HomeProjectsStudents' Projects of the Term → Stereo Recording Techniques

Microphone Techniques for Stereo Recording

Music and other audio sources will generally be recorded using a single pair of microphones that is located at a suitable place. This stereo pair or main microphone will become the basis of the final stereo mix. There may be cases where the stereo pair is the only microphone anyway, which would represent the recording ideal.

The physiological and psychological properties of human hearing makes our brain perceive a spatial sound image by analyzing intensity and runtime differences between both ears. According to these facts, there are three basic principles of setting up a stereo microphone pair:

  1. The coincident pair (intensity stereophony) is created with two identical microphones that, because of their directivity, produce a level gradient between the two stereo channels. The cardioid capsules have to be placed one as close as possible above the other, but with different angles towards the source. In this setup, the soundwaves will reach both capsules simultaneously (coincidentally), but with different intensity. The coincident pair (also named XY) is best chosen for optimal sound localization in the room.
  2. spaced pair (runtime stereophony) uses two identical microphones in a certain distance (base width), located symmetrically in front of the source and adjusted towards the same direction. Soundwaves coming from either side will reach one capsule sooner than the other, so that there will occur runtime differences between both microphone signals, and the ears will perceive a directional and spatial image. This technique is also known as AB stereophony, which is best chosen for a most natural image of the real spatiality.
  3. near-coincident pair (equivalence stereophony) is a combination of both level and runtime differences to create a pleasing room image in addition to good localization of sources. This is possible because of the phenomenon of equivalent directional impression, dependent on level and runtime gradients as parameters. The setup for this recording principle is built with two cardioids in a certain distance, and with the capsules adjusted in a certain opening angle (such as the ORTF technique). For an illustration of the manifold variations offered by this technique, have a look at the Williams curves (developed by Michael Williams, see diagrams below).

Directional impression and localization of sources in the stereo image is almost fully determined by level and runtime differences. The recording angle of the microphone setup in relation to the measurements of the source (NOT identical with the opening angle of the stereo pair!) can be defined as an indicator for the phantom-source image that will be produced when monitoring with two speakers. With a spaced pair for instance, the curve to the left visualizes the influence of the base width on the resulting recording angle. If MS stereophony is used, it is possible to transfer the directional patterns of the capsules via a matrix into a corresponding XY stereo signal, as there are nothing but level differences relevant for this technique.

For more detailed information, please refer to the web page www.hauptmikrofon.de that also offers the application "Image Assistant", based on the Williams curves. For Macintosh users, there is another useful tool available called SR-Angle. These calculations are nice, but anyway you mustn't forget that the directivity of microphones in real life is remarkably dependent on the recorded frequency range.

The following table lists the most important stereo recording principles as a comparison:

XY coincidental stereophony two cardioid capsules @ 135° opening angle
Blumlein coincidental stereophony two figure-8 capsules @ 90° opening angle
MS coincidental stereophony Mid capsule = omni, Side (rear) capsule = figure-8
AB runtime stereophony with common base widths of 50, 70 and 100 cm
ORTF equivalence stereophony base width of 17 cm @ 110° opening angle; recording angle = 96°
level gradient = 8,2 dB (60%), runtime gradient = 0,368 ms (40%)
NOS equivalence stereophony base width of 30 cm @ 90° opening angle; recording angle = 81°
level gradient = 5,42 dB (42%), runtime gradient = 0,573 ms (58%)


[Supplementary audio examples...]

© Erich Thienhaus Institute at the Academy of Music, Detmold | Imprint