mirror/obs-studio
1
0
Fork 0
mirror of https://github.com/obsproject/obs-studio.git synced 2026-02-01 18:12:12 -05:00
Code Issues Packages Projects Releases Wiki Activity
12,812 Commits 18 Branches 236 Tags
4aaffe16ec58a1df3ddb6dcd5994a2bb269397ba
Commit Graph

98 Commits

Author SHA1 Message Date
Lain
106c7aa61f Update copyrights/names 2023-05-20 01:31:18 -07:00
jpark37
7ff5dd63b9 libobs: Fix luma sampling for packed 4:2:2 sources 2023-05-06 16:31:06 -07:00
jpark37
901986cbf4 libobs,obs-filters: Align HDR to SDR disparity 
Adjust colors from BT.1886 to sRGB.
 2023-04-23 23:43:09 -07:00
Norihiro Kamae
94bf325dd9 libobs: Fix VIDEO_FORMAT_V210 shader for GLSL 2023-03-26 05:13:00 -07:00
jpark37
97843dd7d7 libobs: Add VIDEO_FORMAT_V210 
Support 10-bit packed format that DeckLink uses.
 2023-03-26 01:52:44 -07:00
jpark37
463bf0dff5 libobs,UI: Add P216/P416 pixel formats 
Will be useful for ProRes.
 2023-02-21 18:48:44 +01:00
jpark37
0efaf88caf libobs: Add PQ/HLG support for I210/I412 formats 
Seen when playing back ProRes files.
 2023-02-21 18:48:44 +01:00
jpark37
e709797d80 libobs: Add P3 shaders for Mac 2022-11-09 22:34:50 -08:00
jpark37
e119a7024d libobs: Sample video at default chroma location 
Expect left for 4:2:0 SDR and 4:2:2, and top-left for 4:2:0 HDR.

MJPEG will be a half-pixel off, but we'll live with that for now.
 2022-10-09 16:15:16 -07:00
jpark37
51c452e937 libobs: Precompute more accurate matrices 
These are generated with .NET BigInteger fractions, and late divide.
 2022-10-05 08:36:29 -07:00
jpark37
bb02620169 libobs: Add support for reading NV12/YUY2 PQ/HLG 
8-bit HDR shouldn't exist, but OBS looks broken, so just support it.
 2022-09-17 16:36:31 -07:00
jpark37
1a7e5babc2 libobs: Add support for reading I420 HLG 
Sony cameras can create 8-bit HLG videos for some reason.
 2022-09-16 03:04:23 -07:00
jpark37
3a43fc4a07 libobs: Use left chroma location for SDR 
Top-left chroma should be for Rec. 2020/2100.
 2022-07-31 00:10:50 -07:00
Anastasiia Sidorovskaia
ef9c99f5f4 libobs: Set coefficients with higher precision 2022-07-25 17:41:26 -04:00
Jim
29782cd594 libobs: And fix area scaling effect with RGBA 2022-06-04 19:54:52 -07:00
Jim
a33a5d2151 libobs: Fix bilinear lowres RGBA as well 2022-06-04 19:03:20 -07:00
Jim
4a2a06b22f libobs: Fix RGBA format output not working 
Due to a bug in shader parsing, it thinks that because the token
"multiplier" is used here, that the "multiplier" uniform is being used.

This is a workaround for the issue because fixing the parser is probably
going to be much more annoying than just working around the issue for
now.
 2022-06-04 17:47:47 -07:00
jpark37
bacd4713da libobs: Clear low bits when writing P010 
Don't want to rely on consumer to ignore those bits.
 2022-06-04 01:00:21 -07:00
jpark37
11da542a0d libobs: Add max_luminance to obs_source_frame 
Used in situations where source luminance is greater than HDR nominal
peak setting to avoid clipping by applying BT.2408 maxRGB EETF.
 2022-05-27 14:56:47 -07:00
jpark37
4304329d0c libobs: Ignore lower six bits for P010 sources 2022-05-17 02:46:41 -07:00
Norihiro Kamae
dd43635e78 libobs: Fix reserved word in variable names 
The use of the reserved name caused initialization failure on Linux.
 2022-05-11 11:28:20 -04:00
jpark37
0f53dc28bb libobs: Reduce PQ shader math 2022-05-11 03:38:07 -07:00
jpark37
2a0d8d1c9c libobs: Add support for reading I420 PQ 
Not normally a valid combination, but Xbox writes 8-bit HDR videos.
 2022-05-08 14:12:41 -07:00
jpark37
ed835810b4 libobs: Use tabs in format_conversion.effect 2022-05-08 14:12:41 -07:00
Jim
952988d9ec Merge pull request #6231 from mvji/prores_pix_fmt 
Add support for GPU conversion of YUV422P10LE, YUV444P12LE, YUVA444P12LE
 2022-05-04 02:01:27 -07:00
jpark37
a82bbb416f libobs: Fix NaNs when using EETF for HLG 2022-04-26 10:11:28 -07:00
jpark37
ed85307f7e libobs: Clean up color.effect a bit 2022-04-26 10:11:28 -07:00
mvji
d3a8ef7128 libobs: Add support for YUV422P10LE, YUV444P12LE, YUVA444P12LE 2022-04-19 19:37:07 +02:00
jpark37
338608bd67 libobs,UI: Support HLG nominal peak level 
HLG output uses MovieLabs-recommended procedure.

