- Multichannel images are treat as CMYK image when the number of channels are more than 1: when exists, channel 5 is used as alpha. Channels higher than 5 are discarded.
- Multichannel images are treat as Grayscaleimage when the number of channels are equals to 1.
- Device transactions removed (where possible)
- Fix clang-format issues
When writing RGBA16FPx16_Premultiplied format, the image was converted to RGBA64 by clamping the float values. With this patch float values outside the range [0, 1] are preserved.
The following changes have been made:
- Improved writing speed by using scanLine() instead of pixel()
- Optimized memory usage on writing by using ScanlineConverter class
- Added native write support for RGBA32FPx4, RGBA16FPx4, RGB8, Grayscale8 and Indexed8 uncompressed formats
- Grayscale DDS without alpha are loaded in a Grayscale8 image
- Fixed warnings about wrong PITCH reported by GIMP on DDSs saved by this plugin
- Initial support for loading DX10 formats (R16F, RG16F, RGBA16F, RGBPreMulA16F, R32F, RG32F, RGBA32F, RGBPreMulA32F)
- Fixed alignment issues and A8P8 format support of the attached images*
Tested using GIMP and [NVIDIA Texture Tools](https://developer.nvidia.com/texture-tools-exporter) plugin for Photoshop.
(*) The following images (taken from [here](https://github.com/walbourn/directxtexmedia)) cannot be added to read tests due to license issue:
[test8_DWORD.dds](/uploads/449b5a0d886aaf6764af554fe38e2b09/test8_DWORD.dds)
[dx5_logo.dds](/uploads/6f5f27df752890b227ef07e0195435d4/dx5_logo.dds)
[test888_DWORD.dds](/uploads/c8bc355c5749cf203d47e0b3073ad419/test888_DWORD.dds)
- Use of native qfloat16 format
- FP16 and FP32 HDR images are preserved and sRGB Linear profile is set
- Images are always saved as 8-bit so, if valid profile is present the image is converted to sRGB
- It also fixes (left image) a problem (right image) with HDR images:
{width=281 height=157}{width=271 height=169}
I tested the plugin with the images stored [here](https://github.com/walbourn/directxtexmedia). It contains PFM and PHM images but I didn't find any license on the repo, so the images was not added to test cases.
I attach here the images that this patch allows to load:
[grad4d_mono.pfm](/uploads/6e41fcb64d0651a6abd78cffc5ff86b2/grad4d_mono.pfm)
[grad4d_mono.phm](/uploads/348a51476068aa344f67826006cb65c4/grad4d_mono.phm)
[grad4d.phm](/uploads/2ab8330466f4dd0fc2fda00711270ce9/grad4d.phm)
The issue was identified by OSS Fuzz and the feature was not covered by our tests.
- Added earth-cubemap.dds under MIT licenses taken from [Open Toolkit library](https://github.com/mono/opentk/tree/main/Source/Examples/Data/Textures)
- Fix a wrong image size returned by a cubemap image
- Read test skips .license files
Workaround for issue https://github.com/libjxl/libjxl/issues/3983
libjxl_cms is used for color conversion of lossy static images.
Fix loading of lossy grayscale images
Fix loading of grayscale images with odd width
Closes#11
Requires MR !279
On formats that does not support CMYK and does not use the ScanLineConverter class during write operation, the CMYK images must be converted using the color space conversion functions of `QImage` (if ICC profile is valid).
When loading lossless JXL images without ICC profile, the parser failed.
The problem was caused by Boxes being enabled (MR !250). The parser should probably be revised to be more flexible.
CCBUG: 496350
The Portable HalfMap is a format supported by ImageMagick. The test cases was generated by converting pfm to phm using ImageMagick: `convert image.pfm image.phm`.
Added support for:
- (R/O) 1 bit, 3 planes (ICONDOC.pcx) -> test case from internet archive
- (R/O) 2 bits, 1 planes (CGA_FSD.pcx) -> test case from internet archive
- (R/W) 8 bits, 4 planes (dice_rgba.pcx) -> supported by ImageMagick and GIMP 3
Fix write support for:
- Grayscale 8/16-bits (saved as indexed)
- Alpha 8 (saved as indexed)
- RGB 16-bits depth (saved as RGB 24/32)
I keep adding old formats to ensure interoperability with as many programs as possible...
