Rationale: compensate a speaker response curve (gain and phase) on a audio file.

Just came accross REW (Room EQ Wizard) (Linux, Windows, Mac), it let you plot the output of your speakers, then create a filter to compensate the flaws.

Once you created your finely tuned speakers, Export-> Filter as text file.

Then, it’s time to process the audio file with this filter, using the amazing Sox:

sox -q input_file.wav output_file.wav `awk ‘$4==”PK” { gsub(/,/,””,$6 );val=$6;printf ” equalizer “val” “$12″q “$9}’ val=”%’6.3f” REW_filter.txt`

That’s it: you have a outputfile with the correct compensation for your just analysed speaker.

Of course this method is for non-real time need.