Strong artifact with Stereo JBM decoding to MONO output causes high MLD
Basic info
Bug description
Looks like some random/outdated signal is added to the output dmx signal incorrectly at the wrong level. Could be wrong buffer used/buffer not correctly overwritten/out-of-bounds write... The difference is very big in amplitude and causes an MLD of 97.
This does not happen when decoding to stereo. So the error is either hapening in the downmix or later in the JBM path a wrong number of channels is used.
Ways to reproduce
./IVAS_dec_fx -voip -Tracefile out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.jbm-dly_error_profile_4.dec.mono.fx.dectrace mono 48 out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.netsimout-dly_error_profile_4.192 out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.jbm-dly_error_profile_4.dec.mono.fx.wav
./IVAS_dec_flt_ref -voip -Tracefile out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.jbm-dly_error_profile_4.dec.mono.flt.dectrace mono 48 out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.netsimout-dly_error_profile_4.192 out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.jbm-dly_error_profile_4.dec.mono.flt.wav
ssnr.py --print-ctest-measurement --thresh_low -50 out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.jbm-dly_error_profile_4.dec.mono.flt.wav out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.jbm-dly_error_profile_4.dec.mono.fx.wav
wav-diff --no-fail --print-ctest-measurement out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.jbm-dly_error_profile_4.dec.mono.flt.wav out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.jbm-dly_error_profile_4.dec.mono.fx.wav --ref-jbm-trace out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.jbm-dly_error_profile_4.dec.mono.flt.dectrace --cut-jbm-trace out/stereo-noisy_speech-p800-2-cat4-lab_d--stereo-noisy_speech-p800-2-cat4-lab_d-20ms_aligned_files_concatenated--32000--dtx.jbm-dly_error_profile_4.dec.mono.fx.dectraceBitstream shared via box as usual.
(Clear steps or refer to a failing automated test, e.g. with a pipeline link)