GA — 043 / P Black and white

darktable preset inspired by IlfordXP2 Super

Soft gray black and white with gentle midtones, rolled-off highlights and very fine grain.

Film
GA — 043 · sheet, datasheet, sources
darktable
5.6 or later
In darktable
Styles › Pellicules › N&B
File
Pellicules - N&B - Ilford XP2 Super.dtstyle

A darktable 5.6 style built from the Ilford XP2 Super datasheet: native modules, no LUT, measured on real RAW files. In the full pack.

After: beech forest, with the preset inspired by Ilford XP2 Super (GA — 043 / P) applied (simulation). Before: the same beech forest, base development without a preset, in color.
fig. 1 · simulation on a digital RAW, not a film scan Drag, tap or use the arrow keys

Translation into darktable

The 4 modules of the preset, what each one does and its settings
ModuleRoleSettings
color calibration (main instance)Switched off, so chromatic adaptation cannot tint the gray; left at darktable's default settings so it can be switched back onDisabled; “as shot in camera”, CAT16
color calibration (instance “N&B”)Black-and-white conversion with weights integrated from the spectral sensitivity curve of the Harman datasheet (2856 K tungsten light, compensated)Gray = 0.2559·X + 0.2597·Y + 0.4844·Z
tone curveSoft slope (about 0.55 on the datasheet), short toe, long shoulder: highlights that roll off, open shadowsNodes (0, 0) (0.08, 0.055) (0.2, 0.185) (0.35, 0.345) (0.5, 0.5) (0.65, 0.655) (0.8, 0.80) (0.92, 0.925) (1, 0.99)
grainVery fine, soft grain; set by eye because the datasheet gives no RMSCoarseness 900, strength 10, midtones bias 85

Measurements

The measurements come from a few test photos of our own, which are not published on this site. The photos shown here are different.

Before and after the preset, measured in CIELab: L lightness, C chroma, h hue angle
PhotoZoneBaselineMetric BeforeAfterFilm expectation
0775 whole image Base rendering Lsd (contrast) 14.3 14.6 Soft contrast: about unchanged
1014 whole image Base rendering Lsd (contrast) 30.7 30.3 Soft contrast: about unchanged
1014 whole image Base rendering L5 (blacks) 2.6 1.4 Open shadows, no clipping
1014 whole image Base rendering L95 (whites) 87.9 84.9 Long shoulder: softened highlights
0775 whole image Base rendering L95 (whites) 75.2 75.1 Long shoulder: softened highlights
0775 sky Base rendering L 55.3 59.6 Lighter: panchromatic film, blue-sensitive
0989 skin Base rendering L 78.2 73.8 Darker skin
1014 pink T-shirt Base rendering L 47.3 45.0 Darker reds
1014 whole image Base rendering C (neutrality) 8.6 0.0 Neutral gray, no color cast

Base rendering: darktable’s default processing of the RAW file, without the preset. Luminance conversion (black and white): the same photo converted with plain luminance weights, to isolate the effect of the film’s spectral sensitivity.

What the style changes

The style converts the photo to black and white using the spectral sensitivity Harman published for XP2 Super, then softens the tone curve and adds a very fine grain. The film itself is described on the Ilford XP2 Super film page. All figures below come from renders made with darktable-cli 5.6.0 at 1000 px, measured in CIELab.

  • Gray from the film’s spectral sensitivity. A color calibration instance named “N&B” mixes the channels with weights computed from the sensitivity curve on Harman’s datasheet, integrated against the CIE 1931 color matching functions. The weights (0.2559, 0.2597, 0.4844) are very close to those obtained for HP5 Plus (0.2537, 0.2510, 0.4953), and the two curves have a similar general shape.
    • Against the base rendering, the sky gets lighter (L 55.3 → 59.6), while skin (78.2 → 73.8) and a pink T-shirt (47.3 → 45.0) get darker.
  • Soft contrast, long shoulder. The datasheet’s curve has a slope of about 0.55 and a long shoulder (our reading of the graph). Contrast stays about where it was (Lsd 14.3 → 14.6 on one photo, 30.7 → 30.3 on another). The darkest 5 % of the frame open slightly (L 2.6 → 1.4) and the brightest 5 % come down (L 87.9 → 84.9 on photo 1014).
  • Very fine grain. Coarseness 900, strength 10, midtones bias 85.
  • A truly neutral gray. Chroma falls from 8.6 to 0.0.

How it compares with the other black-and-white styles in the archive (1000 px renders, measured the same way):

Measure XP2 Super FP4 Plus HP5 Plus Tri-X 400
Lsd, photo 0775 14.6 16.9 15.8 18.0
Lsd, photo 1014 30.3 32.1 30.6 33.5
L95, photo 1014 84.9 88.5 86.7 90.6
Grain (sky of 0775, 2560 px) 1.9 3.5 5.6 6.8

Grain here is the standard deviation of the image minus its Gaussian blur (radius 4 px), on the sky of photo 0775 (0.3 to 0.7 × 0.05 to 0.3), rendered at 2560 px. Without grain the value is 0.76.

What it leaves alone

  • Your base development. The style contains no exposure, no sigmoid or filmic and no LUT. Those settings stay as you made them.

  • But it switches off your main color calibration. The style contains two color calibration instances:

    • a disabled one with no instance name, which replaces your main color calibration and switches it off. It is set to “as shot in camera” with CAT16, like darktable’s default, so you can switch it back on;
    • the “N&B” instance, which is added and does the black-and-white conversion.

    Why: an active color calibration running after the black-and-white conversion lets its chromatic adaptation tint the gray. Film does not white-balance either.

  • It also replaces your tone curve and grain. When you apply a style from the lighttable or the darkroom, each module without an instance name replaces the one you already had in the same module, and is added otherwise. Your earlier settings remain in the history.

  • Your pipeline order. The style file carries no module order list, so it does not impose one on your image.

  • No color cast and no “print” look. The style stays neutral: nothing in the sources fixes a tint, and it does not reproduce the “enhanced contrast” of a darkroom print. The details of applying styles are in Install and apply darktable styles.

Limits

  • An interpretation, not a calibration. No film was scanned or measured to tune the style. The grain is set by eye: the datasheet gives no granularity, and 1.9 against 3.5 for FP4 Plus is not an RMS ratio.
  • The spectral axis. The datasheet’s axis is printed “Sensitivity”, graduated 0.5 and 1.0, with no unit. We assumed a logarithmic scale, as validated on HP5 Plus. With a linear scale the weights would be (0.245, 0.242, 0.514): a slightly lighter blue and slightly darker skin.
  • The curve was read on a small graph, to about 0.05 in density. It describes the negative, not a print.
  • Soft contrast is not measured by the sources. They call it soft without giving a figure, and reviews are few and use different scenes. Some readers also write “inky blacks”, and the result depends on the lab and the scan.
  • No color cast. A real lab scan can come out warm, cool or purplish.
  • The shoulder. It acts on an image already rendered by sigmoid, so it is an interpretation.
  • The grain. darktable’s grain is procedural noise: it imitates the texture, not the dye clouds.

Tips

  • Apply it early. If you plan to fine-tune the tone curve or grain yourself, apply the style first: it would replace those settings if you made them before.
  • See the spectral conversion alone by switching off the tone curve and grain modules the style set.
  • Back to color: switch the main color calibration back on and switch off the “N&B” instance.
  • Installation: see Install and apply darktable styles.

Get the preset

GA — 043 / P · darktable preset

XP2 Super
for darktable

Full pack · 67 presets · yours to keep
.dtstyle file, darktable 5.6 or later

Try the free pack first · What the full pack contains