View Single Post
Old 6th Aug 2020, 4:06 pm   #1697
mark_in_manc
Octode
 
Join Date: Nov 2011
Location: Manchester, UK.
Posts: 1,872
Default Re: The Audiophoolery Thread.

On the 'feedback' part of this thread - we (I was at Salford) used sub-miniature piezo accelerometers glued to drivers, to derive signals to feed adaptive filters running on boards installed in 286 computers - it was all a rather long time ago. The really difficult thing was establishing a target for the filter adaptation - we thought we wanted to reduce the spatial, freq. and time domain artefacts associated with the way the driver excited the room at low frequency, but working out a cost function was not at all trivial, and doing so whilst not measuring the room to death with large microphone arrays, even more so. Well, they don't call it Permanent Head Damage for nothing

B&O had a rather nifty product which tried to do something like this by measuring its own radiation impedance - the acoustic impedance (pressure / velocity) looking into the room at the location of the source - using not a mic and accelerometer, but a single mic which moved out on a motorised 'stick', to provide an impedance estimate based on a two-microphone finite difference velocity approximation. I suppose this would be around late 1990s, and if I remember correctly, it was on sale in shops.

The accelerometers we used were apparently developed to encourage anti-personnel mines not to waste their time blowing up for a passing tank, I was disturbed to find out much later.
mark_in_manc is offline   Reply With Quote