When using GAUSSIAN_NORMALS, image::resampleImage automatically unitizes
its output. This means that clients (e.g. mipgen) need to ensure that
the source image is in [-1,+1] rather than [0,+1].
Due to an assertion in etc2comp, this unfortunately uses RGB8 instead of
RG11 but it still helps quite a bit (3.3 MB savings in the unzipped KTX
size).
This shrinks file size as follows, although it seems to darken the
rendered result.
```
1398209 Oct 2 16:22 ao.ktx
699172 Oct 2 16:20 ao_etc.ktx
1398209 Oct 2 16:22 metallic.ktx
699172 Oct 2 16:19 metallic_etc.ktx
1398209 Oct 2 16:22 roughness.ktx
699172 Oct 2 16:20 roughness_etc.ktx
```
At build time, mipgen now creates several compressed and non-compressed
variants for albedo.
For ASTC we use the ARM encoder, for S3TC we use the STB encoder. The
latter is somewhat limited so we may wish to investigate other libraries
in the future (e.g., AMD Compressenator).
At run time, we detect which of these variants to download, based on
available WebGL extensions. In practice, this means that desktop
browsers will use the S3TC variant and mobile browsers will use the ASTC
variant.
The makes unzipped albedo go from ~4 MB to 682 KB using S3TC.
Mobile support (ASTC and ETC2) is coming in a future PR, stay tuned.
The compression API now takes block size instead of bit rate. Otherwise
it is quite possible for the encoder to generate a block size that is
not one of the valid block sizes that graphics API's can actually
consume.
this adds a new --type (-t) parameter that is
used to specify cmgen’s output.
faces: outputs cube maps separate faces (default
and current behavior)
equirect: outputs an equirectangular image
ktx: outputs a ktx file
--format=ktx now implies —type=ktx
This was designed as a standalone tool rather than a library, so we are
renaming their "main" function to "standalone_main", which is described
in the tnt README.
We'll also be adding a wrapper to our imageio library.
- internal cross storage is now always horizontal
and seamless.
- source cross images are now always converted to
internal format, conversion includes rescaling
and sampling if scaling down.
- vertical cross is supported again
- fixes#283
- replace hardcoded constants
This tool was producing bad output for the monkey roughness map because
the monkey map has an (unnecessary) alpha channel, and the tool was
assuming that the source image has 3 channels.
This fixes that particular case and changes the tool so that it
operates on single-channel data instead of 3-channel data.
- front and back labels where inverted:
+z is the front of the cube and -z is its back
when looking at the cube from the outside
with a camera pointing towards -z (and the cube
in front of it). This is the GL definition.
- make the synthetic uvXXX debug environment
colors match those of the debug environment
(except for -Z/back which is magenta instead
of black).
- add some documentation about skybox and cube map.
- cmgen now creates reflections maps for all input
assets by default instead of just for equirectangular
images.
- equirectangular decode doesn't mirror -- that's
done by the mirroring code now.
- mirroring can be turned of with --no-mirror
- option --mirror is gone
- faster mirroring code (do it before generating
mipmaps)
- added labels to faces of debug.png environment
- added a debug equirectangular image in third_party
(creative commons license)
- KTX / OpenGL are PX,NX,PY,NZ,PZ,NZ.
- Cubemap::Face is NX,PX,NY,PY,NZ,PZ.
I considered changing the Cubemap::Face enum, but applying ^1 upon
export seems less intrusive.
For each envmap/skybox pair, this will reduce the number of downloads
from 61 files to 3 files.
Note that we are still using sh.txt for coefficients, I'll see if it
would be easy to embed that in the KTX metadata.
Also, I will add KTX support to roughness-prefilter soon.
* Ability to add code prior to prepareMaterial
* Add normal input on shading node
* Add constant float node
* Add multiply and subtract nodes
* Split math nodes out into MathNodes.kt
(1) Suzanne's normal map is actually linear but we were telling GL that
it is SRGB, so the shader was not reading (0.5,0.5,1.0) from a perfectly
flat normal map.
(2) --linear in mipgen was not honored when consuming PNG files.
(3) The build was not passing --linear to mipgen for the normal map.
The libimage API does not include the original image in the returned mip
array, which might make sense when generating individual files, but for
a KTX bundle, we want to include level 0.
I tested the validity of the resulting KTX files by using macOS Preview,
which can read some types of KTX files. (!)
This CL also adds mipgen to the WebGL build, although its outputs are
not yet used by the sample apps.
This remapping can be confused with roughness, because
linear_roughness = roughness*roughness and
lod = sqrt(linear_roughness)
so the value of lod is roughness, but conceptually
it’s not.
Sometimes the source image might have a useless 1-bit alpha channel
which our image-processing pipeline expands into a proper alpha channel.
This consumes memory and breaks an assertion in our encoder.
I ran into this while testing mipgen with the Suzanne textures.
