zooc.dsp.dither

Dithering algorithms.

zooc.dsp.dither.dither_pattern(count, dither_levels)

Generate an unbiased dither pattern list.

The pattern is tiled and then adjusted so that its mean value is close to (levels - 1) / 2.

Parameters:

• count (int) – The total number of dither values to generate.

• dither_levels (int) – The number of discrete dither levels (e.g., 3 for [0, 1, 2]).

Returns:

A list of dither levels.

Return type:

list[int]

zooc.dsp.dither.dither_weights(count, dither_levels)

Generate an unbiased dither weights.

The weights are tiled and then adjusted so that its mean value is close to zero. The noise injected to measurement signal should have an amplitude of weight*LSB.

Not all count & dither_levels combinations produces zero mean. For zero mean, use:

- When dither_levels is odd: count >= dither_levels - 2
- When dither_levels is even: count = even and >= dither_levels - 2

Parameters:

• count (int) – The total number of dither values to generate (measurements).

• dither_levels (int) – The number of discrete dither levels.

Returns:

A list of dither weights.

Return type:

list[float]