9th October 2001 - Microphone Data: Is what you read what you hear ?

A survey of many hundred microphones has been recently published in book form Microphone Data by Rycote Microphone Windshields. Much of the research work was done by Chris Woolf, whose informative lecture was about his experiences gained during the preparation of the data book.

Chris began by describing various types of microphone to illustrate the enormous variety available. He then discussed dynamic range, which is rather ill-defined. The noise floor either has to be measured in a very acoustically quiet environment, or can be estimated by substituting a resistor or capacitor for the transducer. The maximum signal level is determined by the level at which the distortion reaches a certain value, which can be somewhat arbitrarily chosen. Typical examples are THD of 0.25%, 0.5% or 1%. In addition, there is usually some headroom, which is the signal range above the maximum level, but before clipping occurs. Most microphones have a dynamic range of around 90dB. Some specialist types can achieve over 130dB.

Chris then introduced the frequency measurement range, suggesting 5Hz to 40kHz, instead of the more usual 20Hz to 20kHz. To cover off-axis, proximity, bass roll-off, HF boost, EQ and diffuse field effects, the amplitude axis of the frequency response plot needs to cover 80dB, but rarely does because of space considerations in specification sheets. Another variable is the response of the microphone versus the angle of the source to the microphone. This is usually measured by having the microphone under test on a turntable and doing a frequency sweep test at a number of different angles. The results can be plotted as a polar diagram or as a series of separate amplitude responses.

A new way of plotting frequency, amplitude and directional response on one graph was presented. This appears as a 3D surface, along with colours for clarity. Responses of omni, cardioid, figure of eight and rifle microphones were compared.

Next, the time domain response was discussed, with waterfall plots used to show how microphone decays over time vary with frequency.

An extended discussion followed the formal presentation, exploring many microphone subtlies in more detail. This included discussions about noise floor, spark and pistol cartridge test signals, diffuse field response measurements, using two diaphragm or one diaphragm cardioid microphones, the smoothing of response plots, and how to make an acoustically quiet box. A even livelier pub discussion followed the lecture.

Anyone wanting a copy of charts used in presentation, please contact Chris at chris@chriswoolf.co.uk. Information on the Microphone Data book can be found at www.microphone-data.com.

Steven Harris