Smoothing!

</p> <p> I finally got smoothing worked out for FuzzMeasure 1.3. I implemented smoothing using FFT filtering of the frequency data, and the results look great! In the picture above, the blue curve represents a 64k FFT analysis of my iBook's impulse response (including room effects), and the black overlaid curve represents the same analysis after smoothing. </p> <p> This operation actually smooths the information in the frequency domain, and then derives an impulse response to match the smoothed information. As a result, the operation destroys your original impulse. To account for this, users initiate smoothing by actually creating a smoothed copy of the original, rather than destroying the original graph. </p>