Opened 13 months ago

Last modified 13 months ago

#1375 new defect

Decoder crash in decoding the second independent layer in a multilayer bitstream

Reported by: rickxu Owned by:
Priority: minor Milestone:
Component: VTM Version: VTM-10.0
Keywords: Cc: ksuehring, XiangLi, fbossen, jvet@…

Description

VTM10.0 decoder crashed when decoding a multilayer bitstream containing two independent layers. If set "-p 0" in the decoder commandline, the decoder could correctly decode the layer 0 CVLS. But if set "-p 1", the decoder crashes with the error message "DecLib.cpp:1749: The value of sps_video_parameter_set_id shall be the same in all SPSs that are referred to by CLVSs in a CVS.".

The attached zip file contains the bitstream and the config files used to encode the bitstream with below encoder commandline.

EncoderApp.exe -c encoder_lowdelay_vtm.cfg -c two_indep_layers.cfg -c BasketballPass.cfg -l0 -q 37 -l1 -q 27 -f 9 --SEIDecodedPictureHash=1 -b TwoIndependentlayers.bin

Attachments (1)

TwoIndependentLayers.zip (22.2 KB) - added by rickxu 13 months ago.

Download all attachments as: .zip

Change history (3)

Changed 13 months ago by rickxu

comment:1 Changed 13 months ago by rickxu

  • Version VTM-10.0rc1 deleted

comment:2 Changed 13 months ago by ksuehring

  • Milestone VTM-10.1 deleted
  • Version set to VTM-10.0

Unfortunately sub-bitstream extraction is not fully implemented in the VTM decoder. The work-around would be to run the bitstream extractor first and then decode the extracted bitstream.

Possible solutions:
a) remove "-p" option that only works in some cases and always require using the extractor
b) integrate extraction into the decoder

a) should be much easier from the implementation side although it requires an extra step for the user.

Note: See TracTickets for help on using tickets.