diff options
Diffstat (limited to 'src/oklab.h')
| -rw-r--r-- | src/oklab.h | 150 |
1 files changed, 150 insertions, 0 deletions
diff --git a/src/oklab.h b/src/oklab.h new file mode 100644 index 0000000..65c0bd1 --- /dev/null +++ b/src/oklab.h @@ -0,0 +1,150 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/** @file Support for the OKLab/OKLch perceptual color space. + */ +/* + * Authors: + * Rafał Siejakowski <rs@rs-math.net> + * + * Copyright (C) 2022 Authors + * + * Released under GNU GPL v2+, read the file 'COPYING' for more information. + */ + +#ifndef SEEN_OKLAB_H +#define SEEN_OKLAB_H + +#include <array> + +#include "hsluv.h" + +namespace Oklab { + +using Triplet = Hsluv::Triplet; +using Hsluv::to_linear, Hsluv::from_linear; + +/** @brief Convert an OKLab color to the linear RGB color space. + * + * @param oklab_color A triplet containing the L, a, b components. + * @return a triplet containing linear (de-gamma'd) RGB components, all in [0, 1]. + */ +Triplet oklab_to_linear_rgb(Triplet const &oklab_color); + +/** @brief Convert a linear RGB color to OKLab coordinates. + * + * @param linear_rgb_color A triplet containing degamma'd (linearized) R, G, B components. + * @return a triplet containing the (L, a, b) coordinates of the color in the OKLab space. + */ +Triplet linear_rgb_to_oklab(Triplet const &linear_rgb_color); + +/** Convert an OKLab color to a gamma-compressed sRGB color. */ +inline Triplet oklab_to_rgb(Triplet const &oklab_color) +{ + auto rgb = oklab_to_linear_rgb(oklab_color); + for (auto &component : rgb) { + component = from_linear(component); + } + return rgb; +} + +/** Convert a gamma-compressed sRGB color to an OKLab color. */ +inline Triplet rgb_to_oklab(Triplet const &rgb_color) +{ + Triplet linear; + for (size_t i = 0; i < 3; i++) { + linear[i] = to_linear(rgb_color[i]); + } + return linear_rgb_to_oklab(linear); +} + +/** @brief Convert an OKLab color to the OKLch coordinates. + * + * The OKLch coordinates are more closely aligned with the perceptual properties of a color + * and therefore are more convenient for the end user. They consist of: + * L – luminance of the color, in the interval [0, 1] (this is the same as the L in (L, a, b)). + * c – chroma; how far the color is from grayscale. The range of c-values is [0, cmax] but + * cmax depends on L and h; @see Oklab::max_chroma(). + * h – hue. A number in the interval [0, 360), interpreted as a hue angle on the color wheel. + * + * @param ok_lab_color A color in the OKLab color space. + * @return The same color expressed in the Lch coordinates. + */ +Triplet oklab_to_oklch(Triplet const &ok_lab_color); + +/** @brief Convert an OKLch color to the OKLab coordinates. + * + * For the meaning of the Lch color coordinates, @see oklab_to_oklch(). + */ +Triplet oklch_to_oklab(Triplet const &ok_lch_color); +Triplet oklch_radians_to_oklab(Triplet const &ok_lch_with_hue_in_radians); + +/** @brief Convert an OKLab color to an OKHSL representation. + * + * As of late 2022, OK-HSL (hue, saturation, lightness) is not a fully standardized color + * space. The version used here stores colors as triples (h, s, L) of doubles, all in + * the interval [0, 1]. These coordinates have the following meaning: + * + * \li \c h is the hue angle, scaled to the interval [0, 1]. + * \li \c s is the linear saturation. For a given OKLch color (L, c, h), linear saturation + * is the ratio of c to the maximum possible chroma c_max such that (L, c_max, h) + * fits in the sRGB gamut. Therefore, s=1 always corresponds to a maximally saturated + * color and s=0 is always a grayscale color. + * \li \c L is the lightness; it is the same coordinate as the L in OKLab or OKLch. + * + * Note that Björn Ottosson proposes a new, non-standard way of compressing the saturation + * (similar to gamma compression) which results in a different saturation scale. He further + * suggests varying this compression depending on the hue, although he does not standardize + * this proposed OKHSL space or the saturation transfer functions. Instead, he uses a somewhat + * ad-hoc piecewise-linear saturation transfer function in his own picker, which unfortunately + * breaks differentiability of the parametrization, a great advantage of the OKLch color space. + * If an OKHSL space is ever standardized, the behaviour of this function may change. + * See https://bottosson.github.io/posts/colorpicker/ for more details. + * + * @param ok_lab_color A color in the OKLab color space. + * @return The same color expressed in the (h, s, L) coordinates. + */ +Triplet oklab_to_okhsl(Triplet const &ok_lab_color); + +/** @brief Convert an OKHSL color to the OKLab coordinates. + * + * For a description of the OKHSL color scale used here, see oklab_to_okhsl(). + * If an OKHSL space is ever standardized, the behaviour of this function may change. + * + * @param ok_hsl_color A color in the OKHSL coordinates, all normalized to [0, 1]. + * @return The same color expressed in the (L, a, b) coordinates. + */ +Triplet okhsl_to_oklab(Triplet const &ok_hsl_color); + +/** @brief Find the maximum OKLch chroma for the given luminosity and hue. + * + * @param l OKLab/OKLch luminosity, in the interval [0, 1]. + * @param h OKLch hue angle in degrees (interval [0, 360]). + * @return The maximum chroma c such that the color oklch(l, c, h) fits in the sRGB gamut. + */ +double max_chroma(double l, double h); + +/** @brief Helper functions for rendering color strips used in color sliders. + * + * @param h Hue angle in degrees (range 0-360). + * @param s Saturation (percentage of allowed chroma) in the range 0-1. + * @param l OKLab/OKLch lightness in the range 0-1. + * @param[out] map Working store and RGB output showing the range of one of the color coordinates. + * @return A pointer to the raw data of the map. + */ +uint8_t const *render_hue_scale(double s, double l, std::array<uint8_t, 4 * 1024> *map); +uint8_t const *render_saturation_scale(double h, double l, std::array<uint8_t, 4 * 1024> *map); +uint8_t const *render_lightness_scale(double h, double s, std::array<uint8_t, 4 * 1024> *map); + +} // namespace Oklab + +#endif // SEEN_OKLAB_H + +/* + Local Variables: + mode:c++ + c-file-style:"stroustrup" + c-file-offsets:((innamespace . 0)(inline-open . 0)(case-label . +)) + indent-tabs-mode:nil + fill-column:99 + End: +*/ +// vim: filetype=cpp:expandtab:shiftwidth=4:tabstop=8:softtabstop=4:fileencoding=utf-8:textwidth=99 :
\ No newline at end of file |