Workflow
How to Match Footage From Different Cameras in DaVinci Resolve
Most multi-camera mismatches start before any grading: a wrong input transform, a missing setting, a LUT applied twice. Get color management right, then match the parts viewers compare. Here is the order that works.
In short
- Put every camera into one color-managed working space with the correct input transform.
- Use a chart and Color Match if one was filmed.
- Match by hand in this order: black level, exposure, white balance, contrast, saturation, then specific hues.
- Accept that some sensor differences in saturated colors remain and need targeted work.
Why cameras look different
Each camera maker designs its own log curve, its own gamut and its own processing. Sensors respond differently to colors, especially saturated reds, greens and the narrow spectra of LED lights. Settings matter too: exposure index, white balance, ND filters, compression. Two cameras pointed at the same scene with the same white balance setting will not give the same picture, and that is normal.
The goal is not identical pictures. The goal is that a cut between cameras is not noticeable.
Step 1: color management
This step solves most of the problem and causes most of the mistakes.
- Choose a managed pipeline: ACES (with ACEScct as the grading space) or DaVinci Wide Gamut / DaVinci Intermediate with Resolve Color Management. Both bring different cameras into one working space. See ACES vs DaVinci Wide Gamut.
- Check the input transform of every clip. Resolve can read the camera's color space from metadata, but not every file carries it, and some formats or transcodes lose it. Verify each camera format in the clip attributes or the input transform settings.
- Know what the file is. A camera log ProRes, a raw file, a Rec.709 transcode from a drone and a phone's HDR video all need different input transforms.
- Use the raw settings for raw footage. White balance and exposure index of raw formats are best set in the raw decode, not in a node.
- Do not stack transforms. A camera LUT applied to footage that the color management already converted gives a double conversion: too much contrast and saturation.
Step 2: use a chart if you have one
If a color chart was filmed by each camera in the same light, Resolve's Color Match can measure the patches and build a correction. Do this on a separate node, early in the tree. Then check the result on skin, not only on the chart, and remember that the chart only describes the light it was filmed in.
Step 3: match by hand, in order
- Pick the reference camera. Usually the A camera, or the camera that covers the main character.
- Black level. Match the depth of the shadows first. Use the waveform and look at the shadow detail on screen.
- Exposure on the subject. Match the brightness of faces or the key subject, not the frame.
- White balance and tint. On skin or a known neutral in the same light. Watch the vectorscope: skin from both cameras should sit along the same direction, with a similar spread.
- Contrast. Different lenses and sensors give different mid-tone contrast. Adjust around the subject's exposure so it does not move.
- Saturation. Often one camera reads more saturated. Match it overall, then check skin again.
- Specific hues. If greens, reds or a lit sign still differ, use hue-specific tools (Hue vs Hue, Hue vs Saturation, the Color Warper or a qualifier) on that range only.
Step 4: check in playback
Play every cut between cameras on your calibrated display. Many matches that look close in a split screen still jump in motion, and some that look different in stills play fine. The cut decides.
Common mistakes
| Mistake | What you see |
|---|---|
| Wrong or missing input transform | One camera flat and desaturated, or contrasty and over-saturated, with no clean way to match. |
| Double conversion (LUT plus color management) | Crunchy shadows, clipped highlights, neon colors. |
| Data vs video levels mix-up | Lifted or crushed blacks on one source. |
| Matching frame averages | Skin wrong on the wider shots. |
| Fixing a hue problem globally | Skin matches, but the whole frame shifts. |
| Different exposure index without compensating | Different noise and highlight headroom. The highlights clip differently and need separate handling. |
What remains after a good match
Some differences are in the capture: one camera clipped a window that the other held, or noise differs at high exposure index. Grading can hide some of this, but it cannot recreate information that was not recorded. Note these shots for the director and cinematographer instead of forcing a match.
Where AutoBalance fits
AutoBalance works inside Resolve's color management. It reads the project's settings and the clips' input transforms rather than setting them. Its private beta has been tested fully on ARRI LogC3 material, and Balance alone on ARRI LogC4, Sony S-Log3 and X-OCN and Apple Log; other cameras are not tested yet. The compatibility list has the details. It does not replace the color management step above.
FAQ
Should I use the camera manufacturer's LUT to match cameras?
In a color-managed project, use the input transforms instead. Manufacturer LUTs go from one specific log format to one output, and mixing them with color management easily causes double conversions.
Can I match cameras without a chart?
Yes. Match on skin and on neutrals in the same light, in the order above. A chart just makes the first step faster and more precise.
Why do reds and greens still differ after matching?
Sensors render saturated colors differently. A global correction lines up skin and neutrals; specific hues need targeted adjustments on that range.
Sources
- Blackmagic Design: DaVinci Wide Gamut Intermediate (PDF)
- Academy Color Encoding System: ACEScct encoding specification
- Blackmagic Design: DaVinci Resolve Color page (Color Match, Shot Match)
Product names belong to their owners. Features change between releases; check the current documentation for your version.