When shadows look cyan but highlights look warm, do not start with one global color shift. First identify which tonal regions show each cast, then adjust the matching Shadows, Midtones, or Highlights controls in Kdenlive’s Color Balance effect. The manual defines positive values as moving toward the named primary color and negative values as moving toward its complement. Treat that direction as a control rule, not a promise that a particular value will neutralize your shot.
Diagnose the cast, then adjust one region at a time
- Compare the image’s dark, middle, and bright areas separately. Look for plausible neutral objects such as gray fabric or a white wall, but do not assume that every shadow or highlight ought to be neutral. Skin and colored materials are not neutral references.
- If available, inspect an RGB Parade. Kdenlive describes it as separate component graphs useful for spotting casts in shadows and highlights. Compare RGB levels where you expect a neutral area: roughly aligned traces support neutrality; a consistent separation is evidence of imbalance. A parade is evidence to interpret, not an automatic diagnosis. The official Video Waveform Monitor manual explains this display and its limits through the listed component modes.
- Add or select Color Balance on the clip. Confirm the control group by its actual label: RGB Shadows, RGB Midtones, or RGB Highlights. Labels identify the affected tonal range; do not infer the range from control color or position.
- Change only the group associated with the cast you are addressing. For a cyan-looking shadow, consider the red control’s positive direction because cyan is red’s complement. For a warm-looking highlight, inspect the blue-yellow balance and decide whether the image needs movement toward blue or away from yellow. Make a small trial adjustment, then compare the target area and nearby areas before changing another group.
- Decide whether to enable Preserve lightness. Kdenlive documents it as an option to preserve lightness while changing color balance; it is off by default. Compare the image with it off and on rather than assuming either setting is always preferable.
Hypothetical acceptance worksheet
The observations below are invented to demonstrate decisions, not results from a tested clip. They specify no slider values: the manual does not define a universal correction amount.
| Hypothetical observation | Range to address | Direction to consider | Acceptance check |
|---|---|---|---|
| Dark gray jacket looks cyan; RGB Parade traces diverge in the low end | Shadows | Positive red moves away from cyan; judge by image, not direction alone | Jacket looks less cyan without making nearby dark fabric visibly red |
| Skin in the middle looks approximately neutral | Midtones | No correction indicated by this observation | Skin remains plausible while other ranges change |
| Window highlight looks warm; bright neutral trim also appears yellow | Highlights | Consider movement toward blue | Highlight appears less warm without an unwanted blue cast elsewhere |
| Shadow detail appears lighter or darker after a color adjustment | Same adjusted range | Compare Preserve lightness on and off | Choose the version whose brightness remains appropriate, not simply the more neutral one |
Review the whole frame after each isolated change. A successful correction should reduce the intended cast without making skin, neighboring tones, or unaffected regions less believable. These are editorial acceptance checks, not guarantees about how the effect behaves on every shot.
Limits and what to do if the correction fails
A split cast does not prove that the three ranges can be corrected independently without affecting other parts of the image. The manual identifies the controls and their intended regions, but does not promise a particular degree of isolation or provide universal numeric settings. Lighting, colored surfaces, and deliberate creative grading can also make a non-neutral appearance intentional.
If a trial worsens the image, return that control group toward its prior setting before testing another group. Recheck whether the reference object was actually meant to be neutral, then compare the parade with the visible image. If the cast spans several ranges, address the strongest supported imbalance first and reassess; avoid compensating by shifting every group at once. If a neutral-looking result damages natural skin or material color, accept the original character or use a more localized workflow rather than forcing global neutrality.
Three quick questions
Should I correct every unequal RGB trace? No. Unequal values can be expected in colored subjects or intentional lighting. Use the parade with image context and neutral references where available.
Which channel removes cyan from shadows? The documented direction rule says positive red moves balance toward red, the primary complementary to cyan. Whether that improves a particular image remains a visual judgment.
Does Preserve lightness keep every pixel equally bright? The manual says the switch preserves lightness while changing balance. It does not specify a guarantee of identical brightness for every pixel, so verify the result in context.
Sources and next step
- Color Balance — Kdenlive 26.08 Manual 26.08 documentation
- Video Waveform Monitor — Kdenlive 26.08 Manual 26.08 documentation
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 film workspace. The external software controls discussed above belong to their respective products.