hann-rs (Hann Window)

This crate provides functions for computing the Hann window, a function used to taper the edges of a signal window to reduce spectral leakage. The Hann window is defined by the formula w(n) = 0.5 - 0.5 * cos(2π * n / (N - 1)), where n is the index of the current sample and N is the length of the window.

A lookup table for pre-computed Hann windows is provided for common window lengths. This improves performance for repeated calculations with the same window length.

Usage

1. Import the crate:

[dependencies]
hann-rs = "0.1.0"

use hann_rs::get_hann_window;

2. Get a Hann window of a specific length using the get_hann_window function:

let window_length = 1024;
let hann_window = get_hann_window(window_length).expect("Failed to get the Hann window");

If the desired window length is in the lookup table, the precomputed values will be returned. If not, the Hann window values will be computed.

Precomputed Lookup Table

The lookup table, HANN_WINDOW_LOOKUP_TABLE, contains precomputed Hann windows of lengths 256, 512, 1024, 2048, and 4096.

You can add or modify the precomputed window lengths by changing the HANN_WINDOW_PRECOMPUTED_LENGTHS array in the lazy_static block:

const HANN_WINDOW_PRECOMPUTED_LENGTHS: [usize; 5] = [256, 512, 1024, 2048, 4096];

Error Handling

The functions get_hann_window return a Result<Vec<f32>, HannWindowError> type. Errors are returned in the following cases:

• The window length is less than or equal to 1.
• The window length is greater than the allowed maximum.
• Memory allocation errors due to a large window length.

To handle these errors, use the Result type as follows:

match get_hann_window(window_length) {
    Ok(hann_window) => {
        // Use the Hann window
    },
    Err(error) => {
        println!("Error getting the Hann window: {:?}", error);
        // or handle the error in your way
    },
}

Benchmarks of APPROXIMATE results

get_hann_window | 2000 WL | 7.0252 (µs) | 7.0657 (µs) | 7.1153 (µs) | get_hann_window | 4000 WL | 13.496 (µs) | 13.596 (µs)| 13.708 (µs) | get_hann_window (Cached) | 4096 WL | 363.84 (ns) | 369.98 (ns) | 377.30 (ns) |

hann-rs (Hann Window Sum of Squares)

This module provides functions for computing the sum of squares of a Hann window. It utilizes a lookup table for pre-computed sum of squares for common Hann window lengths, improving performance for repeated calculations with the same window length.

Usage

1. Import the module:

use hann_rs::get_hann_window_sum_squares;

2. Get the sum of squares of a Hann window using the get_hann_window_sum_squares function:

let hann_window: Vec<f32> = vec![...]; // The Hann window values
let hann_window_sum_squares = get_hann_window_sum_squares(&hann_window);

If the length of the input hann_window is in the lookup table, the precomputed value will be returned. If not, the sum of squares will be computed using map and sum.

Precomputed Lookup Table

The lookup table, HANN_WINDOW_SUM_OF_SQUARES, contains precomputed sum of squares for Hann windows of lengths 256, 512, 1024, 2048, and 4096.

You can add or modify the precomputed window lengths by changing the HANN_WINDOW_PRECOMPUTED_LENGTHS array in the lazy_static block:

const HANN_WINDOW_PRECOMPUTED_LENGTHS: [usize; 5] = [256, 512, 1024, 2048, 4096];

Benchmarks of APPROXIMATE results

get_hann_window_sum_squares | 2000 WL | 601.14 (ns) | 603.21 (ns) | 605.71 (ns) | get_hann_window_sum_squares | 4000 WL | 1.1465 (µs)| 1.1520 (µs) | 1.1588 (µs) | get_hann_window_sum_squares (Cached) | 4096 WL | 10.583 (ns) | 10.628 (ns) | 10.680 (ns) |