DaVinci Resolve Canon R6 Mark II C-Log3 Import Settings

What Color Space Transform settings does R6 Mark II C-Log3 footage need?

This is the core mechanism, and it's worth writing down once so you're not hunting through a forum thread mid-edit.

Add a Color Space Transform to your first node and set Input Gamma to Canon Log 3. For Input Color Space, match whatever the camera's Color Space menu was actually set to when you recorded, not a default you're assuming:

1. If you recorded with Cinema Gamut selected, use Canon Cinema Gamut as Input Color Space.
2. If you recorded with BT.2020 selected, use Rec.2020 as Input Color Space.
3. If you recorded with BT.709 selected, use Rec.709 as Input Color Space.

In every case, set Output Color Space to DaVinci Wide Gamut and Output Gamma to DaVinci Intermediate. That moves your flat Log source into Resolve's own working space, where your actual grading happens. On your last node, before final output, reverse it: Input Color Space DaVinci Wide Gamut, Input Gamma DaVinci Intermediate, Output Color Space Rec.709 (or Rec.2100 for HDR delivery), Output Gamma Rec.709 (or ST 2084 / HLG for HDR).

This "in, grade, out" pairing is the same two-node pattern Shootlab's walkthrough of grading Canon C-Log footage in Resolve documents for Canon Log workflows generally, using DaVinci Wide Gamut as the intermediate stop between a camera's native color space and a Rec.709 or HDR delivery. The mechanism doesn't change camera to camera. What changes on the R6 Mark II specifically is that step one, confirming Input Color Space, has three possible correct answers instead of one.

Should you shoot Cinema Gamut, BT.2020, or BT.709 in-camera?

There's no single right answer here. It depends on how far you plan to push the grade and what else is on the timeline.

Cinema Gamut is the widest color space of the three, matching what Canon's own Cinema EOS line uses, and it gives you the most room to push saturation and hue in the grade before you run into clipping or banding. It's the closest match to how the R5 records by default, which matters if you're intercutting R6 Mark II footage with R5 footage on the same project and want both cameras to normalize the same way. BT.2020 is a narrower but still wide gamut, standard for HDR delivery pipelines, and a reasonable middle choice if you're not sure yet whether your final delivery is SDR or HDR. BT.709 is the narrowest of the three, essentially the same primaries as your final Rec.709 output, just carrying Log gamma instead of a standard picture profile on top of it. Log-in-BT.709 footage grades with less flexibility than Log-in-Cinema-Gamut, but it's also the fastest to get looking reasonable, since you're not stretching the color as far from your delivery target.

One of the recurring questions in our 100,000+ member editing community is exactly this: which color space should a hybrid shooter actually leave selected on a camera like this one. For most people cutting solo or on a small team, Cinema Gamut is the safer default, because it preserves the most information and costs you nothing extra in a normal grading pass. Reserve BT.709 for situations where you specifically want less grading latitude in exchange for footage that needs a lighter touch, like a run-and-gun setup where nobody downstream is doing a full color pass.

The one setting that's genuinely hard to undo is switching mid-project. If a camera's Color Space menu gets bumped from Cinema Gamut to BT.2020 partway through a shoot day, whether by a second shooter, a preset left on from a different job, or an accidental menu tap, nothing in the recorded file makes that change obvious later. The fix isn't a camera setting, it's a habit: log the Color Space menu on your shot sheet the same way you'd log a lens change, and check it again after handing the camera to anyone else on set. That single habit prevents the mixed-color-space timeline problem covered later in this guide.

Whichever color space you pick before you shoot decides your Input Color Space setting after you shoot, and there's no menu inside Resolve that tells you which one it was. Note it on set, or check the file's metadata through Clip Attributes, before you build a Color Space Transform on a guess.

Why does R6 Mark II footage look different over HDMI than internal recording?

Regardless of which color space you selected in the camera's menu, HDMI output on the R6 Mark II always converts to a fixed color space before it leaves the port, and that catches more editors than the color science itself does.

Rodrigo Polo, who has published Log-to-Rec.709 conversion breakdowns for Canon cameras for years, tracked this behavior down directly in Canon's own documentation. "Canon finally address the issue in the EOS R6 Mark II user manual, which clearly states that even though you can select 'Cinema Gamut' in their menu, it will only use Cinema Gamut with internal recordings, the HDMI output will always be Rec/BT.2020," he wrote after comparing Canon's manuals across camera generations. Older Canon manuals, including the original R5's, never stated this outright, which is part of why editors kept assuming Cinema Gamut carried through to an external recorder when it doesn't.

Practically, that means a clip recorded internally to your card with Cinema Gamut selected, and a clip recorded at the same moment to an external monitor over HDMI, come from the identical sensor and identical exposure but need two different Color Space Transform inputs to match. The HDMI clip needs Rec.2020 as Input Color Space, with Rec.709 as the fallback tag on older HDMI devices that don't support Rec.2020 at all. This is the same underlying limitation Polo documented on the original R5 and R6, detailed separately in his writeup on the R5, and Canon's R6 Mark II manual is simply the first one to state it in writing rather than leaving editors to discover it through mismatched footage.

