heathen, here's a quickie translation, but please "see below" re: German vs. U.S. ways of thinking about these issues.
------------
4.3.3 DIN method
The DIN method is unknown in actual practice, and represents a former proposal from German [public] broadcasting for an "equivalence-stereo" arrangement [see below --ds]. Despite this, I have given consideration to this type of setup because it forms a mid-point between ORTF and NOS, and thus represents the countless possibilities for setups having roughly equal components of level and arrival-time differences [again, see below --ds]. The benchmark data for this method are: 20 cm spacing between the microphones; angle between the two main axes, 90°. This leads to a stereophonic recording angle of 101°, with a 52.4%-to-47.6% relationship of level to arrival-time differences [yep, see below --ds].
------------
Here's the "below" part: In U.S. textbooks, it is generally explained that stereophonic localization can be based either on the precedence effect (when two microphones are spaced apart, direct sound from any given source will arrive at the closer microphone first, unless the source is exactly on the center line between them) or the effect of the microphones' directionality, assuming that they're aimed away from one another. Most people here are familiar with that dichotomy, I think. You can go back to the 1970s or even late 60s and find books that showed and compared the two methods. Both had a lot of tradition behind them, and important practical pros and cons could be discussed (e.g. regarding mono compatibility, which was a big thing back in the LP-and-mostly-mono-radio era)(back then most pop music stations were still on the AM band, and many people still listened to FM radio in mono either in their cars or on portable radios)(and airplay was the primary driver of record sales).
In U.S. books and articles starting maybe in the late 1970s (but definitely by the 80s) the "or" became an "or/and"--where the "and" meant directional microphones that were spaced apart and angled apart, but neither spaced nor angled apart as much as in the traditional A/B or X/Y approaches. The authors seemed to feel as if they were being forced to concede the existence of such "neither fish nor fowl" methods, though. They were based on trial and error, and no handy theoretical explanations could be offered for the setups that worked well vs. the ones that didn't. The only advice was to try, listen, not go to extremes, and pray for luck. The ORTF method was sometimes mentioned, a little grudgingly as it seemed to me.
Where they took it somewhat farther in Germany was to reason out that mathematical formulas, based both on the trigonometry of microphone patterns and on experiments that quantified the precedence effect (i.e. how far--to what actual angle--do you hear a source being shifted to the left or right when it arrives at your ears X milliseconds sooner rather than later?), could allow any given angle between two microphones of a given pattern to be translated into a roughly equivalent distance between the same two microphones, or vice versa. Using an equivalence formula, intermediate spacings and angles could be interpolated between any two A/B vs. X/Y endpoints. You can even select your favorite point along the curve of those interpolated settings, depending on the relative effect of "arrival time differences" versus "level differences" that you want to apportion.
That changed the whole game. Now you could choose your desired stereophonic recording angle (based on your distance from the sound sources and the angular width that they represent from your microphones' point of view), then choose how much you wanted to rely on arrival-time differences vs. level differences (each of which gives a recording a different spatial "feel"), and the formulas could tell you the spacing and angle that would give you what you wanted. Suddenly it was no longer guesswork. But that approach didn't find its way to the States until ... frankly, I don't know whether most people here understand it even today. Instead they rely on Williams' or Sengpiel's charts or on Wittek's "stereo assistant" on line. All of which is fine fine fine.
But the result is a big gap between American and German writing about this topic that still persists. In the German writing there are three basic categories of two-mike stereo recording methods, all of which are brightly-lit and clearly defined: (a) methods that rely on arrival-time differences only (i.e. spaced omnis), (b) methods that rely solely on the directionality of the microphones, which creates differences in level between channels for each distinct sound source, depending on the angle from which its direct sound reaches the microphones (i.e. X/Y), and (c) "equivalence" methods which rely on calculated, or at least calculable, tradeoffs between principles (a) and (b). All three categories are understood to be fundamentally knowable and predictable, although certain non-ideal behaviors of microphones (e.g. large dual-diaphragm cardioids) still need to be taken into account.
Meanwhile in most American writing there are still mostly just the two opposite poles (spaced vs. coincident), with the in-between area given less respect, even though it's where so much of the fun is--like as if "some people like to swim in those waters, but there are no firm guidelines, and you could get your ass bitten if you aren't careful". That attitude implicitly encourages people to rely on cookbook formulas as safe havens, while others express their indomitable American spirit by doing the exact opposite, declaring that they don't need no steenkin' formulas.
--best regards
Topic: so-called "DIN" stereo recording method(s) (Read 4983 times)