dcpomatic/run, branch support-080626 DCP-o-matic DCP tools https://git.carlh.net/cgit/dcpomatic/atom?h=support-080626 2026-04-23T14:51:08Z Fix reading of RGB0 images. 2026-04-23T14:51:08Z Carl Hetherington cth@carlh.net 2026-04-22T20:24:09Z urn:sha1:e54f93bf7d144737df4b0331017164360c9121ff We would allocate 3 bytes per pixel instead of 4. I couldn't see a way to get FFmpeg to tell us about this (unless we used FFmpeg's stride somehow maybe?) Fix bug causing mangled audio analyses in some cases (#3155). 2026-04-21T20:26:25Z Carl Hetherington cth@carlh.net 2026-04-21T20:26:25Z urn:sha1:824350b6e66ce7ef8a5449b60ce1abb2e4f462a5 The duplicate ID detection was broken due to using the video stream index rather than its ID for checks. DYLD_LIBRARY_PATH tweaks for macOS. 2026-04-20T09:13:35Z Carl Hetherington cth@carlh.net 2026-04-20T09:13:35Z urn:sha1:ceeeaf107dd484a984be1737b38f2d301462a8bf Fix erroneous hints about ccaps having more than 3 lines if they contain formatting. 2026-04-15T20:30:36Z Carl Hetherington cth@carlh.net 2026-04-15T20:30:36Z urn:sha1:6b8e08be8000824e2af0f24dbc86199dd427fcf6 Bump private test data version. 2026-04-08T08:07:56Z Carl Hetherington cth@carlh.net 2026-04-08T08:07:56Z urn:sha1:0b98de7d1c2c0ae40c6ffa14dd8ebbad3e0b08f9 Fix environment to load libs from lib64. 2026-02-01T00:37:50Z Carl Hetherington cth@carlh.net 2026-02-01T00:37:50Z urn:sha1:9ad55d5c76043afa89ef6fcb692899c610fa82d6 Fix mismatched audio reel length (#3123). 2025-12-31T20:38:12Z Carl Hetherington cth@carlh.net 2025-12-30T01:38:54Z urn:sha1:36c5c4cdc1c5f47125118edba2593e851ce0756b Fix artefacts on the edge of YUV images with odd widths (#3119). 2025-12-15T07:21:56Z Carl Hetherington cth@carlh.net 2025-12-15T00:39:56Z urn:sha1:6e8f8563f4e4e6ff04b632aff8d6ae95de945649 This also modifies an existing test to expose the bug. Try AVSEEK_FLAG_ANY if _BACKWARD fails. 2025-12-08T19:50:53Z Carl Hetherington cth@carlh.net 2025-12-07T01:17:34Z urn:sha1:9bd8b3304f35f0b27bf18f2167302f1fd5d3f92b We usually want to do _BACKWARD as this should seek to the keyframe before the seek position. On some files this fails, and then it seems that _ANY is a good next thing to try. Missing check for duplicate IDs. 2025-11-03T00:50:10Z Carl Hetherington cth@carlh.net 2025-11-03T00:10:39Z urn:sha1:b3056163b9da4425aab25e3c9630a805a48d30ff Without this call we don't spot duplicates that are found in the initial scan.