EqualizerAPO 1.4.2 - Double Precision Edition - CI Test
Pre-release
Test of a new automated CI pipeline for building EqualizerAPO, currently supporting 1.4.2 on 64-bit. Now includes all components (notably the Editor and other executables), all rebuilt with a newer FFTW (3.3.10 up from 3.3.5), and support for an automated installer build. Should allow easier development of fixes and improve support for contributions.
Note there may still be issues with convolution/mono in specific setups, hopefully fixed in the next build
Features
- Double precision processing filter engine (64bit audio processing) for improved fidelity with more complex setups
- Additional usage of AVX instructions to reduce processing latency
- Removes the need for static VC++ dlls in the install folder
- Compiled with VS 2022 for all components (minus libsndfile), using the latest Windows SDK, AVX2 and various optimizations.
- Built with AOCL-FFTW 3.3.10 (optimized for AMD/modern CPUs)
- Updated Qt 6.9 for GUI components
- Uses fp-precise instead of fp-fast for floating point math
- Automated build pipeline for all components and the installer via Github Actions
Requires at minimum a CPU with AVX2 support, anything from 2013 (Haswell) onwards for Intel, and anything from 2015 onwards (Excavator/Ryzen) for AMD. AVX-512 support will be readded in a future release to the new pipeline.
Installation
Install via the included setup file.
Alternatively download via Github Actions (https://github.com/TheFireKahuna/equalizerAPO64/actions/runs/19897955139), requires a signed in Github account.
If you have a missing DLL error, make sure you have installed the Microsoft Visual C++ 2022 Redistributable from their website here.