mrchirster: this sounds like a very good test for Magic Lantern bitrate settings.
Can you go to the same place (to have the same scene, similar lighting if possible, same lens and same settings) and do some tests?
1) What QScale was used (ML displays it as a tiny number near the red dot)? Just look at some approximate values, or range. Bitrate setting should be "FW default" or whatever you have used in your original shoot.
2) Record a clip with:
- default bitrate
- with CBR setting pushed as far as possible without recording having to stop.
- with QScale -16, even if it will stop recording after 1 second or so.
3) Take a few silent pictures during recording (at any compression setting; it doesn't matter; these are the uncompressed 422 frames, before upscaling and entering into H264 engine).
Can you go to the same place (to have the same scene, similar lighting if possible, same lens and same settings) and do some tests?
1) What QScale was used (ML displays it as a tiny number near the red dot)? Just look at some approximate values, or range. Bitrate setting should be "FW default" or whatever you have used in your original shoot.
2) Record a clip with:
- default bitrate
- with CBR setting pushed as far as possible without recording having to stop.
- with QScale -16, even if it will stop recording after 1 second or so.
3) Take a few silent pictures during recording (at any compression setting; it doesn't matter; these are the uncompressed 422 frames, before upscaling and entering into H264 engine).