sunbeamgls
pfm Member
Using this (flat) speaker and measurement technique, a theoretically perfect (according to Toole) in-room response would show a smooth 1dB/octave (10dB) fall off between 20Hz and 20kHz. The modal resonances in this untreated room create some large deviations from this ideal at the bottom end, most notably the really big peak at 36Hz. By applying some filters in the 20Hz to 200Hz region, you can compensate for these modal peaks to get a pretty close to perfect in-room response.
This chart shows the measured (uncorrected) response for one speaker, the 1dB/octave target, the filters required to map the measured response to the target and the predicted response with the filters applied. The "after" measurements confirm that the measured response with the filters in place is pretty close to the predicted response (and the target).
Excellent description of what's going on. Thanks.