Originally Posted by cjwhitehouse
If you have the right software and use a logarithmic sine sweep, you can make a continuous plot of THD% against frequency. WinMLS can do this among other software packages. It can also show the contributions of individual harmonics at each frequency.
THD% as a pure number unfortunately tells you nothing about how distorted something sounds. You have to take into account the ear's equal loudness contours, transfer function of the ear and psycho-acoustic masking effects to get anywhere near a metric that will tell you how bad a given level of distortion sounds.
Measuring in-room will also completely mess up any objective measurement of distortion level as the room will boost/cut fundamental and individual harmonic levels in a way that renders the results pretty meaningless for anyone but the originator.
Thanks, CJW, and nice to see your input in this forum, as always.
You can use TrueRTA to calculate in-room harmonic distortion manually, but as CJ points out, there are sympathetic vibrations and room influences involved.
You can run a sweep to get the in-room FR at the mic position, save it on the graph and refer to it in your calculations of the individual harmonics, adding/subtracting for the rooms boost/cut influences. This, IMO, is a decent enough reference for those who are never going to consider dragging everything out to a parking lot, although much more tedious.