Unable to create QCTools concurrent to capture

[iamdamosuzuki]

I'm not sure if this option is supposed to be working, but it fails every time I try to create a QCTools report concurrent with capture.

I've never successfully done this, but currently have access to a powerful computer so I wanted to stress test it. Unfortunately, the error crashes immediately whenever I try to run a capture with concurrent QCTools report creation.

Hit enter to start recording

Close the playback window to stop recording.
qcli dev build, Oct 12 2023 07:09:34
filters selected: signalstats aphasemeter astats ssim 
Guessed Channel Layout for Input Stream #0.0 : 7.1
[Parsed_pan_0 @ 0x7fdf47a0d880] [SWR @ 0x7fdf38e38000] Input channel layout has a different number of channels than the number of used channels, ignoring layout
Input #0, matroska,webm, from 'pipe:':
  Metadata:
    ENCODER         : Lavf59.27.100
  Duration: N/A, start: 0.000000, bitrate: 9216 kb/s
  Stream #0:0: Audio: pcm_s24le, 48000 Hz, 8 channels, s32 (24 bit), 9216 kb/s
    Metadata:
      ENCODER         : Lavc59.37.100 pcm_s24le
  Stream #0:1: Video: v210 (v210 / 0x30313276), yuv422p10le(bottom first), 720x486, 29.97 fps, 29.97 tbr, 1k tbn
[Parsed_pan_0 @ 0x7ff3ec716900] Pure channel mapping detected: 0 1
[Parsed_pan_0 @ 0x7ff3ed107f40] Pure channel mapping detected: 0 1

analyzing input file... -
                                                   0 of 100 %[matroska,webm @ 0x7ff3ed00fd40] File ended prematurely
[matroska,webm @ 0x7ff3ed00fd40] Seek to desired resync point failed. Seeking to earliest point available instead.
/usr/local/bin/vrecord: line 2666:  6130 Broken pipe: 13         tee >("${PLAYER_COMMAND[@]}")
      6132 Segmentation fault: 11  | "${QCLI_COMMAND_PIPE[@]}"
av_interleaved_write_frame(): Broken pipe
[Parsed_pan_5 @ 0x7fdf47809600] [SWR @ 0x7fdf19298000] Input channel layout has a different number of channels than the number of used channels, ignoring layout
Finishing stream 0:0 without any data written to it.
Finishing stream 1:1 without any data written to it.
Error writing trailer of pipe:: Broken pipe
Error closing file pipe:: Broken pipe
mv: /Volumes/CaptureDisk01/20251105/Test02_vrecord_input.log: set owner/group (was: 501/0): Operation not permitted
mv: /Volumes/CaptureDisk01/20251105/Test02_vitc_frame_timecodes.txt: set owner/group (was: 501/0): Operation not permitted
mv: /Volumes/CaptureDisk01/20251105/Test02_frame_eia608data.txt: set owner/group (was: 501/0): Operation not permitted
[Parsed_drawtext_30 @ 0x7ff3ed257200] [FILE @ 0x70001020d598] Cannot read file '/tmp/vrecord.sC44Yu.vrecord_input.log': No such file or directory
[Parsed_drawtext_30 @ 0x7ff3ed257200] The text file '/tmp/vrecord.sC44Yu.vrecord_input.log' could not be read or is empty
[Parsed_drawtext_30 @ 0x7ff3ed257200] [FILE @ 0x70001020d598] Cannot read file '/tmp/vrecord.sC44Yu.vrecord_input.log': No such file or directory
[Parsed_drawtext_30 @ 0x7ff3ed257200] The text file '/tmp/vrecord.sC44Yu.vrecord_input.log' could not be read or is empty
Vrecord is analyzing your video file. Please be patient.
qctools XML /Volumes/CaptureDisk01/20251105/Test02.mkv.qctools.mkv is empty or does not exist!
QCTools analysis is complete.
Checking file conformance against FFV1 video policy...
 2024-11-05T16:06:22 - File did not pass vrecord policy check for FFV1 video and may not conform to digital preservation standards. Try another file?
 2024-11-05T16:06:22 - See /Volumes/CaptureDisk01/20251105/Test02_mediaconchreport.xml for a full MediaConch policy report.

[iamdamosuzuki]

I dug into this a bit and it seems like it's being caused by a Segmentation fault with qcli. I couldn't exactly determine that the tee and piping to qcli (around line 2666) is correct, but, but it looks like it should be right.

This could possible be related to the segmentation fault I mentioned in the other vrecord ticket (#851), which has an open issue here: bavc/qctools#842

Sorry to have made all these tickets. I'm happy to keep this one open until the issues with qcli are addressed if that's ok with you.

[privatezero]

Sounds good! Thanks for testing!

[iamdamosuzuki]

This has been fixed with QCTools 1.4 and this QCTools pull request: bavc/qctools#849 . Thanks all!