The R6 Mark II's HDMI port converts everything to Rec.2020 before it leaves the camera, no matter what color space you picked for internal recording. Apply Cinema Gamut to HDMI-recorded footage anyway and you'll get a visible mismatch against internally recorded clips of the same scene, one that a correct Color Space Transform input, not a creative grading fix, is what actually resolves.

Resolve Color Management or a manual Color Space Transform node, which one?

Both reach the same destination. The right one depends on how consistent your R6 Mark II footage actually is clip to clip.

Resolve Color Management, set under Project Settings, Color Management, with Color Science set to DaVinci YRGB Color Managed, lets you tag a clip's Input Color Space and Input Gamma once through Clip Attributes or the Camera RAW tab, and Resolve applies the conversion automatically on every node downstream, no Color Space Transform effect required. That's the fast path if your whole shoot used one consistent camera setting.

The catch on the R6 Mark II specifically is that its three-way color space choice makes RCM's speed advantage less reliable than it is on a single-color-space camera like the R5. Batch-tagging a folder of clips as Cinema Gamut works fine until one clip in the middle of the folder was actually shot with BT.2020 selected, whether by a second shooter, a preset that got left on, or a menu change mid-shoot day that nobody flagged. A manual Color Space Transform node forces you to look at, and set, the Input Color Space on each clip or clip group individually, which is slower for a consistent shoot but catches an inconsistent one that automatic batch tagging would silently apply wrong. Our guide to color matching different cameras in DaVinci Resolve covers the broader normalization workflow if your R6 Mark II footage is mixed with a second camera body on the same project.

What happens if you intercut R6 Mark II footage recorded in different in-camera color spaces on the same timeline?

This is the situation the RCM-versus-manual-node decision above is really protecting you from, and it's common enough on a multi-day or multi-shooter project that it's worth walking through directly instead of leaving it as a hypothetical.

Say a two-day shoot started with Cinema Gamut selected, and on day two a second shooter picked up the same camera body with BT.2020 left selected from a previous project. Both cards land in the same Media Pool bin, both clips read as Canon Log 3 in the metadata, and nothing in Resolve's file browser flags the mismatch on its own; the in-camera color space choice isn't exposed as a separate visible column the way frame rate or resolution is. Grade the whole bin with one Color Space Transform preset built for Cinema Gamut, and the BT.2020 day-two clips come out with a color shift the wheels won't fully correct, because the node is decoding the wrong gamut into DaVinci Wide Gamut before you ever touch a wheel.

The practical fix is to sort before you grade, not during. Split clips into bins by their actual recorded color space, confirmed through Clip Attributes or on-set notes as covered earlier in this guide, before building a single Color Space Transform preset and applying it broadly. Resolve's Color Groups, applied per bin or per group of clips, let you park a correctly configured Color Space Transform on a group-level node and have every clip in that group inherit it, while a second group with a different Input Color Space setting handles the mismatched clips separately. That's a middle path between full Resolve Color Management, which tags every clip individually but assumes your batch tagging was accurate, and a fully manual node on each clip, which catches every mismatch but doesn't scale past a handful of clips.

If the mixed footage is coming from genuinely different cameras rather than one R6 Mark II with an inconsistent menu setting, for example R6 Mark II clips cut against R5 or Cinema EOS footage, treat that as a separate normalization pass entirely. Our guide to color matching different cameras in DaVinci Resolve covers building a shared look across camera brands once each source is individually corrected to DaVinci Wide Gamut, which is the step this guide's two-node pattern gets you to regardless of which color space or camera the clip started in.

A bin full of Canon Log 3 clips that all show the same codec and resolution can still hide two different in-camera color space settings, and Resolve won't warn you about the mismatch before you grade. Sorting by confirmed color space before applying any Color Space Transform preset, rather than after a color mismatch shows up in review, is the cheaper fix in every case.

Should you use Canon's official LUTs instead of a Color Space Transform?

You can, and it's a genuinely faster path for some workflows, but it carries the same in-camera-color-space risk the manual node approach solves for.

Canon publishes free Cinema Gamut-to-Rec.709 LUTs for cameras that shoot Canon Log, and Filmmaking Elements' rundown of free LUTs for the R6 Mark II specifically points to Canon's own downloadable .cube files as the source, the same format Resolve reads natively. Drop one onto a node as a single conversion and you get a usable Rec.709 image without touching an Input Color Space or Output Gamma menu.

The gap is that a Cinema Gamut LUT assumes the clip it's converting was actually recorded with Cinema Gamut selected, and it has no way to know or check that. Feed it a BT.2020 or BT.709 clip from the same camera and you get a mismatch, the same problem a Color Space Transform would throw if you picked the wrong Input Color Space, except the LUT gives you no dropdown to catch the mistake before it's baked into your grade.

A one-step LUT and a two-node Color Space Transform solve the same math, but only the Color Space Transform gives you a separate, checkable setting for which color space you're actually converting from. If every clip in your project used the same in-camera color space, either approach lands you in the same place. If you're not certain of that, or you're mixing HDMI and internal sources, the manual node's visible Input Color Space setting is the safer default.

What ISO should you shoot C-Log3 at on the R6 Mark II?

