This release mostly fixes (security related) bugs, including
* 12 decoder bugs found by oss-fuzz
* encoder bug CVE-2021-0561
Bump revision to stay ahead of -stable.
Enable debug package while here.
with namn@
Spotted through this warning on startup:
"GlobalShortcutX: No XInput support, falling back to polled input. This
wastes a lot of CPU resources, so please enable one of the other methods."
https://www.mumble.info/blog/mumble-1.4.230/
Lots of cool changes!
qmake was replaced with cmake and upstream already merged a few handful fixes
I sent during 1.4.0rc1 tests (they should land in the next 1.4.x release).
https://github.com/FNA-XNA/FNA/releases/tag/22.01https://github.com/FNA-XNA/FAudio/releases/tag/22.01https://github.com/FNA-XNA/FNA3D/releases/tag/22.01
New Features:
Update to FNA3D 22.01
Update to FAudio 22.01
Added SetVideoTrackEXT extension
Added SrgbEXT extension
Fixes:
The FNA3D log is now hooked later in FNAPlatform initialization, to try and load SDL before FNA3D
Partially works around a regression in the new dyld in macOS Monterey
Fixed DDSFromStreamEXT mip loading to calculate each mip level in full
New Features:
Added support for SRGB textures/renderbuffers
Removed Features:
The Metal renderer has been removed, in favor of Vulkan/MoltenVK
Fixes:
Vulkan: Various fixes for device support checks
Image: Log errors from stb_image
GZDoom's music system as a standalone library for easier reuse by
other projects.
port from Timo Myyrä with some small tweaks by me. Thanks Timo!
OK solene@
Ports still pick up rnnoise just fine, but we can do better.
Remove NO_TEST=Yes while here: tests are empty and don't fail, but this
might change in the future.
NO_TEST reminder from sthen
It is a dependency of an upcoming port, but it will also be picked up by
audio/mumble to enable optional noise suppression there as well.
Feedback sdk tb phessler
OK sdk
---
RNNoise is a noise suppression library based on a recurrent neural network.
A description of the algorithm is provided in the following paper:
J.-M. Valin, A Hybrid DSP/Deep Learning Approach to Real-Time Full-Band Speech
Enhancement, Proceedings of IEEE Multimedia Signal Processing (MMSP) Workshop,
arXiv:1709.08243, 2018.
https://arxiv.org/pdf/1709.08243
While it is meant to be used as a library, the code for a simple command-line
tool is provided as an example. It operates on RAW 16-bit (machine endian) mono
PCM files sampled at 48 kHz.
