Color correction

On this page 7

Color correction is the post-production process of adjusting exposure, white balance, contrast and saturation so that video footage looks technically correct, natural and consistent from shot to shot. It removes problems introduced by cameras, lighting and changing conditions, and it brings the image within the legal signal range of the delivery standard.

Color correction is usually the first stage of the color pipeline and precedes color grading, which shapes the creative look. The distinction matters in practice: correction aims at an objective, neutral baseline measured with scopes, while grading makes subjective, story-driven choices on top of that baseline.

How color correction adjusts the image

Color correction works by remapping pixel values in the red, green and blue channels, or in derived luma and chroma components. The core controls act on tonal ranges: the shadows, the midtones and the highlights can each be shifted in level and in color independently.

The classic lift, gamma and gain controls adjust the black point, the midtone curve and the white point. Offset moves the whole signal uniformly, which is especially useful on log footage. Because these operations are mathematical, the order matters: in most node or layer pipelines, normalization and balance come before any creative adjustment.

ControlMain tonal rangeTypical use
LiftShadows and black levelSet true black without crushing detail
GammaMidtonesBrighten or darken faces and mid-gray
GainHighlights and white levelPlace whites just below clipping
OffsetEntire signalBalance log footage and exposure
SaturationColor intensityRestore or reduce overall chroma
CurvesAny selected rangePrecise contrast shaping
Primary correction controls

Beyond these primary corrections, secondary corrections isolate parts of the image with qualifiers, shape masks or tracking windows. A colorist might, for example, remove a green cast from skin only, or reduce a bright sky without affecting the foreground.

Reading scopes during color correction

Color correction relies on video scopes because human vision adapts to color casts and display brightness. Scopes show the signal objectively, so decisions remain valid on any calibrated display.

ScopeWhat it showsDecision it supports
Waveform monitorLuma level across the frame widthBlack and white levels, exposure
RGB paradeSeparate red, green and blue waveformsWhite balance and color casts
VectorscopeHue and saturation around a circleSkin tones and saturation limits
HistogramDistribution of pixel valuesClipping and tonal spread
Scopes used in color correction

A neutral gray or white object should show equal red, green and blue levels on the parade. Skin tones of every complexion fall close to the skin tone indicator line on the vectorscope, which makes it a reliable reference for balancing faces.

Color correction compared with grading and LUTs

Color correction, color grading and LUTs are often confused because they use the same tools. The difference lies in purpose and in how adaptive each method is.

ProcessGoalNatureTypical stage
Color correctionAccurate, consistent, legal imageObjective, scope-drivenFirst pass
Shot matchingSame look across angles and takesComparativeAfter correction
Color gradingMood, style and storytellingCreative, subjectiveAfter matching
LUTFixed transform, often log to displayStatic, non-adaptiveInput or monitoring
Correction, grading and LUTs

A technical LUT can convert camera log footage to Rec. 709, but it cannot correct a shot that is underexposed or carries a color cast. Correcting exposure and balance before the LUT, or using a color-managed pipeline such as ACES, prevents the transform from amplifying errors.

A practical color correction workflow

A reliable workflow moves from global to local and from technical to creative. Each step builds on a clean result from the previous one.

  1. Calibrate the display and set the working color space, such as Rec. 709 at gamma 2.4 for broadcast.
  2. Normalize log or raw footage with a color space transform or input LUT.
  3. Set black and white levels on the waveform, keeping detail in shadows and highlights.
  4. Neutralize white balance using the RGB parade and a known neutral.
  5. Match shots within each scene against a hero shot.
  6. Apply secondary corrections for skin, sky and distractions.
  7. Check legal levels against the delivery specification before export.

Legal ranges depend on the standard. For 8-bit Rec. 709 video in limited range, black sits at code value 16 and nominal white at 235, and many broadcasters reject signals that exceed these limits beyond small tolerances.

Common color correction mistakes

Most correction errors come from trusting the eye over the scopes or from pushing footage beyond what it recorded.

SymptomLikely causeFix
Crushed shadowsLift set too lowRaise lift until shadow detail returns
Clipped highlightsGain pushed past the limitLower gain or use a highlight roll-off
Banding in gradientsHeavy correction on 8-bit footageShoot 10-bit or apply gentler curves
Shifting skin tones between shotsNo shot matchingMatch against a hero shot on the vectorscope
Image too blue or warmUncalibrated or adapted eyesVerify on the RGB parade and reference display
Mistakes and fixes

The idea in brief

Color correction is cleaning up the picture before styling it. A camera can record a scene too dark, too orange or too blue. Color correction fixes those mistakes so the image looks like what people actually saw.

It is a bit like tuning an instrument before a concert. Grading is the music itself, but correction makes sure every note starts in tune. That is why scenes filmed at different times still look as if they belong together.

A practical scenario

A corporate video crew films an interview with two cameras next to a window as clouds pass. Camera A was set to 5600 K and camera B to auto white balance, so the two angles disagree: one looks bluish, the other drifts orange, and the window brightness changes from take to take.

In correction, the colorist converts both log sources to Rec. 709, sets black and white levels on the waveform, and neutralizes a gray wall on the RGB parade. He then matches camera B to camera A, using the vectorscope to align the speaker’s skin tone, and adds a tracked window to hold the face exposure steady. The result is an interview that cuts seamlessly between angles and stays within broadcast legal levels, ready for a light creative grade.