If moving different tone equalizer sliders changes the foreground and sky together, first check whether the guided mask gives those areas different brightness levels. The sliders do not identify objects or image positions: they adjust pixels according to the brightness of the mask at each pixel. A sky and foreground can therefore respond together if their mask brightness falls in the same range, even when they are far apart in the frame.
Check what the controls are addressing
In darktable, the guided mask is a monochrome map used to divide the image into areas of similar luminosity. In the simple tab, each slider corresponds to a brightness zone in that mask. In the advanced tab, the curve’s horizontal axis is mask brightness, and its vertical axis specifies the exposure adjustment. The histogram behind that curve represents the mask, not the original image. See the official tone equalizer documentation for the module’s behavior and controls.
Use this ordered diagnosis for a landscape where you want to lift a dark foreground without brightening the sky:
- Identify the two areas you want to control separately. Do not choose slider positions based only on where those areas appear in the frame or how dark they look in the original photo.
- Show the guided mask using the display exposure mask control. Compare the mask values in representative points within the sky and foreground. The mask preview and, when available, the cursor’s mask-intensity readout are clues to the ranges each area uses.
- Open the advanced tab and inspect the mask histogram behind the curve. Ask whether the distribution is overly bunched and whether it extends across the mask-brightness range relevant to the adjustments you need. The manual recommends spreading it across the full tonal range and covering as many control points as possible for flexibility; use that as a qualitative guide, not a numeric threshold or pass/fail test.
- If the two areas are not usefully separated, go to the masking tab and adjust the mask controls, then inspect the preview and histogram again. If the mask distribution is already suitable, adjust the curve or simple-tab zones corresponding to the observed mask values, making a small change and checking both areas.
Keep two decisions separate: first diagnose whether the mask assigns useful tonal ranges; then decide whether to change the mask or the exposure curve. Mask exposure compensation shifts the mask histogram left or right; the manual describes using it to recenter after changing image exposure. Mask contrast compensation expands or compresses the histogram. These are distribution controls, not object-selection controls.
Record a landscape diagnosis
The following is an author-created worksheet for a hypothetical landscape, not a darktable screenshot, guaranteed result, or software test. The example assumes the mask preview suggests that sky and foreground occupy overlapping ranges. Record observations from your own image rather than treating these sample values as defaults. The numbers below are illustrative mask-intensity readings, not a claim about how darktable measures or counts pixels.
| Diagnostic pass | Hypothetical observation | Decision to record |
|---|---|---|
| Initial mask preview | Sky and foreground sample points show similar mask-intensity readings | The selected points do not show a useful separation; do not assume different sliders will isolate them. |
| Initial advanced histogram | In this hypothetical example, the distribution is tightly grouped near the middle and does not extend across the mask-brightness range needed for the intended curve adjustments | Inspect masking controls and compensation before trying to solve the overlap with a steeper curve. Judge spread qualitatively; the manual's advice to cover more control points is a flexibility recommendation, not a countable threshold. |
| After a proposed mask adjustment | In this hypothetical comparison, sky and foreground sample points show clearly different mask-intensity readings, and the distribution extends farther across the relevant curve range | The regions now have distinct readings in this example; try modest curve adjustments and check both areas. Wider distribution may offer more flexibility, but this remains a hypothetical observation, not a predicted effect. |
| After a proposed curve adjustment | Foreground is brighter; sky appears nearly unchanged in the hypothetical review | Record whether the intended separation holds in the image; do not infer that every pixel in either region received the same adjustment. |
The comparison separates two questions: did the mask make the areas distinguishable, and did the chosen curve produce an acceptable result? A broader histogram may offer more control points to work with, but it does not guarantee a clean visual separation.
When separation still fails
If sky and foreground remain close in mask intensity after adjustment, or if changing the mask creates unwanted transitions, do not keep moving unrelated curve points in the hope of selecting by location. Tone equalizer control follows mask brightness. The manual says the mask filters aim to preserve sharper boundaries while smoothing detail within tonal regions, and that balancing those aims may require experimentation. Inspect the mask preview while changing masking controls.
Filter choice is a separate decision from whether the two areas have distinct mask values. The masking tab offers options including guided filter, average guided filter, eigf, and averaged eigf; they concern how the mask is smoothed. The manual notes differences in exposure sensitivity and blur behavior, but does not prescribe one option for every landscape. After checking whether the mask ranges are separable, consider filter behavior if the mask boundary or local contrast is the problem. The masking-tab documentation describes these options, smoothing diameter, edge refinement, and filter diffusion.
If the regions genuinely have similar tonal values and cannot be separated adequately by this mask workflow, the tone equalizer’s brightness-based controls may not provide the independent, location-based adjustment you need. Consider a different local adjustment method when the requirement is explicitly spatial or object-based, and review its selection boundary separately. Do not assume a filter change will make two same-tone objects independently addressable.
Common questions
Do the top sliders always mean foreground and the bottom ones sky?
No. Their usual shadow-to-highlight interpretation depends on the mask histogram being spread across its tonal range. Read the mask histogram and the mask values for the areas you care about.
Should I change exposure compensation or mask contrast compensation first?
Use the symptom to choose what to inspect: exposure compensation shifts the distribution; contrast compensation spreads or compresses it. Make one change at a time and recheck the histogram. The manual does not establish a universal starting value.
Should I choose a different filter to separate the regions?
Only if smoothing or edge behavior is the issue. First establish whether the mask gives the regions distinguishable brightness ranges; filter choice is not an object selector.
Sources and next step
This guide interprets public documentation and uses explicitly hypothetical acceptance examples, not application tests. Controls and availability can change; verify the cited manual against your installed version.
For work created in Imagild, carry the input requirements and acceptance checklist into the photo editor. The external software controls discussed above belong to their respective products.