experience does lead me to notice the three vs four speakers he mentioned and sort of laugh and note how things have truly changed since 1986. At that time, we AES scientists and members, were all being told by the bean counters no one would afford/purchase multiple amplifiers and 4 or more loudspeakers for their home. [Edit to add: I misinterpreted the paper in regard to the ambisonics section - but will let this stand for continuity of the thread] (read below posts)
It turned out that the bean counters were all BS in the hindsight of the way it turned out, which allows me the personal insight to realize that even "experts" cannot anticipate the inventions to come and the willingness of consumers to pay.
This is a good fundamentals type paper, and Lipshitz was well regarded by the AES community. Thanks for posting this heathen.
(BTW- I am aware that this post will not go far in convincing the tapers who don't frequent this site to stop calling it Audio Engineering section)
https://meyersound.com/news/cirque-du-soleil-alize/ Interesting new system for Cirque using Meyer technology (excerpt relating to this discussion):
System optimization and spatial processing are supplied by ten Galileo GALAXY Network Platforms, with six dedicated to the P.A. and four running in Spacemap Go mode to manage approximately 64 outputs. Everything is connected via a Milan AVB network.
Only three loudspeakers, eh?
Topic: Stereo Microphone Techniques...Are the Purists Wrong? (Read 2910 times)