You should use this version instead; it's better.
Note that this uses a bilinear transform and so is not accurate at high frequencies.
Apply an A-weighting filter to a sound stored as a NumPy array.
Use Audiolab or other module to import .wav or .flac files, for example. http://www.ar.media.kyoto-u.ac.jp/members/david/softwares/audiolab/
Translated from MATLAB script (BSD license) at: http://www.mathworks.com/matlabcentral/fileexchange/69
@ThomQu
at lower sample rates.(Edit: Actually I guess it's not yet. I have a better version with MZTi correction that I haven't pushed yet. I thought I pushed it a long time ago.)