- If peak luminance is greater than 1000, use maxRGB EETF to 1000.
- Otherwise, don't tonemap.
- Then use normal HLG conversion procedure with gamma 1.2 (1000 nits).
 2022-04-14 09:36:44 -07:00
jpark37
08f50a7d22 libobs: Add more color handling to default/opaque 2022-04-13 06:23:11 -07:00
jpark37
06111d5b10 libobs: Add high-precision sRGB support 2022-04-08 17:19:23 -07:00
jpark37
0ed0f2cdb4 libobs: Add I010/P010 support, TRC enum 2022-04-03 00:01:25 -07:00
jpark37
c455aaedba libobs: Add color spaces to deinterlace shaders 2022-04-02 20:14:31 -07:00
jpark37
a8bc994f07 libobs: Add color spaces to scale shaders 2022-03-26 13:00:34 -07:00
jpark37
87ab39c412 libobs: Render main texture for active color space 
Preview will draw SDR white luminance from settings (default 300 nits)
when displayed on an HDR monitor rather than CCCS 80 nits.
 2022-03-23 22:35:27 -07:00
jpark37
4e6765a01b libobs: Remove DrawSrgbDecompressPremultiplied 
Technique is no longer referenced, and doesn't seem useful.
 2021-10-02 05:53:27 -07:00
jpark37
10cf411f99 libobs: DrawSrgbDecompress for default_rect.effect 
Necessary for upcoming fix to browser source alpha.
 2021-09-30 17:35:06 -07:00
jpark37
c9766d8e28 libobs: Add DrawSrgbDecompress default technique 
Useful when the texture does not support SRGB conversion on load.
 2021-07-11 08:26:05 -07:00
jpark37
05b507d900 libobs: DrawSrgbDecompressPremultiplied technique 
Necessary for an upcoming fix to browser source.
 2021-07-11 08:11:12 -07:00
jpark37
b70161bc67 libobs: Add DrawOpaque for rect effect 
Needed by Syphon Client.
 2021-07-04 10:35:54 -07:00
jpark37
a8f0a27a3a libobs: Remove DrawAlphaBlend technique 2021-05-11 13:12:39 -07:00
jpark37
de7cec3dee libobs: Add DrawAlphaBlend technique 
Useful when fixed-function blend does not provide enough precision.
 2021-05-05 09:43:34 -07:00
jpark37
840e2b3d43 libobs: Add DrawNonlinearAlpha technique 
This allows OBS to mimic other programs that incorrectly use alpha.

Requires premultiplied blend state.
 2021-04-24 17:18:42 -07:00
jpark37
9bdb16aa78 libobs: Fix Area shaders missing for RGB output 
Area downscale setting currently only works with YUV outputs. This adds
the missing DrawAlphaDivide technique.
 2019-08-31 00:25:24 -07:00
jpark37
af01e044a2 libobs: Fix Lanczos calculations 
- Fix: Ensure (1, 1) coordinate gets clamped.
- Fix: Increase weight precision by premultiplying UV in VS.
- Cleanup: Group coordinates 012/345 instead of 024/135.
- Cleanup: Remove unnecessary branches.

NVIDIA RTX 2080 Ti, Intel GPA, SetStablePowerState

256x224 -> 1323x1080: 123 us -> 123 us
 2019-08-25 10:00:23 -07:00
jpark37
3485c4cdac libobs: Simplify bicubic weight calculations 
Also increase weight precision by premultiplying UV in VS.

Intel HD Graphics 530, Intel GPA, SetStablePowerState

256x224 -> 1323x1080: 1221 us -> 1020 us
 2019-08-25 10:00:10 -07:00
jpark37
9f5d218e16 libobs: Remove unnecessary divides from Lanczos 2019-08-14 21:36:23 -07:00
jpark37
93f1ab789d libobs: Fix dark lines using Lanczos 
When texel samples are not exactly on texel centers, weight calculations
will involve a divide by a number very close to zero, resulting in
precision issues. Restore normalization of weights to compensate.
 2019-08-14 21:00:09 -07:00
jpark37
3d6f5c8ad6 libobs: Add YUV alpha formats 
This will allow YUV alpha formats to be converted to RGBA on the GPU.
 2019-08-11 11:26:22 -07:00
Jim
31a902b3af Merge pull request #2018 from jpark37/yuv-simplify2 
libobs: Separate textures for YUV input
 2019-08-10 22:41:28 -07:00