How to Make Car Edits in DaVinci Resolve

If you’re trying to figure out how to make car edits in DaVinci Resolve without the whole project getting messy halfway through, the issue usually is not one tool. It is the workflow.

A lot of the time, the footage itself is fine. The problem is that the edit starts without much structure, then every change starts affecting something else. Speed changes start breaking clips. Stabilization gets overwritten. The grade starts to feel forced. By the end, the video still works, but it took more effort than it should have.

What I want is a workflow that stays clean from the rough cut all the way through export. That means making decisions in the right order, keeping the color process organized, and not asking DaVinci Resolve to fix problems that should have been handled earlier.

Build the selects first

The first thing I do is go through all of the footage and pull the clips I think I might use. Then I drop those into a selects bin.

This is one of those small changes that has a big impact. Instead of digging through everything while trying to cut, I already have a smaller group of usable shots in front of me.

If some of the footage was shot at 60fps and I am editing on a 29.97 timeline, this is also where I interpret those clips to 29.97 in clip attributes. That gives me clean slow motion right away before I start building the cut.

What’s really happening here is simple. You are separating decisions. First you decide what footage is usable. Then you decide how it fits into the sequence. That tends to hold together better than trying to do both at the same time.

Cut to the music and build flow

Once the selects are ready, I drop in the music track and mark the beats. Then I use those markers as a guide for the rough cut. From there, I scrub through the clips, find the best part of each shot, set in and out points, and start building the timeline.

As I am doing that, I am thinking about flow. I want shots to connect through movement, direction, or a clean visual handoff. I am also thinking about momentum. I want the sequence to build, not just feel like random clips stacked together.

A good clip can still feel wrong if it does not connect well to the one before it or after it. That is where a lot of edits start to feel loose.

What I do with drone shots

For drone shots, I usually bring them closer to normal speed, then slow them slightly if I want smoother movement.

That usually feels better than forcing every drone clip into heavy slow motion. The movement still feels controlled, but it keeps a little more life in it.

Fine cut and retiming

Once the rough cut is working, I move into the fine cut. This is where I tighten timing, trim frames, swap shots, and clean up the pacing.

If a slowed clip starts to break apart, I switch the retime process to Optical Flow and test what gives the cleanest result.

The reason this works is that retiming only looks good when the footage can actually support it. Once frame interpolation starts falling apart, the shot starts to feel fake very quickly.

Stabilize after the cut feels solid

After that, I move into framing and stabilization. For most gimbal shots, I handle stabilization on the Edit page. If I need more control, especially on tougher shots or speed ramps, I go into Fusion. I usually start with Translation or Similarity and adjust from there.

That order matters. A lot of people stabilize too early, then keep changing the edit afterward. Once timing, framing, or retimes change later, the correction work can start getting messy.

I also keep the render cache set to User so I can cache clips when playback starts slowing down. Smooth playback makes it much easier to judge whether the shot is actually working.

Set up color the right way

Once the edit feels solid, I move into color. In this workflow, I am using DaVinci YRGB with DaVinci Wide Gamut and DaVinci Intermediate as the working space. I also keep 3D LUT interpolation set to Tetrahedral because the image quality is better.

Then I group clips by camera type so each camera gets its own input transform before everything lands in the same working space.

If Sony and DJI clips are not normalized cleanly into the same space first, you end up fighting those differences shot by shot.

Input transforms

For Sony A7 IV footage shot in S-Log3, I use a Color Space Transform from Sony S-Gamut3.Cine and S-Log3 into DaVinci Wide Gamut and DaVinci Intermediate.

For DJI Mini 4 Pro footage shot in D-Log M, I use another transform node from D-Gamut and Rec.709 into DaVinci Wide Gamut and DaVinci Intermediate.

Start the grade with exposure and contrast

Before I make final color decisions, I turn proxies off and view the footage at full quality if my system can handle it. Then the first thing I adjust is exposure, usually with the HDR Global wheel. After that, I add contrast, set the pivot around 0.336, and add a little midtone detail, around 10.

From there, I shape the image with the primary wheels. This is where the image usually starts to feel dialed in, because I am not chasing style yet. I am just getting the footage into a strong baseline first.

Why I use a separate curves node

Once exposure and contrast feel right, I add a gentle S-curve in the next node. I like doing this in a separate curves node because it gives me more control over how the contrast rolls off through the image.

