York University Music Technology Group
Using ordinary microphones to emulate a Soundfield.
Want to use existing microphones to avoid buying a pukka Soundfield? This is possible using three figure-8's with an omni or two figure-8's and an omni for B format if you can do without Z-axis information. However, you should be aware that even mounting the microphones absolutely as close as possible will not give as good a result as a real Soundfield mic, but it's good enough to experiment with and Nimbus does pretty well all their recordings this way.
You will need two (or three) figure-8 mics of the same model and a good omni that sounds as near the same as the figure-8's as possible. Try to get physically small units. Mount the as clos as possible to each other, with the figure-8's mounted so that the capsules are sitting at right-angles to each other. If you are making a horizontal only version, put the omni in the centre with one figure-8 above and pointing forwards, the other below pointing side-to-side. Avoid obstructing the acoustic pathways to the capsules with cable, mic stands etc. as far as is physically possible and what obstructions there are, make as symmetrical as possible. Now you will need to align them.
Aligning the microphones
How do you align the rig? Well, preferably with an anechoic chamber, and a plane wave tube, and lots of expensive test gear - but there are ways to do without that. First you need to get the polarities of the capsules the same. To do this, if you haven't got something like the Phase Check System from Canford Audio, you will need something to make a click - like, by snapping your figures. Stand so that you are positioned at 45 degrees left (ie. half way between the two horizontal figure-8's) and snap your figures whilst recording the outputs of the three (or four) capsules into your PC/Mac/digital workstation. Now bring up the clicks on the screen and compare the waveforms. They should all look the same, at least in terms of the direction (up or down) of the start transient. If not, change the polarity of whichever is needed to make them all the same. Ideally, try and make them give a positive output for a positive pressure change. Now go to somewhere with a very reverberant acoustic and work with the microphone array well away from the walls or floor. Place a speaker well away from the microphone and play pink noise through it at a good level, well above the noise floor of the room - face the speaker away from the mic, preferably pointing towards a diffusing surface (curved wall, architectural decoration with a lot of bumps, rpg diffusor). Note the output level from the figure of eight pointing forwards (towards the speaker). Rotate the mic by 90 degrees so the other mic is facing forwards the speaker and adjust for the same output. Adjust the omni for the same output as well - if the space is truly reverberant, this will result in the W output being the required 3dB down. To check this, repeat the proceedure in a dry acoustic - this can be outdoors well away from walls, etc - and face the speaker towards the mic, possibly at a smaller distance. Now you should see the signals from the figure eights at 3dB above the W when they are pointing directly at the speaker. That should just about ensure that things will work correctly. Good luck!Music Technology GroupWelcome page
Last updated; 15th. October 2001 by Dave Malham.
If you have any suggestions, comments or requests you can reach
me at
dgm2@york.ac.uk
To return to the YorWeb home page click here