Packed YUV output in VTM decoder doesn't work anymore

[crhelmrich]

Specifying --PYUV=1 in the VTM decoder command-line used to allow packed YUV output during decoding, see also VVCSoftware_VTM/doc/pyuv_format.pdf

In commit ​https://vcgit.hhi.fraunhofer.de/jvet/VVCSoftware_VTM/-/commit/6c7dae5424ec18f7dfca2b7b8e4ff6adf07e9d82#c6b7bcb864f352a5ff7be94d8c36d7f1fefae3ab
(Jan 22, 2020, author yiwenchen) a

CHECK( writePYUV, "Not supported" );

was introduced in source/Lib/Utilities/VideoIOYuv.cpp which completely disables support for PYUV output, even though the respective code is still present and, when removing the above CHECK, decoding to PYUV format seems to work fine (at least on the 10-bit SDR CTC bitstream I tested it with).

I'll submit a merge request soon which simply deletes the above source code line with the CHECK, but maybe the author of that code can shed some light on the reasoning behind this check, i.e., under which circumstances the PYUV output doesn't work properly.

Thanks,

Christian

[crhelmrich]

Sorry, it seems the original author of this CHECK wasn't yiwenchen, as noted above, but V. Seregin when integrating JVET_O1164_RPR on Aug 13, 2019:
​https://vcgit.hhi.fraunhofer.de/jvet/VVCSoftware_VTM/-/commit/f38104fd0db7d9c9c3352ccc867d23f14992f75f#c6b7bcb864f352a5ff7be94d8c36d7f1fefae3ab