RecipesCam

3dtours/RecipesCam

Fork 0

Commit Graph

Author	SHA1	Message	Date
3dtours	6d60d452e0	One move of the light for the whole app: a mask's EXPOSURE and its four tone knobs stop being a second opinion A gradient mask had its own tone formula and its own exposure, and both disagreed with the frame's. Before any of this was tidied, the mask ran a smoothstep luma lift with an arbitrary 0.55..1.35 chroma clamp while the frame moved the knots of a four-zone ramp, and the mask's exposure was a stop on sRGB-encoded values while the frame's was a stop on light. Two names, four moves, and the same slider meant different things depending on whether the pixels were inside the shape you drew — the divergence §3.3 of the Android port's compat doc warns about. The maths is one string now (TONE_MATH_SKSL, interpolated by both passes): the ramp the four knots build, the hue-preserving rebuild behind it, the transfer pair, and exposureMove. A mask calls the same functions the frame calls. The rebuild carries the chroma instead of re-scaling it. Lightness takes the curve and the colour rides the difference — the channel differences move by ONE shared scale k, pulled back only where the cube has no room left. The doc's ratio (R_new = R_old * Luma_new / Luma_old) was the old reading and it is exact only while nothing clips: a channel past 1.0 stops being scaled with its neighbours and the hue goes with it. Measured on a flat patch frame, a skin tone at 24.0° came back at 48.0° at HIGHLIGHT +100, a warm white at 37° at 57.4°, and under L = 0.5 the same ratio multiplied a near-black pixel's cast by x30 — colour noise amplified, which is why the 0.55..1.35 clamp was there. The scale is the chroma's own now: over 135 knob combinations on seven colours and five greys, the ramp moves the luma and the hue does not move at all (Δ < 1e-9°). EXPOSURE gets the same treatment, which is what the second half of the request was: the linear domain decides where the luma is going, and the pixel is rebuilt onto it through the same lightMove. The old pass multiplied the three channels in linear light, so +1 EV clipped them by three different amounts: measured on the scratchpad probe, 29.2° of hue drift on a skin tone at +1 EV and 33.3° at +2, against 0.00° here. The stop is applied as a ratio on the pixel's own encoded luma rather than pointed straight at the encoded linear target, which is what makes the knob exactly the identity at 0 EV — the transfer does not commute with the luma weights, so pointing at it brightened a colour by a couple of code values even at zero. A grey is the knob it always was: 128 through +1 EV is 176, the same number the linear per-channel multiply put there, so nothing a user has dialled in moves. Verified: `npx tsc --noEmit` clean, `npm run build` clean. highlight-knee-check now runs EXPOSURE_SKSL for real — compiled with CanvasKit and four pixels pushed through it, agreeing with the twin to a code value on a grey at +1 EV (176), a skin tone at +1 EV and +2 EV, and a shadow at -2 EV; it also pins the hue, the cube, the identity at 0 EV and the black pixel that has no light to move. mask-wb-check compiles the mask pass and pushes the same stops through it: 128 through +1 EV is 176, through -1 EV is 92, +2 EV lands the channel on the ceiling at 255 and holds the hue within 3°. auto-tone-check, preview-match-check, white-level-check, raw-develop-check, half-check and roll-walk-check all pass. Live on the built bundle in a 1440x950 browser: the LIGHT panel's EXPOSURE +1 EV takes the mid grey of a flat patch frame from 0.502 to 0.690 (a stop on light gives 0.686) and moves no patch's hue at -1 EV (Δ 0.00°), and a linear gradient mask's EXPOSURE +1 EV and HIGHLIGHT +100 move the pixels inside the mask (luma 185.9 -> 211.1 and 185.9 -> 197.5) while the corner outside it does not move at all (220.2 -> 220.2), with no console error. Co-authored-by: PenguinHarness <noreply@penguin.local>	2026-09-29 18:04:50 +07:00
3dtours	3b92e4e2d2	A gradient mask can carry the LIGHT column's white balance now: COLOR TEMP and TINT, the same two rulers, read on the mask's own pixels. The pair was asked for as the two rows the develop column already has, so it is the same pair and not a second opinion about what a kelvin means: the gain comes from one function, colorUtils.whiteBalanceGain(temperature, tint), which the frame-wide colour matrix now calls too — the RGB kelvin fit of kelvinToRGB, the symmetric ±0.08 magenta/green tint, and the division by the product's own Rec.709 luma that keeps a cast from being a brightness move. A mask that never touched the pair reads 5500K / 0 (readMasks' own defaults, clamped to the ruler's ends), whose gain is exactly (1,1,1), so nothing moves and every recipe stored before this reads the same. The measured rule holds inside a mask exactly as it does on the whole frame: at 10000K the mask's pixels came out R/B 1.25 -> 1.72 against 0.994 -> 0.994 outside it. The Kelvin fit is a curve, so it is resolved on the JS side and handed over as a gain: maskUniforms grows one more float4 array (wb[i], three gains and a zero pad) between the spatial pair and the frame size, in declaration order like every other array in that buffer, and the shader multiplies the mask's colour by it right after EXPOSURE — one clamp, three multiplies by one for a mask that leaves the pair alone, and no second copy of the fit in SkSL. The spatial build is a second shader text with a second signature and reads the same array. App.tsx draws the rows where the panel's other rows already are, and TINT rides the existing maskKnobRow helper: same store unit (±10), same ±100 slider the frame-wide row wears since the develop knobs were deepened. COLOR TEMP keeps its own scale (2500..10000, step 100, "10000K"), because a temperature is not a percentage, and it carries the same swatch the frame-wide row paints. Verified: npx tsc --noEmit; npm run build; the repo's check set (highlight-knee, auto-tone, half, white-level, preview-match, library, scan-nav, roll-walk) all pass; and scripts/mask-wb-check.mjs, which is new here — it transpiles the gradientMask graph into a temp dir and checks the gain (neutral at 5500K/0, warm at 10000K, cool at 2500K, luma-preserving, ±TINT symmetric), the clamps, the uniform offsets (the gain lands where the shader reads wb[i], the frame size one array further on), and then compiles the real shader through CanvasKit and pushes a mid-grey through it: neutral leaves 128, 10000K warms it, +TINT magenta-ises it, and both the plain and the spatial builds and a two-mask list read it. The UI was driven on the built bundle too (a linear mask dragged on the preview, then the two rulers moved through their own range inputs): the mask column reports 11 rows, opens at 5500K / 0, takes 10000K and +40 ±100-scale TINT, the swatch follows, and the preview's own pixels warm inside the mask and nowhere else. ponytail: the mask's WB is not skipped for monochrome stocks the way the frame-wide matrix skips it — a local kelvin on a B&W frame is a tint someone asked for by hand, not the colour leak that rule exists to stop. Add the same isMonochromeBase guard (and pass the base filter into maskUniforms) only if that turn out to read wrong on a live B&W recipe. Co-authored-by: PenguinHarness <noreply@penguin.local>	2026-09-29 17:35:19 +07:00

