Commit graph

792 commits

Author SHA1 Message Date
HifiExperiments
6156495345 stupid, shadows still broken 2019-12-12 23:04:47 -05:00
SamGondelman
1736252f0d trying and failing to fix models 2019-12-12 23:04:47 -05:00
HifiExperiments
e0a29d9367 material entities support cullFaceMode and support gltf doubleSided 2019-12-12 23:01:57 -05:00
HifiExperiments
bae2ba7bae merge with master 2019-11-02 01:08:56 -07:00
Anthony J. Thibault
92f1b59cc8 Full legacy API and FST support 2019-10-18 16:30:32 -07:00
SamGondelman
9c1b139d5f merge with master 2019-06-25 23:46:39 -07:00
Ken Cooke
4c5c61231d Only use QVector resize when actually needed 2019-06-25 12:57:55 -07:00
Ken Cooke
0a34745ff9 Use QVector resize() instead of reserve() to silence Debug builds about size not being set.
Causes unneeded default initialization, but no obvious way to avoid.
2019-06-25 12:53:07 -07:00
Ken Cooke
75aea88f20 CR feedback 2019-06-24 11:51:03 -07:00
Ken Cooke
87e0f5b2bb Runtime dispatch for AVX2 2019-06-23 10:40:15 -07:00
Ken Cooke
8653118b6c Full SIMD implementation of packBlendshapeOffsets() using AVX2.
6x speedup over the existing (partial SIMD) version.
60x speedup over the original (pure GLM) version.
2019-06-23 10:32:57 -07:00
Ken Cooke
cceff21cd0 Pull packBlendshapeOffsets() into separate function 2019-06-23 08:57:56 -07:00
Ken Cooke
87c680382f Strip-mining optimization to improve cache utilization.
For each mesh: init, accumulate, and pack using a recycled offset buffer.
2019-06-23 08:42:38 -07:00
Ken Cooke
9e309b095d Fix bug that was packing zero offsets that were never used 2019-06-23 08:00:13 -07:00
Sam Gondelman
a7842fe3ea
Merge branch 'master' into proceduralMesh 2019-05-21 09:51:33 -07:00
SamGondelman
0039fb2f54 attempt to fix material crash 2019-05-17 14:51:42 -07:00
SamGondelman
6d3e23a7fe merge with master 2019-05-16 20:03:10 -07:00
SamGondelman
de90c5088c (✿◠‿◠) 2019-05-16 19:33:07 -07:00
Anthony J. Thibault
3ae33353d9 Fixed unused variables warnings 2019-05-16 16:38:47 -07:00
Anthony J. Thibault
98fe059d97 Fix blender::run crash
Pass a shared_ptr of the HFMModel to the Blender.  This will prevent the HFMModel
from being destroyed on the main thread if an Avatar changes their skeletonModelURL.

Also the _blendshapeOffset hash in Model has been eliminated, it was not necessary and
was also a source of data races.

The body of Blender::run has been updated slightly to reduce the number of allocation necessary for temporary QVectors.
2019-05-16 13:12:16 -07:00
jennaingersoll
4c6c0acc56
Merge pull request #15454 from ctrlaltdavid/M22246
Case 22246: Picks, PickType, RayPick JSDoc update
2019-05-13 16:16:17 -07:00
SamGondelman
eb8c722080 fix runtime switch 2019-05-13 09:58:55 -07:00
Brad Davis
17632e3c50 PR feedback 2019-05-07 10:36:14 -07:00
Brad Davis
2b1267ffeb Disable the use of tbb::parallel_for in Blender 2019-05-06 16:53:37 -07:00
Shannon Romano
29e415a104
Merge pull request #15498 from SamGondelman/crash
Case 22444: Try to fix materialMapping crash
2019-05-02 16:29:39 -07:00
SamGondelman
50e538a73c try to fix materialMapping crash 2019-05-01 17:23:19 -07:00
SamGondelman
5a184bd584 fix model bounds 2019-04-30 13:38:11 -07:00
David Rowe
73fd61dcf0 Picks, PickType, RayPick JSDoc update 2019-04-25 13:19:07 +12:00
SamGondelman
2d5cfbf4f5 fix oven crash 2019-04-16 15:16:26 -07:00
David Rowe
38e21dcaa2 Doc review 2019-04-11 11:04:02 +12:00
David Rowe
ca3f45163f Merge branch 'master' into M21989 2019-04-11 10:37:34 +12:00
Sam Gateau
51d1c4b1f3
Merge pull request #15357 from SamGondelman/cauterize
Case 21417: Fix cauterization of entities that are children of MyAvatar's head
2019-04-10 12:29:40 -07:00
SamGondelman
a50cca006e use # instead of ? for material URL name selection 2019-04-08 23:45:47 -07:00
SamGondelman
27338c3639 fix cauterization of entities that are children of MyAvatar's head 2019-04-08 19:48:11 -07:00
Sam Gondelman
88c278d357
Merge pull request #15319 from SamGondelman/materialBaker
Case 21764: Oven bakes model materials directly to separate material description
2019-04-05 17:08:22 -07:00
Andrew Meadows
72ac35c6f7 avoid nan in Model::_scale 2019-04-04 08:28:03 -07:00
David Rowe
2a17ad3da5 AvatarList and AvatarManager JSDoc 2019-04-04 10:11:54 +13:00
SamGondelman
770b863604 fixing embedded textures and materialMapping priority 2019-04-02 19:04:23 -07:00
SamGondelman
7b0854c121 support selecting material in json by name with ? 2019-04-02 14:13:00 -07:00
Sam Gondelman
0f99280c15
Merge pull request #15265 from hyperlogic/bug-fix/blendshape-precision
Improve blendshape precision for small models
2019-03-27 10:51:22 -07:00
Anthony Thibault
08c6acdf99 Improve blendshape precision for small models
For avatars authored in meters, the max component length would often be less then 1.
In that case, the blendshape offset was not normalized before quantization, resulting in loss of precision.
This change will normalize the offset for all cases, except when the max component length is 0.
2019-03-26 11:21:13 -07:00
SamGondelman
c1c45d8a01 revert some now unnecessary changes that were impacting performance
(cherry picked from commit b6984de16c2fd17f04ea72de7b339c31db6467ab)
2019-03-22 11:55:04 -07:00
SamGondelman
2b32b77bed handle case when clusterMatrices.size() == 0 2019-03-11 17:32:40 -07:00
SamGondelman
b24b7fed3d the root node isn't the first onegit add ../.git add ../. 2019-03-11 15:34:41 -07:00
Shannon Romano
8be93fd16c
Merge pull request #14891 from kencooke/glm-pack-normals-speedup
Case 21114: SIMD optimized blendshape packing
2019-03-07 15:28:40 -08:00
SamGondelman
fd88ec0d16 need to copy meshStates on main thread 2019-03-04 15:25:02 -08:00
SamGondelman
708632ee82 possible fix for model crash 2019-03-04 10:00:26 -08:00
Shannon Romano
5f99158c25
Merge pull request #14977 from SamGondelman/rot
Case 21354: Add "modelScale" property to fix pal
2019-02-22 16:36:17 -08:00
SamGondelman
6d6cd42adb fix bound scaling 2019-02-22 12:00:00 -08:00
Anthony Thibault
e54754c4ce
Merge pull request #14926 from luiscuenca/flowCpp
Cpp implementation of the Flow script
2019-02-21 17:10:05 -08:00