AIFF-C files with 32-bit or 64-bit floating point samples gives an error: "Unsupported bit-depth". The files were exported using QuickTime 5.0.2 on MacOS 9.2.1 (a very old system). See the attached zip file for examples.
The files have the bit depth field set to 16, but in reality they are 32-bit or 64-bit floating point files. QuickTime 5 probably had a bug because it was writing the wrong bit depth. Modern macOS 13 Ventura / QuickTime version 10.5 can open these files. Also, Audacity 3.1.3 can open these files. Do you want to support these kind of files?
Maybe if type is "fl32" or "fl64" then always use 32 or 64 bit depth and ignore the bit depth field in the file? Or maybe if the compression type is not "NONE", then always derive the bit depth from the compression type field and ignore the bit depth field?
AIFF-C files with 32-bit or 64-bit floating point samples gives an error: "Unsupported bit-depth". The files were exported using QuickTime 5.0.2 on MacOS 9.2.1 (a very old system). See the attached zip file for examples.
The files have the bit depth field set to 16, but in reality they are 32-bit or 64-bit floating point files. QuickTime 5 probably had a bug because it was writing the wrong bit depth. Modern macOS 13 Ventura / QuickTime version 10.5 can open these files. Also, Audacity 3.1.3 can open these files. Do you want to support these kind of files?
Maybe if type is "fl32" or "fl64" then always use 32 or 64 bit depth and ignore the bit depth field in the file? Or maybe if the compression type is not "NONE", then always derive the bit depth from the compression type field and ignore the bit depth field?
quicktime5.zip