Struct sonogram::SpecOptionsBuilder

pub struct SpecOptionsBuilder { /* private fields */ }

A builder struct that will output a spectrogram creator when complete. This builder will require the height and width of the final spectrogram, at a minimum. However you can load data from a .wav file, or directly from a Vec memory object.

Example

  let mut spectrograph = SpecOptionsBuilder::new(512, 128)
    .set_window_fn(utility::blackman_harris)
    .load_data_from_file(&std::path::Path::new("test.wav"))?
    .build();

Implementations

impl SpecOptionsBuilder

pub fn new(num_bins: usize) -> Self

Create a new SpecOptionsBuilder. The final height and width of the spectrogram must be supplied. Before the build function can be called a load_data_from_* function needs to be called.

Arguments

• num_bins - Number of bins in the discrete fourier transform (FFT)

pub fn load_data_from_file(self, fname: &Path) -> Result<Self, SonogramError>

Load a .wav file to memory and use that file as the input.

Arguments

• fname - The path to the file.

pub fn load_data_from_memory(self, data: Vec<i16>, sample_rate: u32) -> Self

Load data directly from memory - i16 version.

Arguments

• data - The raw wavform data that will be converted to a spectrogram.
• sample_rate - The sample rate, in Hz, of the data.

pub fn load_data_from_memory_f32(self, data: Vec<f32>, sample_rate: u32) -> Self

Load data directly from memory - f32 version.

Arguments

• data - The raw wavform data that will be converted to a spectrogram. Samples must be in the range -1.0 to 1.0.
• sample_rate - The sample rate, in Hz, of the data.

pub fn downsample(self, divisor: usize) -> Self

Down sample the data by the given divisor. This is a cheap way of improving the performance of the FFT.

Arguments

• divisor - How much to reduce the data by.

pub fn channel(self, channel: u16) -> Self

Set the audio channel to use when importing a WAV file. By default this is 1.

pub fn normalise(self) -> Self

Normalise all the sample values to range from -1.0 to 1.0.

pub fn scale(self, scale_factor: f32) -> Self

Scale the sample data by the given amount.

pub fn set_window_fn(self, window_fn: fn(_: usize, _: usize) -> f32) -> Self

A window function describes the type of window to use during the DFT (discrete fourier transform). See (here)[https://en.wikipedia.org/wiki/Window_function] for more details.

Arguments

• window - The window function to be used.

pub fn set_step_size(self, step_size: usize) -> Self

This is the step size (as the number of samples) between each application of the window function. A smaller step size may increase the smoothness of the sample, but take more time. The default step size, if not set, is the same as the number of FFT bins. This there is no overlap between windows and it most cases will suit your needs.

pub fn build(self) -> Result<SpecCompute, SonogramError>

The final method to be called. This will create an instance of [Spectrograph].