Test failures with ffmpeg 7.0

[lfos]

Running the test suite with 08d7f33 (Add compatibility with FFMPEG 7.0, 2024-05-08) and with ffmpeg 7.0, there seem to be several test failures:

The following tests FAILED:
	30 - Clip:verify parent Timeline (Failed)
	34 - Clip:time remapping (Failed)
	35 - Clip:resample_audio_8000_to_48000_reverse (Failed)
	92 - FrameMapper:resample_audio_48000_to_41000 (Failed)
	93 - FrameMapper:resample_audio_mapper (Failed)
	172 - Timeline:Multi-threaded Timeline Add/Remove Clip (Failed)

The failures themselves all look fairly similar and contain output such as

[SWR @ 0x7df3fcb2e380] Input channel layout "" is invalid or unsupported.
[SWR @ 0x7df3fcb2e380] Context has not been initialized

@eclipseo Any idea? Do you remember whether you manually ran the full test suite with ffmpeg 7.0 and with the final version of the changes proposed in PR #954?

It'd also be good to extend the GitHub workflows to run the build and full test suite with ffmpeg 7.0.

[VVD]

Is this fixed in 0.4.0?

Changelog for OpenShot 3.3 says:
FFmpeg Compatibility: Upgraded support for newer FFmpeg versions ensures stability on modern systems.
https://www.openshot.org/blog/2024/12/22/new_openshot_release_330/

[lfos]

I just checked. Not fixed. It looks like there still the same issues.

[stale]

Thank you so much for submitting an issue to help improve OpenShot Video Editor. We are sorry about this, but this particular issue has gone unnoticed for quite some time. To help keep the OpenShot GitHub Issue Tracker organized and focused, we must ensure that every issue is correctly labelled and triaged, to get the proper attention.

This issue will be closed, as it meets the following criteria:

• No activity in the past 90 days
• No one is assigned to this issue

We'd like to ask you to help us out and determine whether this issue should be reopened.

• If this issue is reporting a bug, please can you attempt to reproduce on the latest daily build to help us to understand whether the bug still needs our attention.
• If this issue is proposing a new feature, please can you verify whether the feature proposal is still relevant.

Thanks again for your help!

[VVD]

Probably this is not fixed.

[rathann]

@eclipseo Any idea? Do you remember whether you manually ran the full test suite with ffmpeg 7.0 and with the final version of the changes proposed in PR #954?

It'd also be good to extend the GitHub workflows to run the build and full test suite with ffmpeg 7.0.

Unfortunately, the test suite failures were silently ignored in our package builds, so @eclipseo did not catch this, sorry about that. The problem is, we have no H.264 decoding support out-of-the box in Fedora, so many of the tests are failing because the Sintel trailer video is encoded with H.264 and FFmpeg is unable to find any decoder for H.264 video. If you replaced the sample video with something encoded with an unencumbered codec like VP9 or AV1 (e.g. https://commons.wikimedia.org/wiki/File:Big_Buck_Bunny_Trailer_1080p.ogv), we could run the full suite. Regardless, I'm fixing our package so that we skip only the tests that fail due to missing H.264 codec. That way, we can at least catch some regressions.

[lfos]

@rathann - The error is not specific to Fedora. My original report was filed based on us noticing the errors when packaging OpenShot for Arch Linux. And #1007 seems to suggest the issue exists under Windows too when using ffmpeg 7.

It'd be great to have the tests fixed instead of having to selectively ignore tests.