Delaying Spot Microphone Signals for Runtime Compensation

In the era of digital audio recording, it is no longer a problem to digitally compensate the runtime difference of acoustical signals between spot microphones and the stereo microphone pair (that naturally is more remote from the source). But there is no agreement if it is an advantage to do so, nor is it clearly defined how long the delay should be: does an exactly calculated delay, i.e. perfect equalization of stereo and spot microphone signal runtimes, improve the sound? Or would it be better to overcompensate the spot microphone signal to make it reach the ears even after the stereo pair, like an early reflection?
While working on this project, we systematically analyzed the influence of delayed spot microphones on the sound of a recording. There were 30 different setups to be examined in the course of the project, that was performed with the guidance of Prof. Rainer Maillard.

The basic setup consisted of a spaced microphone pair with a base width of 45 cm and a mono spot microphone. The three microphones were of the same manufacturer and therefore had identical amplifier circuits, while the capsule types and thus the directional patterns varied. Now we recorded three audio examples with the same source, but with variable distance between spot microphone and stereo pair (2 meters, 5 meters and 8 meters). Then three different mixes were made of every recording: one without any delay, another one with exactly calculated delay (determined by sound velocity) and one with exact delay plus an additional 10 milliseconds. The level proportions between the microphones were regulated individually for each spot microphone distance.

Several examples of our recordings are available here for download. Each example contains a short passage of audio in all of the three mix variants: first without delay, then with exact delay and finally with 10 ms of overcompensation. You will easily detect significant differences in sound, even though the files are in quality reduced MP3 format. Judge by yourself...

• Speech Distance spot mic - stereo pair: 5 meters
  (recorded at Great Hall, Musikverein, Vienna)

	without delay
	with exact delay		134 kB each / MP3 file @ 128 kBit/s, Stereo, 0:09 min
	10 ms overcompensation





• Piano Distance spot mic - stereo pair: 2 meters
  (recorded at New Auditorium, Academy of Music, Detmold)

	without delay
	with exact delay		133 kB each / MP3 file @ 128 kBit/s, Stereo, 0:09 min
	10 ms overcompensation





• Voice Distance spot mic - stereo pair: 5 meters
  (recorded at New Auditorium, Academy of Music, Detmold)

	without delay
	with exact delay		115 kB each / MP3 file @ 128 kBit/s, Stereo, 0:07 min
	10 ms overcompensation





• Kettle Drum Distance spot mic - stereo pair: 8 meters
  (recorded at Music Hall, Hamburg)

	without delay
	with exact delay		71 kB each / MP3 file @ 128 kBit/s, Stereo, 0:05 min
	10 ms overcompensation





• Organ Distance spot mic - stereo pair: 8 meters
  (recorded at Martin Luther Church, Detmold)

	without delay
	with exact delay		133 kB each / MP3 file @ 128 kBit/s, Stereo, 0:09 min
	10 ms overcompensation





• Saxophone Distance spot mic - stereo pair: 2 meters
  (recorded at Martin Luther Church, Detmold)

	without delay
	with exact delay		182 kB each / MP3 file @ 128 kBit/s, Stereo, 0:12 min
	10 ms overcompensation





• Violin Distance spot mic - stereo pair: 2 meters
  (recorded at the foyer of New Auditorium, Academy of Music, Detmold)

	without delay
	with exact delay		175 kB each / MP3 file @ 128 kBit/s, Stereo, 0:11 min
	10 ms overcompensation

To compare two audio examples in a representative way, we recommend the freeware tool ABX Comparator, which allows so-called "blind tests" and was already used in the auditive comparison of DSD and PCM coded material, performed at our institute. Unfortunately this software is only able to playback in wave format, so you will have to convert the MP3 files. A comfortable and fast tool for converting is MP3 and WAV Solutions (free demo version).

(ds)