darktable preset inspired by FujifilmVelvia 100

What the style changes

The style adds five darktable modules. It aims at Velvia 100’s reputation: very strong saturation just under Velvia 50’s, electric blues, cooler greens and yellows, a discreet magenta bias, strong contrast and very fine grain. The sources are detailed on the Fujifilm Velvia 100 film page. All figures below come from renders made with darktable-cli at 1000 px, measured in CIELab, on the same RAW files as the other styles.

• Clear contrast, noticeably darker blacks. The tone curve is steep in the midtones. Contrast rises from 14.3 to 18.8 (+31 %), and the darkest 5 % of the frame fall from L 2.6 to 0.6.
• Very strong saturation, below the Velvia 50 style. A color balance rgb instance named “film” raises saturation, chroma and vibrance together. Sky chroma goes from 27.5 to 41.1 (+49 %) and sea chroma from 22.2 to 36.0 (+62 %). In the same test, the Velvia 50 style reaches 43.3 on the sky and 40.1 on the sea.
• Blues barely pulled toward indigo. rgb primaries and color equalizer move the sky from 273.6° to 275.9° and the sea from 270.1° to 276.4°, a smaller move than the Velvia 50 style (+6° to +12°).
• Discreet magenta. The “film” instance tints the shadows slightly magenta and keeps the highlights clean. The pink T-shirt gains 52 % chroma (33.7 to 51.3).
• Slightly pinker skin. Skin hue moves from 44.2° to 42.3°, in the direction that reviews report on light skin.
• Very light grain. Grain is added at coarseness 800 and strength 8. The datasheet’s RMS 8 makes grain nearly invisible on film, so this is an aesthetic choice.

For comparison, the t3mujinpack style for this film, rendered on the same files, is more moderate: contrast 16.5, sky chroma 35.3, sea chroma 30.6. It leaves the shadows open (L5 2.8) and moves skin hue slightly the other way (45.3°).

What it leaves alone

• Your base development. The style file holds only the five modules above: no exposure, no sigmoid or filmic, no white balance, no color calibration and no LUT. Those stay as you set them.
• Your pipeline order. The style file carries no module order list, so it does not impose one on your image.
• Your own color balance rgb. The style’s instance is named “film”, so it is added next to yours without changing it (unless the photo already has an instance named “film”, for example from an earlier application of the style).

What it does not leave alone: applied from the lighttable or the darkroom, each module of the style replaces the instance of the same name on the photo, or is added if there is none. The style’s unnamed modules, rgb primaries, color equalizer, tone curve and grain, therefore overwrite any settings you had already made in those modules. Your earlier settings stay in the history stack. Exporting with a style, or rendering with darktable-cli as our measurements do, adds the style’s modules as separate instances instead. Details in Install and apply darktable styles.

Limits

• An interpretation, not a calibration. No film was scanned or measured to tune the style. The sources disagree on Velvia 100’s saturation relative to Velvia 50, and some reviews confuse Velvia 100 with Velvia 100F.
• The drift of greens and yellows comes from research summaries, not from a measurement. The measurements cover no green or yellow zone, so the cooler greens and yellows are unchecked.
• The slope read on the datasheet curve is approximate.
• Light skin turns clearly red, which matches the reviews but may look excessive.
• Narrow latitude and long-exposure color shifts are not simulated. The grain is added lightly although RMS 8 makes it nearly invisible on film.

Tips

• For a gentler look, lower the saturation of the “film” color balance rgb instance.
• Mind the skin. On portraits, switch off color equalizer or reduce its orange hue shift.
• Installation: see Install and apply darktable styles.