No. In darktable’s AgX module, Read Exposure estimates the black and white relative exposures from the settings in the exposure module; it does not analyze picture content. The white and black pickers, by contrast, use selected image areas. Before relying on the estimate, check whether an enabled exposure instance actually affects the image before AgX in the pipeline. The AgX manual explains the button’s input and instance selection.
Check the conditions, then choose a range method
Use Read Exposure when the exposure module’s settings represent the input that reaches AgX. If exposure comes after AgX, or does not affect the image, the manual says using the button makes no sense. Its estimate can account for in-camera exposure compensation, in-camera highlight preservation where supported by darktable, and manual exposure adjustments already handled through that module. It is not applied automatically.
- In darktable’s pipeline, confirm the exposure module is before AgX. The manual says modules before AgX operate in scene-referred space; its recommended workflow sets overall mid-tone exposure with the exposure module first.
- List every exposure instance and note its order, enabled state, and whether it is masked. Do not assume the visibly selected instance is the one Read Exposure will use.
- Record each instance’s exposure value and whether any in-camera compensation or highlight preservation has already been handled in the exposure module. This is a practical recordkeeping check, not a separate darktable feature.
- If the input is actually influenced by the module, click Read Exposure and record its result. If your aim is instead to set endpoints from visible image content, use the white and black pickers on suitable areas, or the auto tune levels picker for both. AgX projects the selected input range into its curve; values beyond the selected black or white endpoint are clipped to the corresponding output boundary.
Compare the results in a worksheet
The table below is a hypothetical diagnostic, not a prediction of exact darktable readings. The numbers are example observations to help decide which method matches the task. Repeat the comparison on your image and keep the selected areas and instance state in your notes.
| Hypothetical observation | What it suggests | Decision to record |
|---|---|---|
| Exposure is before AgX; instance A is enabled and unmasked; its value is +0.7 EV; instance B is masked at -0.3 EV; Read Exposure returns a range | The manual says the button uses the first enabled, unmasked instance, so A is the relevant instance in this example | Compare the estimate with the intended input, not with B’s masked value |
| Every exposure instance is masked; the first instance is A at +0.7 EV | The manual says the first instance is used when all instances are masked | Record that exception; do not interpret the button as sampling a visible area |
| A picker on a chosen bright region and one on a chosen dark region give endpoints different from the button’s result | The methods rely on different inputs: module settings versus selected image areas | Choose the range-setting method that matches whether you want to account for the exposure-module input or define endpoints from image content |
Limits and next steps if it fails
A different picker result does not by itself show that Read Exposure is malfunctioning: the controls answer different questions. If the estimate seems irrelevant, first check pipeline order, then inspect every exposure instance’s enabled and mask state. Reassess whether the exposure module actually changes the signal reaching AgX and whether in-camera compensation has been processed there. If not, do not treat the button’s estimate as a content-based measurement; use the pickers for content-selected endpoints instead.
The manual documents which instance is read, but does not define a content-analysis algorithm for the button or a guaranteed numerical relationship between its estimate and picker results. Nor does it say that the estimate is automatically applied. Keep those limits separate from the worksheet’s proposed comparison.
Common questions
Does Read Exposure find the brightest and darkest pixels? No. The manual distinguishes its settings-based estimate from the content-reading pickers.
Which exposure instance does it use? The first enabled, unmasked instance; if all are masked, the first instance.
Should I use the button for every image? Only when exposure-module settings influence the input to AgX. For image-area-based endpoints, use the relevant pickers. [The recommended workflow also places mid-tone exposure adjustment before AgX](https://docs.darktable.org/usermanual/5.6/en/module-reference/processing-modules/agx/).
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.