I need to edit the soundtrack of a captured digital video recording (raw dv format, i.e. audio is pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s).
For some reason I don’t understand some audio samples have spurious zero amplitude. Imagine a half-period [0:pi] of a sine wave which is represented by about 50 audio samples. In my recording, samples 3, 4, 8, 9, 13, 14, etc are zero instead of having the correct non-zero value.
I’m puzzled as to what could have created these, but regardless I would like to remove these zeros and interpolate the true values from the neighbouring samples.
Is this possible from within audacity, or otherwise how could one solve this?