This plugin adds read-only support for the Scitex SCT format only. This format is also supported by Photoshop in read and write (SCT test cases were created by Photoshop).
[Scitex HandShake Formats Specifications](/uploads/a3e213e48349d898b260375d6c052521/Scitex_HandShake_Formats.pdf)
Highlights of the patch:
- Supersede MR !249
- Added FP16 and FP32 images support thus preserving HDR values (read / write, required libjxl 0.9+).
- Added Gray8 and Gray16 support (read / write).
- Indexed images are saved as Gray8 when palette is gray scale.
- Binary images are saved as Gray8 (does JXL natively support binary images?).
- Simplified writing process by partially removing the use of additional buffers.
- Added XMP metadata support by decoding/encoding Boxes.
- Changed maximum image size in pixels in accordance with JXL feature level 5 (still limited to 256 megapixels).
Compatibility:
- Older versions of this plugin load FP images correctly as UINT16 (obviously losing HDR info).
- HDR images saved with this patch are also loaded correctly by Gimp and Photoshop.
- Grayscale images saved with this patch are also loaded correctly by Gimp and Photoshop.
Compilation modifiers for cmake file:
- `JXL_HDR_PRESERVATION_DISABLED`: disable the FP support (behaves like previous versions).
- `JXL_DECODE_BOXES_DISABLED`: disable metadata reading (behaves like previous versions).
The following changes are done:
- Fixed an error when copying image from sequential device
- Return error when the image exceed 4GB size on writing*
- Enabled JXR tests
(*) Note that when writing an image larger than 4GiB, the JXRLib does not give any error but the resulting saved image id wrong.
Where possible, QIODevice::peek has been used instead of transactions or instead of using ungetchar() for sequential access devices and seek() for random access devices.
Furthermore:
- RAS format gained the ability of read on sequential devices.
- Removed unused code in XCF (still related to ungetchar and sequential devices).
- These changes should prevent errors like the ones fixed by MR !258
When reading from a sequential device, the peekHeader() method in
the PCX readers reads the header its defined little endian into
arch-specific endianness for multibyte types.
Being a "peek" method, it then it tries to push back the bytes into
the device after reading for its next use, but it doesn’t convert
multibyte types correctly from arch-specifice endianness to the
initial little endian format.
Subsequent reading of the data from the device will thus lead to
incorrect values for multibyte types on the next use.
This patch reuses the same technique as the TGA reader to read the
whole header as bytes before deserializing it, so that the bytes
can be pushed back into the sequential device in the same order.
- Added support for `Size` and `Format` options and slightly improved format detection from canRead().
- Removed conversion to ARGB32 on load (improved performace with RGBA images).
- Added result checks on writing.
With this MR, all plugins have minimal support for options.
- Added support for ```Size``` and ```Format``` options and slightly improved format detection from canRead().
- Added PCXHEADER::isValid() method to consolidate header consistency checks in one place.
Let Qt rotate the image when the ImageAutotransform option is set to true.
In tests it also solves the image size control with the value returned by the options with certain rotations.
EXR, HDR, JXR and PFM formats support High Dynamic Range images (FP values grater than 1).
In summary, here is the list of changes:
EXR, HDR, JXR and PFM: When working with FP formats, the clamp between 0 and 1 is no longer done.
EXR: Removed old SDR code and conversions. Due to the lack of a QImage Gray FP format, Gray images are output as RGB FP (recently added code for Qt 6.8 has been removed).
PFM: Due to the lack of a QImage Gray FP format, Gray images are output as RGB FP.
HDR: Added rotation and exposure support.
With this patch, EXR, JXR, HDR, PFM behave like Qt's TIFF plugin when working with FP images.
This fixes the following compiler warning with mingw-w64 toolchain in Windows.
raw.cpp:436:9: warning: "DEFAULT_QUALITY" redefined
436 | #define DEFAULT_QUALITY (C_IQ(3) | C_OC(1) | C_CW(1) | C_AW(1) | C_BT(1) | C_HS(0) | C_FLAGS(1))
| ^~~~~~~~~~~~~~~
wingdi.h:1142:9: note: this is the location of the previous definition
1142 | #define DEFAULT_QUALITY 0
| ^~~~~~~~~~~~~~~
DEFAULT_QUALITY macro is used with CreateFontW API.
https://learn.microsoft.com/en-us/windows/win32/api/wingdi/nf-wingdi-createfontw
