sound_split

sound spectrum analyzer

description

A linux cuda implementation of Fast recursive spectral analysis using variable window　(H.Nakatsuji).

screenshot

compile

prepare cuda environment

sudo apt-get install build-essential
wget https://developer.nvidia.com/compute/cuda/9.2/Prod2/local_installers/cuda_9.2.148_396.37_linux
sudo bash cuda_9.2.148_396.37_linux
wget https://developer.nvidia.com/compute/cuda/9.2/Prod2/patches/1/cuda_9.2.148.1_linux
sudo bash cuda_9.2.148.1_linux

and add this to .bashrc

export PATH=/usr/local/cuda/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

prepare libraries

sudo apt install libgtk-3-dev
sudo apt install libpulse-dev

clone source and compile

git clone https://github.com/lithium0003/sound_split.git
cd sound_split

edit cuda compute capability in Makefile

NVCCFLAGS := -O3 -I ~/NVIDIA_CUDA-9.2_Samples/common/inc/ -gencode=arch=compute_61,code=sm_61

and run make

make

how to use

$ ./sound_split [optinos] input.wav
options
  --mono: force load as mono
  --start [time]: skip first [time] samples in wav file
  --length [time]: analyze only [time] length
  --filter: filter mode
time option
  10.05 -> 10.05sec
  1min20sec or 1h10m5.02sec will be ok

License

These codes are licensed under CC0.

reference

可変窓を用いた高速再帰的スペクトル解析　中辻秀人　著