> The data-format also seems to remain as what it was in the source file... Thanks for the bug report; the problem is that the channel extraction code forgot to include the new header info. I think the current tarball works correctly.