Preserve the sequence of changes
Record what was changed for the first video and whether the new problem begins with the next one. Keep the two items’ known scan characteristics separate. A setting chosen to address missing interlace metadata on one file should not be assumed appropriate for every later programme.
For an illustrative pair, one older interlaced recording and one progressive sample can need different handling. Their similar resolution or container extension does not establish the same scan structure. Do not describe an unverified example as a test performed by this article.
Return to the appropriate baseline
Consult the current QMPlay2 video-filter guidance and restore the normal supported detection behaviour where relevant. Use the same second-video passage for the comparison. Keep decoder and output choices fixed initially so the effect of removing the manual override remains understandable.
If the image improves after that single reversal, record the result and why the first override was introduced. Preserve a note for the unusual original source instead of leaving a global change active indefinitely. A source-specific exception should be recognisable the next time it is needed.
Investigate remaining defects separately
If the second video still looks wrong with the expected baseline, keep that outcome and describe the actual artefact. Colour, scaling, motion cadence and field-related combing are different observations. Do not continue changing deinterlacing to address every picture complaint.
Include player version, platform, decoder/output route, known source scan type and the before-and-after filter state in a support note. Redact private paths. This article applies the publisher’s explicit revert instruction; it does not guarantee that every source identifies its scan structure correctly or that one automatic choice repairs every file.