When I add handles to the curve, I can shape the shadows and highlights more gradually instead of forcing a hard contrast move. That helps the blacks feel richer without getting crushed, and it helps the highlights stay bright without turning harsh.

Refine saturation before white balance

Once I have a good foundation, I start refining color. I usually check saturation before white balance because it helps me see color casts more clearly.

For saturation, I like using a separate node with the color space set to HSV. Then I isolate the saturation channel by turning off channels 1 and 3. That gives me more control over how saturation responds.

If I want to adjust the overall saturation, I can work through the gamma wheel. If I want to control hotter or more saturated parts of the image, I can use the gain wheel.

Use chromatic adaptation for white balance

For white balance, I have used a few different methods in Resolve, but lately I have been using chromatic adaptation when I want more precise control.

I like it because it lets you define the source illuminant and target illuminant intentionally instead of just pushing temperature and tint until it feels close.

That is useful when the image is almost right, but still feels a little off.

Add look nodes to unify the sequence

The first node in my tree is for noise reduction, but I usually leave it off until the end because it is heavy on performance. Once the primary grade feels good, I move into the look.

This is where I add look nodes across the group, and a lot of the time that includes LUTs before the final output transform. The point is to unify the sequence so everything feels like it belongs together.

Build separation in the secondary grade

After that, I move into the secondary grade. This is where I start creating separation between the car and the background.

Usually that means an inverted power window using the HDR Global scale to lower exposure outside the subject, plus a little sharpening. If I need more depth, I use a gradient to darken the foreground.

If the background stays too bright, the car never really separates. Once you control that, the subject starts to read more clearly.

Clean up the blacks

A lot of times I also desaturate the blacks. I isolate the darker parts of the image, clean up the key, and use Saturation vs Saturation in the curves window to pull that saturation down slightly.

It is subtle, but it helps the shadows feel cleaner and more controlled.

Finish with glow, split tone, and trim

Once that is done, I move into finishing effects. Usually that includes glow and split tone.

With glow, I am trying to soften or contain the highlights a little and sometimes bring a little life back into the paint, especially on white cars. If you push it too hard, the shot gets ugly fast. The same goes for split tone.

I usually use split tone to create a little more separation between warmth and coolness and add that last bit of polish. I also keep Protect Neutrals checked so the image maintains its original color tone.

Then the final node is what I call trim. This is the last cleanup pass. I go back mainly with the HDR wheels and tighten up anything that still feels slightly off.

Add noise reduction at the end

For temporal noise reduction, I usually start with three frames, set it to Better, use Medium motion estimation, and then adjust from there.

What I am looking for is the point where it starts helping without making the image feel plastic. If you overdo it, the shot starts to feel fake very quickly.

I also like saving that as a shared node so I can update or toggle it across similar clips without going one by one.

Copy grades without breaking your work

Once the hero shot is fully dialed in, I save it as a still and apply it to matching shots. One important detail here is to make sure you are applying color only, not all settings.

If you apply everything, you can overwrite tracking data and lose stabilization work you already did. That is a brutal fix if you catch it late.

Most clips do not need a full rebuild after that. Usually they just need a few smart refinements like window sizing, exposure adjustments, or white balance tweaks.

Matching DJI and Sony footage

If I am matching footage from different cameras like DJI and Sony, I still use the same general process. Match exposure first. Match white balance. Use the scopes as a guide.

My 4K export workflow

For the final export, I set my in and out points and move over to the Deliver page. Then I change my timeline settings to 4K even though I edited on a 1080 timeline.

After that, I export a master in Apple ProRes 422 HQ. That becomes my master file. From there, I bring it into HandBrake and make an H.264 version for YouTube.

That gives me a much smaller file while still holding solid quality, which makes uploads easier.

Final thoughts

A clean car video workflow in DaVinci Resolve is really about order.

Build the selects first. Cut with intention. Retime and stabilize after the edit is working. Normalize footage into the same working space before grading. Then shape exposure, contrast, color, separation, and finishing in a way that does not fight the rest of the project.

What I like about this workflow is that it keeps things from falling apart every time I tweak something. The edit stays cleaner. The grade stays more controlled. By the time I get to export, I am refining the work instead of rescuing it.