Plot renders of sRGB sweep in chromaticity diagram

analyze.py

import colour.models
import colour.plotting
import pathlib
from matplotlib import pyplot
from PIL import Image


def plot_chromaticity_image(filename):
    image = Image.open(filename)
    traces = extract_traces(image)

    fig, ax = colour.plotting.diagrams.plot_chromaticity_diagram(show_diagram_colours=False, show=False)
    srgb = [[1.0, 0.0, 0.0], [0.0, 1.0, 0.0], [0.0, 0.0, 1.0], [1.0, 0.0, 0.0]]
    xy = colour.XYZ_to_xy(colour.sRGB_to_XYZ(srgb))
    ax.scatter(xy[:, 0], xy[:, 1], marker='o', c=srgb, label='sRGB')

    for trace in traces:
        trace = [t for t in trace if sum(t) > 0] # skip [0, 0, 0]
        xy = colour.XYZ_to_xy(colour.sRGB_to_XYZ(trace))
        ax.plot(xy[:, 0], xy[:, 1], '-', color='lightgray', alpha=0.3)
        ax.scatter(xy[:, 0], xy[:, 1], marker='o', s=5, c=trace)

    ax.set(xlim=(-0.1, 0.8), ylim=(-0.1, 0.9), aspect='equal')
    ax.set(title=filename.stem)
    return fig, ax


def plot_oklch_image(filename):
    image = Image.open(filename)
    traces = extract_traces(image)

    fig, ax = pyplot.subplots()

    srgb = [[1.0, 0.0, 0.0], [0.0, 1.0, 0.0], [0.0, 0.0, 1.0], [1.0, 0.0, 0.0],
        [1.0, 1.0, 0.0], [0.0, 1.0, 1.0], [1.0, 0.0, 1.0]]
    oklch = colour.Oklab_to_Oklch(colour.XYZ_to_Oklab(colour.sRGB_to_XYZ(srgb)))
    ax.scatter([1.02]*len(oklch), oklch[:, 2], marker='o', c=srgb, label='sRGB')

    for trace in traces:
        trace = [t for t in trace if sum(t) > 0] # skip [0, 0, 0]
        oklch = colour.Oklab_to_Oklch(colour.XYZ_to_Oklab(colour.sRGB_to_XYZ(trace)))
        ax.plot(oklch[:, 0], oklch[:, 2], '-', color='lightgray', alpha=0.3)
        ax.scatter(oklch[:, 0], oklch[:, 2], marker='o', s=5, c=trace)

    ax.set(xlim=(0.0, 1.04), xlabel='Oklch L (luminance)',
        ylim=(360.0, 0.0), ylabel='Oklch h (hue angle)')
    ax.set(title=filename.stem)
    return fig, ax


def extract_traces(image):
    tl = 147, 111
    br = 1780, 916
    num_boxes = 51, 19
    box_size = (br[0] - tl[0]) / num_boxes[0], (br[1] - tl[1]) / num_boxes[1]
    lines = []
    for box_y in range(num_boxes[1]):
        line = []
        for box_x in range(num_boxes[0]):
            px_x = tl[0] + box_x * box_size[0] + box_size[0] / 2
            px_y = tl[1] + box_y * box_size[1] + box_size[1] / 2
            color = image.getpixel((int(px_x), int(px_y)))
            color = tuple(c / 255 for c in color)
            line.append(color)
        lines.append(line)
    return lines


if __name__ == "__main__":
    for filename in pathlib.Path('.').glob('*.png'):
        fig, ax = plot_chromaticity_image(filename)
        fig.savefig(filename.stem + '_chromaticity.svg')
        pyplot.close(fig)
        fig, ax = plot_oklch_image(filename)
        fig.savefig(filename.stem + '_oklch.svg')
        pyplot.close(fig)

Feel free to use this script any way you please. I place it in the public domain.