kfrlib/kfr

Health v2 · maintenance only

Adoption not part of the score

1917 stars · 281 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

KFR is a fast, modern C++ digital signal processing framework providing FFT/DFT, FIR/IIR/biquad filter design and processing, sample rate conversion, convolution, and audio file I/O. It uses a SIMD abstraction layer (SSE, AVX, AVX-512, ARM NEON, RISC-V RVV) for high performance and is available under GPLv2+ or a commercial license.

Use cases

• compute FFT of audio or scientific signals in C++
• design and apply biquad, FIR, and IIR filters
• resample audio with high-quality sample rate conversion
• read and write WAV, FLAC, AIFF, MP3 audio files
• build audio plugins with fast SIMD-optimized DSP
• process gravitational-wave or sensor data with fast DFT
• implement convolution reverb in an audio application

When to choose

• you need high-performance FFT or filtering in C++ with SIMD acceleration
• you target multiple architectures (x86, ARM, RISC-V) from one codebase
• you want a header-friendly DSP library with no heavy runtime dependencies
• you need audio file encoding/decoding plus DSP in one library

When to avoid

• your project is not GPL-compatible and you cannot purchase a commercial license
• you need a DSP library in a language other than C++
• you only need simple one-off FFTs where a lighter library suffices
• you target RISC-V CPUs with VLEN < 128 or non-Linux RISC-V systems

Facets

library · maturity stable

audio-processing math image-processing sdk developer-tools cross-platform windows cpp wasm dsp fft simd sample-rate-conversion filters header-only audio-codecs avx neon risc-v audio signal-processing scientific-computing linux macos ios android

Member repositories

For agents

markdown · JSON · MCP: product_card(name="kfrlib/kfr")