Large deviations observed with the extrapolation of shifted channel by TD-ITD compensation in stereo
Basic info
Bug description
This issue is already observed in issue #1463 (closed) .
It seems that something is erroneous with the extrapolation of the missing part of the signal when shifting the lagging channel by the ITD number of samples. As en example frame=842 for the same item and MDCT-stereo operation point shared in #1463 (closed):