| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2025-10-15 | Add another new test reference for FFmpeg 8. | Carl Hetherington | |
| Commit 045a8b15b19ec7f872fb01cfb986faeaa26cb8bb changed how AC3/EAC3 decoders in FFmpeg are flushed, and this caused a considerable change to the audio result in this test (in particular, the audio is about 3db louder after the FFmpeg change). I'm assuming the previous result was just wrong. | |||
| 2025-09-23 | Return quite close to original approach for "no colour conversion". | Carl Hetherington | |
| There's a few things going on here: 1. Improve the regression test for 3042. Previously we made a DCP from the reporter's _original_ prores file (before they converted it to XYZ) and compared the result to a reference J2K file of uncertain origin. This seems wrong because: a) We never got confirmation from the reporter that the fix worked for them, so any arbitrary reference is dubious. b) It doesn't seem to reflect their actual complaint, which was that they got a different result when making a DCP from XYZ TIFFs compared to their "XYZ" Prores. The new test makes a DCP from their TIFFs and "XYZ" Prores and compares the result. 2. Revert to the old approach to "no conversion" handling. In the good old days we did everything -> RGB48LE except XYZ12LE -> XYZ12LE, and that's what we do again here. 3. Change the YUV->RGB conversion from Rec.601 to Rec.709 for the "no conversion" case. This fixes the 3042 regression test. The supposed "XYZ" Prores is yuv444p12le according to ffprobe. So I think what we have here is actually a file that was converted to XYZ and then back to YUV by Resolve. I experimented with using the raw YUV values and considering them as XYZ but this was clearly wrong. I think 3 is probably what I should have done in the first place. | |||
| 2025-08-12 | New/improved pixel format decision when converting to XYZ. | Carl Hetherington | |
| For a long time we would keep XYZ12LE, if that's what we have, otherwise ask FFmpeg to switch to RGB48LE. Then in 1d5c211dadb9a9dc2318adce86ca9c31b367cabe I tried to fix the case of an XYZ source mis-tagged as YUV. I changed things so that with no colour conversion we'd always ask FFmpeg to convert to XYZ. This meant that RGB sources with no colour conversion would get treatment by FFmpeg due to the RGB -> XYZ switch. Here we're going back to the more-or-less the "long time" behaviour when there is a conversion (keep XYZ12LE but otherwise convert to RGB48). When there's no conversion, keep RGB (to avoid the FFmpeg conversion from RGB -> XYZ) but convert everything else to XYZ. | |||
| 2025-05-26 | Fix incorrect colours with "no colourspace conversion" sources that are not ↵ | Carl Hetherington | |
| AV_PIX_FMT_XYZ12LE. Prepare the image for where it is going to next. I don't understand the difference in this situation between XYZ12LE and RGB48LE (it seems they are both 16-bit-per-channel) but there you go. | |||
| 2025-05-19 | Tidy up finding required files during tests. | Carl Hetherington | |
| In particular, prefer not to use lots of fallback attempts. It seems that something funny is going on with how we find Liberation Sans on Linux, meaning that we were sometimes using the OS copy of the font when we didn't mean to. | |||
| 2024-12-09 | Fix decoding of SSA subs in files decoded by FFmpeg (#2904). | Carl Hetherington | |
| Suddenly we have 8 commas, not 9, perhaps because of 29412821241050c846dbceaad4b9752857659977 in ffmpeg (although that's strange, because it was a long time ago). | |||
| 2024-11-01 | Merge remote-tracking branch 'origin/main' into v2.17.x | Carl Hetherington | |
| 2024-11-01 | Obtain audio length by scanning through the file if required (#2880). | Carl Hetherington | |
| 2024-06-25 | Fix thinko in run/tests on macOS. | Carl Hetherington | |
| 2024-06-24 | Fix running tests on arm64. | Carl Hetherington | |
| 2024-05-29 | Merge remote-tracking branch 'origin/main' into v2.17.x | Carl Hetherington | |
| 2024-05-21 | Allow running tests from trees made with cdist. | Carl Hetherington | |