Author

SHA1

Message

Date

3dtours

6d60d452e0

One move of the light for the whole app: a mask's EXPOSURE and its four tone knobs stop being a second opinion

A gradient mask had its own tone formula and its own exposure, and both
disagreed with the frame's. Before any of this was tidied, the mask ran a
smoothstep luma lift with an arbitrary 0.55..1.35 chroma clamp while the frame
moved the knots of a four-zone ramp, and the mask's exposure was a stop on
sRGB-encoded values while the frame's was a stop on light. Two names, four
moves, and the same slider meant different things depending on whether the
pixels were inside the shape you drew — the divergence §3.3 of the Android
port's compat doc warns about.

The maths is one string now (TONE_MATH_SKSL, interpolated by both passes): the
ramp the four knots build, the hue-preserving rebuild behind it, the transfer
pair, and exposureMove. A mask calls the same functions the frame calls.

The rebuild carries the chroma instead of re-scaling it. Lightness takes the
curve and the colour rides the difference — the channel differences move by ONE
shared scale k, pulled back only where the cube has no room left. The doc's
ratio (R_new = R_old * Luma_new / Luma_old) was the old reading and it is exact
only while nothing clips: a channel past 1.0 stops being scaled with its
neighbours and the hue goes with it. Measured on a flat patch frame, a skin
tone at 24.0° came back at 48.0° at HIGHLIGHT +100, a warm white at 37° at
57.4°, and under L = 0.5 the same ratio multiplied a near-black pixel's cast by
x30 — colour noise amplified, which is why the 0.55..1.35 clamp was there. The
scale is the chroma's own now: over 135 knob combinations on seven colours and
five greys, the ramp moves the luma and the hue does not move at all (Δ < 1e-9°).

EXPOSURE gets the same treatment, which is what the second half of the request
was: the linear domain decides where the luma is going, and the pixel is
rebuilt onto it through the same lightMove. The old pass multiplied the three
channels in linear light, so +1 EV clipped them by three different amounts:
measured on the scratchpad probe, 29.2° of hue drift on a skin tone at +1 EV
and 33.3° at +2, against 0.00° here. The stop is applied as a ratio on the
pixel's own encoded luma rather than pointed straight at the encoded linear
target, which is what makes the knob exactly the identity at 0 EV — the
transfer does not commute with the luma weights, so pointing at it brightened a
colour by a couple of code values even at zero. A grey is the knob it always
was: 128 through +1 EV is 176, the same number the linear per-channel multiply
put there, so nothing a user has dialled in moves.

