Hey gang,
I'm using Vista and the IMediaObject, IMediaBuffer, blah blah blah for
AEC. However, if the near-end adjusts either their mic gain or
speaker level the far-end hears their own echoed speech. I think it
will eventually train but it takes a magnitude of minutes to train.
Sooooooo. I get the AecQualityMetrics_Struct stats and look at the
Convergence flag. It never changes from 1 to 0. And even if I were
to detect when AEC is diverged, I'm not sure what to do. Does anyone
here know:
1. How I can detect divergence in Vista. And...
2. What to do about it when I do detect divergence.
/Loren