
Use Pick neutral spots in RawTherapee's Film Negative tool to estimate the film's red and blue exponent ratios from two trustworthy scene neutrals. Choose one lighter and one darker region, with neither clipped. Recheck white balance afterward. This is a roll-specific color-response starting point, rather than a request to make every photographed object gray.
Key Takeaways
- The two samples represent neutral objects in the photographed scene, not simply light and dark pixels.
- Film-channel ratios and white balance perform different jobs.
- Reusing a roll calibration still requires reviewing each frame.
Find a frame that contains usable evidence
An orange-looking negative does not identify which photographed surface was neutral. A pale wall might have been cream, and a dark coat might have been blue. Use objects whose original neutrality is supported by the scene or other records. If that evidence is absent, leave the sample choice uncertain instead of forcing an attractive result.
Choose a frame from the roll with two suitable regions of different brightness. Avoid reflections, clipped illumination and visible damage. The target is a shared material response, so a convenient sample from an unrelated roll is not an equivalent reference.
First establish a usable positive-conversion input through the negative preparation guide. A scanner's transparency mode does not itself supply the scene-neutral pair required here.
Estimate the ratios before final white balance
The RawPedia Film Negative reference documents the Color-tab tool and its Pick neutral spots control. Enable Film Negative, activate that picker, and select the lighter and darker scene neutrals. Supported inputs include non-RAW photographs and Bayer or X-Trans RAW files; do not infer support for every other RAW arrangement.
The tool models different film-channel responses. Its reference exponent sets the green-channel response, while red and blue ratios describe their relationship to it. The neutral-pair operation therefore has a different purpose from clicking one patch with a white-balance picker.
Adjust white balance after choosing the neutral pair. The documentation places that calculation later in the processing pipeline, so a previously chosen white balance may need revisiting. Do not change several unrelated contrast tools simultaneously while judging whether the pair was credible.
Carry the starting point across the roll
Copy the processing profile to other photographs from the same roll as a proposed comparison. Inspect a bright frame, a darker frame and one with recognizable skin or other supported colors. Do not use the original calibration frame as the only acceptance evidence.
A shared starting profile is not a promise that every exposure or final print appearance should match. Separate the reusable film response from subsequent frame-specific decisions. The RAW preparation guide explains why a rendered positive and its editable source remain different assets.
If one frame fails, inspect clipping and sample validity before recalibrating the entire roll around it. Keep the original capture and identify the frame that supplied the pair. Accept the calibration as a documented working interpretation, with unresolved colors left uncertain; no conversion test or historical color reconstruction is claimed here.
Related tasks
- Why Does My Scanned Film Negative Look Orange and Reversed?
- Should I Develop a RAW Photo Before Using AI Image Editing?
Sources and scope
- RawPedia: Film Negative: Current English documentation covers Bayer, X-Trans and non-RAW inputs; film-negative inversion uses reference sampling and dedicated settings rather than simple channel inversion.
Based on official documentation retrieved and actually opened on October 23, 2026 (Asia/Shanghai). The examples and acceptance procedures are proposed, not application tests. Web-tool readable source extracts are retained in research/photo-data-076-080-web-evidence.json; no raw HTTP status is inferred from search availability.
Prepared by Imagild Editorial, the company editorial team responsible for these guides. AI assisted research and drafting. This is documentation-based guidance with proposed examples, not a claim of hands-on testing or an independent product review. Report a correction.
Cover: William Liu / Pexels, Pexels License. This is an illustrative source photograph, not an application screenshot or a generated test result.