Verified: `npx tsc --noEmit` clean, `npm run build` clean. highlight-knee-check
now runs EXPOSURE_SKSL for real — compiled with CanvasKit and four pixels
pushed through it, agreeing with the twin to a code value on a grey at +1 EV
(176), a skin tone at +1 EV and +2 EV, and a shadow at -2 EV; it also pins the
hue, the cube, the identity at 0 EV and the black pixel that has no light to
move. mask-wb-check compiles the mask pass and pushes the same stops through
it: 128 through +1 EV is 176, through -1 EV is 92, +2 EV lands the channel on
the ceiling at 255 and holds the hue within 3°. auto-tone-check,
preview-match-check, white-level-check, raw-develop-check, half-check and
roll-walk-check all pass. Live on the built bundle in a 1440x950 browser: the
LIGHT panel's EXPOSURE +1 EV takes the mid grey of a flat patch frame from
0.502 to 0.690 (a stop on light gives 0.686) and moves no patch's hue at -1 EV
(Δ 0.00°), and a linear gradient mask's EXPOSURE +1 EV and HIGHLIGHT +100 move
the pixels inside the mask (luma 185.9 -> 211.1 and 185.9 -> 197.5) while the
corner outside it does not move at all (220.2 -> 220.2), with no console error.

Co-authored-by: PenguinHarness <noreply@penguin.local>

2026-09-29 18:04:50 +07:00

3dtours

3b92e4e2d2

A gradient mask can carry the LIGHT column's white balance now: COLOR TEMP and

TINT, the same two rulers, read on the mask's own pixels.

The pair was asked for as the two rows the develop column already has, so it is
the same pair and not a second opinion about what a kelvin means: the gain comes
from one function, colorUtils.whiteBalanceGain(temperature, tint), which the
frame-wide colour matrix now calls too — the RGB kelvin fit of kelvinToRGB, the
symmetric ±0.08 magenta/green tint, and the division by the product's own Rec.709
luma that keeps a cast from being a brightness move. A mask that never touched
the pair reads 5500K / 0 (readMasks' own defaults, clamped to the ruler's ends),
whose gain is exactly (1,1,1), so nothing moves and every recipe stored before
this reads the same. The measured rule holds inside a mask exactly as it does on
the whole frame: at 10000K the mask's pixels came out R/B 1.25 -> 1.72 against
0.994 -> 0.994 outside it.

The Kelvin fit is a curve, so it is resolved on the JS side and handed over as a
gain: maskUniforms grows one more float4 array (wb[i], three gains and a zero
pad) between the spatial pair and the frame size, in declaration order like every
other array in that buffer, and the shader multiplies the mask's colour by it
right after EXPOSURE — one clamp, three multiplies by one for a mask that leaves
the pair alone, and no second copy of the fit in SkSL. The spatial build is a
second shader text with a second signature and reads the same array.

App.tsx draws the rows where the panel's other rows already are, and TINT rides
the existing maskKnobRow helper: same store unit (±10), same ±100 slider the
frame-wide row wears since the develop knobs were deepened. COLOR TEMP keeps its
own scale (2500..10000, step 100, "10000K"), because a temperature is not a
percentage, and it carries the same swatch the frame-wide row paints.

Verified: npx tsc --noEmit; npm run build; the repo's check set (highlight-knee,
auto-tone, half, white-level, preview-match, library, scan-nav, roll-walk) all
pass; and scripts/mask-wb-check.mjs, which is new here — it transpiles the
gradientMask graph into a temp dir and checks the gain (neutral at 5500K/0, warm
at 10000K, cool at 2500K, luma-preserving, ±TINT symmetric), the clamps, the
uniform offsets (the gain lands where the shader reads wb[i], the frame size one
array further on), and then compiles the real shader through CanvasKit and pushes
a mid-grey through it: neutral leaves 128, 10000K warms it, +TINT magenta-ises it,
and both the plain and the spatial builds and a two-mask list read it. The UI was
driven on the built bundle too (a linear mask dragged on the preview, then the
two rulers moved through their own range inputs): the mask column reports 11 rows,
opens at 5500K / 0, takes 10000K and +40 ±100-scale TINT, the swatch follows, and
the preview's own pixels warm inside the mask and nowhere else.

ponytail: the mask's WB is not skipped for monochrome stocks the way the
frame-wide matrix skips it — a local kelvin on a B&W frame is a tint someone
asked for by hand, not the colour leak that rule exists to stop. Add the same
isMonochromeBase guard (and pass the base filter into maskUniforms) only if that
turn out to read wrong on a live B&W recipe.

Co-authored-by: PenguinHarness <noreply@penguin.local>

2026-09-29 17:35:19 +07:00

2 Commits