The Open Source KFR framework has been updated to version 3.0.
Brief summary of the major improvements in this release:
FFT performance has been improved by up to 15%. Also, optimized non power of 2 FFT is also supported, including FFT of prime sizes.
Visual Studio and GCC are now fully supported and tested in addition to Xcode and clang.
- AVX512 support has been added for latest Intel CPUs.
- Ability to compute loudness according to EBU R128 standard.
- Audio file reading/writing support has been improved and now includes FLAC file reading, as well as more portable WAV file reading/writing and sample format conversion.
- Other performance and stability improvements.
KFR framework provides highly optimized algorithms for audio and DSP, including FFT, FIR, IIR, Biquad filters, audio file reading/writing, high quality sample rate conversion, oscillators. All algorithms and functions are SIMD-optimized for x86 and ARM.