| 2024-05-20 | Use $HOME rather than hard-coded user name. | Carl Hetherington | |
| 2024-05-19 | Fix thinko in previous. | Carl Hetherington | |
| 2024-05-19 | Missing setup of DYLD_LIBRARY_PATH. | Carl Hetherington | |
| 2024-05-16 | Merge branch 'main' into v2.17.x | Carl Hetherington | |
| 2024-05-15 | Fix macOS environment setup. | Carl Hetherington | |
| 2024-05-10 | Make sure build/test exists. | Carl Hetherington | |
| 2024-04-22 | Unify indentation at 4 spaces. | Carl Hetherington | |
| 2024-04-22 | Use run/environment for paths in run/tests. | Carl Hetherington | |
| 2024-04-18 | Missed update to private test repo version.v2.16.80 | Carl Hetherington | |
| 2024-03-19 | The private test repo got some v2.18.x updates. | Carl Hetherington | |
| 2024-02-03 | Don't emit subtitle images that have a zero dimension (#2743). | Carl Hetherington | |
| They cause problems later when trying to blend them into the image. | |||
| 2024-01-15 | Inspect most DCPs made during tests with ClairMeta (#76). | Carl Hetherington | |
| 2024-01-15 | Inspect most DCPs made during tests with dcp_inspect (#76). | Carl Hetherington | |
| 2024-01-15 | Missing PRIVATE_GIT update from previous. | Carl Hetherington | |
| 2023-12-09 | Support YUV422P9LE in alpha_blend() (#2671).notar1 | Carl Hetherington | |
| 2023-09-29 | Support alpha blending 64-bit RGB onto things. | Carl Hetherington | |
| 2023-09-29 | Fix errors with WAVs containing markers (#2617). | Carl Hetherington | |
| I'm not 100% sure about this but they seem to end up giving audio packets with no channels and no frames. Here we handle such packets better. | |||
| 2023-08-26 | Set up packet duration correctly when encoding using FFmpeg (#2588). | Carl Hetherington | |
| It seems strange that this is necessary (maybe I'm missing some other way that the packet duration is supposed to be set up). Without this, when the mov muxer writes the trak tags it uses an incorrect length value because the length value is calculated from trk->end_pts, which in turn is calculated from the last-seen pts + the duration of the last packet. If that packet is marked as length 0 the length of the track is 1 frame short, so the export is missing a frame. | |||
| 2023-08-19 | Yet more private file fixing. | Carl Hetherington | |
| 2023-08-19 | More private test files tinkering. | Carl Hetherington | |
| 2023-08-19 | Private test files repo change. | Carl Hetherington | |
| 2023-06-01 | Bump required private git revision. | Carl Hetherington | |
| 2023-04-14 | Update dcpomatic-test-private for some v2.17.x changes that don't affect main. | Carl Hetherington | |
| 2023-03-29 | Stop running ./check_sums now we're using git annex. | Carl Hetherington | |
| 2023-03-29 | Always make 16-channel MXFs. | Carl Hetherington | |
| 2023-03-21 | Bump private git sha.v2.16.48 | Carl Hetherington | |
| 2023-03-18 | Bump private test git for use of git-annex. | Carl Hetherington | |
| 2023-02-27 | Fix tests. | Carl Hetherington | |
| 2022-11-22 | Bump private git so that we use Liberation Sans rather than Arial for a ↵ | Carl Hetherington | |
| subtitle test. | |||
| 2022-11-22 | Only check test private data if --check is passed. | Carl Hetherington | |
| 2022-11-22 | Check that the private test data repo is as we expect before running tests ↵ | Carl Hetherington | |
| on Linux/macOS. | |||
| 2022-09-05 | Only use --log_level=test_suite when running from cscript. | Carl Hetherington | |
| 2022-07-05 | Allow running tests on Fedora. | Carl Hetherington | |
| 2022-07-01 | Support Fedora libpath when running tests. | Carl Hetherington | |
| 2022-03-09 | Improve ratings dialog to allow only valid values (#2199). | Carl Hetherington | |
| 2021-12-14 | Add missing library search path for macOS tests.v2.15.179 | Carl Hetherington | |
| 2021-05-21 | Fix path to test libraries on macOS. | Carl Hetherington | |
| 2021-01-08 | Fix test running on macOS. | Carl Hetherington | |
 |
index : dcpomatic | |
| DCP-o-matic DCP tools | cth@carlh.net |
| summaryrefslogtreecommitdiff |