Reads raw audio data from any common audio or video format. Use read_audio_bin to get raw PCM audio samples, or read_audio_fft to stream-convert directly into frequency domain (spectrum) data using FFmpeg built-in FFT.
read_audio_fft( audio, window = hanning(1024), overlap = 0.75, sample_rate = NULL, start_time = NULL, end_time = NULL ) read_audio_bin( audio, channels = NULL, sample_rate = NULL, start_time = NULL, end_time = NULL )
path to the input sound or video file containing the audio stream
vector with weights defining the moving fft window function. The length of this vector is the size of the window and hence determines the output frequency range.
value between 0 and 1 of overlap proportion between moving fft windows
downsample audio to reduce FFT output size. Default keeps sample rate from the input file.
position (in seconds) to cut input stream to be processed.
number of output channels, set to 1 to convert to mono sound
Currently read_audio_fft automatically converts input audio to mono channel such
that we get a single matrix. Use the
plot() method on data returned by read_audio_fft
to show the spectrogram. The av_spectrogram_video generates a video that plays
the audio while showing an animated spectrogram with moving status bar, which is
# Use a 5 sec fragment wonderland <- system.file('samples/Synapsis-Wonderland.mp3', package='av') # Read initial 5 sec as as frequency spectrum fft_data <- read_audio_fft(wonderland, end_time = 5.0) dim(fft_data)#>  512 860# Plot the spectrogram plot(fft_data)#>  1024 430#>  2048 215