Primary color correction is the stage of color grading that adjusts the entire image at once, setting overall exposure, contrast, white balance and saturation for every pixel in the frame. It works on the full tonal range through controls such as lift, gamma and gain, offset, curves and RGB balance, without isolating any region or hue.
In a typical color grading workflow, primaries come first: they create a neutral, technically correct and consistent base, after which secondary color correction, power windows and creative looks are added. A clean primary pass makes every later step simpler and reduces artifacts caused by pushing isolated corrections too hard.
How primary corrections act on the image
Primary tools apply a mathematical transfer function to each color channel, so the same rule changes every pixel with a given value. The core model divides the tonal range into shadows, midtones and highlights, each controlled separately but with smooth, overlapping influence so that adjustments never create hard boundaries.
The classic lift, gamma and gain controls are the foundation. Lift raises or lowers the black point while pinning white, gain scales the signal toward white while pinning black, and gamma bends the midtones. A fourth control, offset, shifts the whole signal up or down equally and is the main tool for log footage and film-style printer-light grading.
| Control | Main tonal range | Typical use |
|---|---|---|
| Lift | Shadows, pivoting at white | Set black level, remove color cast in shadows |
| Gamma | Midtones | Brighten or darken faces and middle gray |
| Gain | Highlights, pivoting at black | Set white level, balance highlight color |
| Offset | Entire range equally | Exposure and overall balance on log footage |
| Contrast and pivot | Spread around a chosen point | Expand or compress tonal range |
| Saturation | All colors | Overall color intensity |
Each control exists as a color wheel with a master luminance slider. Moving a wheel toward a hue adds that color to its range and subtracts the complementary color, which is how a blue cast in shadows is neutralized by pushing lift slightly toward orange.
Reading scopes during primary correction
Primary correction is guided by video scopes, not by the monitor alone, because eyes adapt quickly to color casts. The waveform monitor shows brightness levels across the frame, the RGB parade shows each channel side by side, and the vectorscope shows hue and saturation around a central neutral point.
| Question | Scope to use | What to look for |
|---|---|---|
| Are blacks and whites in range | Waveform monitor | Signal within legal limits, no clipping |
| Is the image neutral | RGB parade | Channels aligned on white and gray objects |
| Is saturation reasonable | Vectorscope | Trace within target boxes, skin near skin line |
| Do two shots match | Parade and vectorscope | Similar shapes and levels across shots |
For Rec. 709 delivery, black typically sits near 0 percent or 0 IRE and diffuse white near 90 to 100 percent on the waveform, with legal video ranges of 64 to 940 in 10-bit code values. HDR delivery uses different targets measured in nits, but the principle of setting endpoints first remains the same.
Primary versus secondary color correction
The defining difference is scope of selection. Primary corrections touch every pixel, whereas secondary color correction first isolates a hue, saturation, luminance range or shape and adjusts only that selection. The two are complementary rather than competing.
| Aspect | Primary color correction | Secondary color correction |
|---|---|---|
| Area affected | Entire frame | Qualified color range or window |
| Main tools | Wheels, curves, offset, contrast | HSL keys, power windows, tracking |
| Order in workflow | First | After primaries are balanced |
| Main risk | Clipping or crushed detail | Noisy edges and visible masks |
| Purpose | Balance and consistency | Targeted fixes and emphasis |
A practical rule follows from this: fix globally what is global. A white balance error affects the whole frame, so correcting it with a secondary key on one object leaves the rest of the image wrong and creates a mismatch.
A typical primary workflow
In applications such as DaVinci Resolve, colorists often apply a color space transform or input LUT first so that log footage is interpreted correctly. Primary corrections then follow a stable order that prevents one adjustment from undoing another.
- Normalize the input from log or camera space to the working color space.
- Set black and white points with lift, gain or offset.
- Adjust contrast around a sensible pivot, usually near middle gray.
- Remove casts by balancing the RGB parade on neutral objects.
- Set overall saturation, checking skin tones on the vectorscope.
- Match shots in the scene to a chosen hero shot.
Working in a scene-referred color space such as ACES or a wide-gamut log space keeps adjustments behaving consistently across cameras. Applying exposure changes in a linear or log domain also tends to look more natural than pushing gamma on display-referred Rec. 709 data.
Common mistakes in primary correction
- Grading on an uncalibrated monitor, which bakes the display’s error into every shot.
- Crushing blacks or clipping highlights while chasing contrast, losing detail that cannot be recovered.
- Correcting casts by eye only, since the eye adapts and the scopes do not.
- Starting with a creative look before the base is neutral, which makes shot matching much harder.
- Overusing saturation, which pushes colors outside the delivery gamut and makes skin look artificial.
The simple explanation
Primary color correction is the first cleanup of a video image. The colorist makes the whole picture correctly bright, gives it proper contrast and removes unwanted tints. Every part of the frame changes together.
Think of adjusting the brightness, contrast and color settings of a television, but with far finer control. Only after this base is right does the colorist work on single areas, such as a sky or a face. A good foundation makes the rest easier.
How it is used in practice
A documentary team films an interview with two cameras in a room lit by a window and a tungsten lamp. Camera A looks slightly blue and underexposed, while camera B looks warm and flat. In the grading suite, the colorist first converts both log clips to Rec. 709 with a color space transform.
Using the waveform, the colorist sets black near 0 percent and the brightest highlight on the face near 70 to 75 percent. On the RGB parade, a gray wall is used as a neutral reference, and the gain and lift wheels are nudged until the three channels align. Saturation is then brought to a level where skin sits on the skin line of the vectorscope.
After about ten minutes of primary work per camera, both angles cut together seamlessly. Only then does the colorist add a subtle vignette and a light secondary on the subject’s face, which stay small because the primary pass already solved the main problems.