While on the topic of images, I'd like to just toss in some
image processing trivia...
If you take the FFT of a (single) image (I know the thread was
on the subject of stereo images) and throw away all the amplitude
information in the FFT, you can still reconstruct the entire
image at full resolution from just the phase information. There
will be a constant offset (ie. brightness, for example), but
otherwise you have the complete image. I think I have this
I'm wondering - and I have no real basis to believe that this
bit of image processing trivia has any relationship to audio - if
it would be possible to reconstruct the original stereo signal
from just the phase information. And if this were possible,
we might even build DSP tools to assess the quality/accuracy
of a measured stereo audio "image". If there was any debate
if a system were creating a proper stereo audio image, such
tools could reverse a stereo signal and then compare the
phase-generated stereo signal to the source...
Hah. Easier to use your ears for sure!
OK - enough blue sky out-of-the-box craziness for a Friday
The Audio Archive
mailto:[log in to unmask]