Ignacio
58617584d4
Apply patch provided in issue #224 . Add support for ARM64.
9 years ago
Ignacio
4ef408b591
Remove debug code.
9 years ago
Ignacio
592c8b8b0a
Merge pull request #225 from Skylark13/master
...
Fix compilation on VS2013 (non-CUDA)
9 years ago
Jean-S�bastien Guay
b254961b22
- Removed Library.cpp and Library.h from nvcore VS2013 project
...
- Also added project references to nvtt in nvdecompress and nvimgdiff projects, lack of which prevented them from compiling
- Change to nvtt.sln is an auto-modification from VS, pretty sure it makes no difference
9 years ago
Ignacio
d32f0e0173
Merge pull request #223 from cmdrf/no-include-library-h
...
Do not include Library.h
9 years ago
Ignacio
701f8224dc
Merge pull request #222 from cmdrf/gccforeach
...
Update NV_FOREACH for C++11
9 years ago
cmdrf
7b031aa4bf
Do not include removed Library.h
9 years ago
cmdrf
d01d72f8e7
Update NV_FOREACH for C++11
9 years ago
Ignacio
0d205015ae
Merge pull request #221 from cmdrf/snprintf-vs2015
...
Do not redefine snprintf on VS2015
9 years ago
Ignacio
7aa0390f32
Merge pull request #220 from cmdrf/cmakecuda
...
Fix CMake CUDA introspection
9 years ago
Ignacio
68e6cbdd40
Merge pull request #219 from cmdrf/gcclinkorder
...
Fix Link Order for GCC in CMake
9 years ago
cmdrf
fa062aa298
Do not redefine snprintf on VS2015
9 years ago
cmdrf
205f492e2e
Fix CMake CUDA introspection
9 years ago
cmdrf
899a99a106
Fix link order for Linux
9 years ago
cmdrf
766edcb1a6
Fix more link order for Linux
9 years ago
Ignacio
b37ca4498d
Merge pull request #217 from pzychotic/master
...
Fixed compiler flag being set on wrong target.
9 years ago
Thomas Roß
9ad63e7f3a
Fixed compiler flag being set on wrong target.
9 years ago
Ignacio
81336cc3e9
Remove unnecessary files.
9 years ago
Ignacio
e9d44813d2
Merge pull request #216 from esok/master
...
Added support for gcc and c++11 features.
9 years ago
Ignacio
d4fbfd0ff4
Add project notes.
9 years ago
Ignacio
4c7ff58885
Update vc12 projects
9 years ago
Ignacio
a083337473
Merge changes from The Witness.
9 years ago
esok
6bc3224fe7
Added support for gcc and c++11 features.
9 years ago
castano
7e2a9d1adb
osx fixes. Fix issue 211.
10 years ago
castano
2d6fc0e304
DXT1 compressor tweaks.
10 years ago
castano
d019cd7080
Merge changes from the witness.
10 years ago
castano@gmail.com
4cb60cc5ba
Fix vc12 build.
10 years ago
castano@gmail.com
9722101114
Cleanup vc12 project.
10 years ago
castano@gmail.com
3f0bffaadf
Add vc12 solution.
10 years ago
castano
6e69560969
Tag dll and binaries with 32/64 suffix.
10 years ago
castano
8f4ebf1d93
Remove missing project.
10 years ago
castano
132443013a
Minor updates to C# bindings. Work in progress.
10 years ago
castano
ac87eefa6d
Apply patch to update C# wraper. Fixes issue 184.
10 years ago
castano
f705b2eaa1
Fixes issue 93.
10 years ago
castano
6ade0e27b7
Fixes issue 202.
10 years ago
castano
7e74261f92
Fixes issue 204.
10 years ago
castano
57298c71c8
Fix issue 194.
10 years ago
castano
b06a6f46a5
Fixes issue 207.
10 years ago
castano
63d6036e3a
Fix issue 206.
10 years ago
nathaniel.reed@gmail.com
ab316deeaa
Add BC7 support. It's incredibly slow - ~60 seconds to compress a 512x512 image, on a Core i7 - but it works.
...
- Added AVPCL compressor to projects and got it building with VC9 and VC10.
- Removed unused command line interface & file read/write code from AVPCL.
- Convert AVPCL to use NV vector math lib, asserts, etc.
- Convert AVPCL to use double instead of float.
- Added 4x4 symmetric eigensolver, for AVPCL; it's based on the existing 3x3 one, but I had to rewrite the Householder reduction stage. As with ZOH, using the eigensolver (instead of SVD) gives a ~25% speedup without significantly affecting RMSE.
- Encapsulate ZOH and AVPCL stuff into their own namespaces to keep everything separate.
- Added some missing vector operators to the nvmath lib.
11 years ago
nathaniel.reed@gmail.com
f2fa0517b5
Ported over James Arvo's SVD code from ZOH. It's a good deal slower than the eigensolver, and no better in RMSE - so didn't use it, but left the code in place for future reference.
...
Also replaced doubles with floats in the eigensolver code. Speeds up BC6 another 5%. No change to RMSE in test suite.
11 years ago
nathaniel.reed@gmail.com
474239c784
Add BC6 support to nvtt lib and utils.
...
- Use 3x3 eigensolver for initial fit in ZOH. Slightly better perf and RMSE than power method.
- Remove use of double precision in ZOH - speeds up by 12%.
- Fixed RGBM encoding that was broken for HDR images.
- Use gamma-2.0 space for RGBM for HDR images (improves precision in darks).
- Use UNORM instead of TYPELESS formats when saving a DX10 .dds file. The TYPELESS formats break most viewers.
- Cleaned up warnings in ZOH code.
- Command-line utils will warn if you give them an unrecognized parameter.
- Added VS2010 profiling results.
11 years ago
nathaniel.reed@gmail.com
77188a42ac
Fix VC10 project files. Fix code to build with CUDA 5.5 (the cudaMemcpyToSymbol overload that takes the symbol name as a string was removed).
11 years ago
castano@gmail.com
57b99a1595
Apply CubeSurface fold/unfold patch. Fixes issue 199.
11 years ago
castano@gmail.com
b2b367ef5f
Fix xyzToCieLab bug and add CIE Lab DeltaE 1994 color differencing.
...
Fixes issue 197.
Fixes issue 198.
11 years ago
castano@gmail.com
022944e49b
Apply patch to fix issue 192.
11 years ago
castano@gmail.com
27024253ce
Handle 3d textures in nvcompress.
11 years ago
castano
22bc0b949d
Fix assert. Fixes issue 190.
11 years ago
castano
ac0759efc8
Update cmake file. Fixes issue 189.
11 years ago
castano@gmail.com
94d0635285
Merge changes from the Witness.
11 years ago