The Color Space Transform math assumes the footage was exposed the way the gamma curve was actually built for, and Resolve can't fully repair an ISO chosen wrong on the shoot day.

Per Canon's own product manual, "Canon Log offers a dynamic range of approx. 1600% at ISO 800 or higher," and the manual notes directly that dynamic range narrows below that point even though noise also drops. Canon's separate ISO Speed in Movie Recording page confirms the default automatic ISO setting range with Canon Log active runs from ISO 800 up to 25600, with the camera's minimum automatic speed also floored at 800 specifically because Canon Log is engaged. You can manually push ISO down to 100 with expanded settings, but doing so trades away the extra highlight headroom the whole curve exists to capture.

Exposing C-Log3 below ISO 800 doesn't just make footage darker, it throws away the extra highlight headroom that's the entire reason to shoot C-Log3 in the first place. If R6 Mark II footage still clips highlights or crushes shadows after a correct Color Space Transform, check the original ISO before spending more time on the color node. A curve that never got the exposure it was built around won't grade the way one that did.

Does noise get worse at higher ISO, and how do you fix it in Resolve?

Every stop above C-Log3's ISO 800 baseline buys correct exposure at the cost of more visible noise once that flat Log image gets stretched into a normal-contrast grade, and the Color Space Transform itself does nothing to reduce it.

DaVinci Resolve's Temporal and Spatial Noise Reduction tools, including the AI-driven UltraNR mode, are Studio-exclusive. Blackmagic's own Studio product page lists these among the Neural Engine features that separate Studio from the free version. Our full guide to DaVinci Resolve's noise reduction settings covers the specific Temporal frame counts, threshold values, and the tradeoff between denoising strength and softened detail in more depth than fits here.

Node placement matters as much as which mode you choose. Apply Noise Reduction on a node before your Color Space Transform pushes the clip out of Log, while the image is still flat and the noise pattern is closer to what the sensor actually captured, rather than after contrast and saturation have already amplified it. R6 Mark II footage shot near ISO 800 rarely needs this step at all; footage pushed several stops higher for a dim location usually does.

Staying close to ISO 800 on the R6 Mark II, rather than pushing it several stops over just because the range technically allows it, keeps most shoots out of the noise reduction workflow entirely. That's a shooting-day decision, not a Resolve setting, and it's the cheapest fix available.

What does independent lab testing show about the R6 Mark II's rolling shutter and dynamic range?

Everything so far in this guide is about Resolve settings. This section is about what the sensor and codec actually deliver before any of those settings touch the image, because a Color Space Transform can't fix a physical limitation of the camera.

DPReview's lab testing measured the R6 Mark II's rolling shutter at 18 milliseconds, or roughly 1/56 of a second, per DPReview's in-depth review of the camera. That's not fast by mirrorless standards, noticeably slower than a stacked-sensor camera like the Sony a9, which reads closer to 1/150 of a second, though DPReview's testing found it quick enough that most moderately fast pans and handheld motion don't show heavy skew. Whip pans, fast vehicle passes, or anything moving through frame quickly are where 18ms starts to show as a visible wobble or lean in vertical lines, an effect a Color Space Transform node has no ability to correct, since it's a timing artifact from how the sensor is scanned, not a color problem.

On dynamic range, DPReview's testing found the R6 Mark II performs comparably to its full-frame peers overall, but noted something worth knowing before you push shadows hard in a C-Log3 grade: in the deepest shadow regions, the review observed some detail smudging that reads as in-camera noise reduction being applied even before the file leaves the sensor, similar to what Canon applied on the EOS R5. Practically, that means the extra dynamic range C-Log3 promises at ISO 800 and above is real, but the very bottom of that range may already carry some softening you can't undo in Resolve, no matter which noise reduction mode you reach for afterward.

An 18-millisecond rolling shutter and shadow-region noise reduction baked in at the sensor level are physical limits of the R6 Mark II, not settings a Color Space Transform, Resolve Color Management, or a noise reduction node can correct after the fact. If a shot needs whip pans or extreme low-light shadow detail, that's a shooting-day and lens-choice decision, the same category as the ISO floor covered earlier, not something to chase on the color page.

Is C-Log3 the only Log option on the R6 Mark II, and how does that compare to the R5?

Yes, C-Log3 is the only Canon Log gamma curve available on the R6 Mark II, which simplifies one decision this camera forces on you even as it adds the color-space choice covered above.

Canon Log 2, which targets a wider dynamic range for a heavier VFX or color-intensive pipeline at the cost of needing more careful exposure and noise handling, is reserved for higher-tier bodies. The EOS R5 Mark II offers it alongside C-Log3, per Canon's own R5 manual documentation of its Log options and its Mark II successor's expanded Log menu, but the original R5 and the R6 Mark II covered in this guide do not. If you're shooting an R6 Mark II, the decision isn't which Log curve to use, it's which color space to pair the one available curve with, the question this guide spends most of its space answering.

C-Log3 is the format Canon's own documentation and most working colorists point hybrid shooters toward, and on the R6 Mark II it's not a choice among Log curves at all, it's simply the one your footage arrives in. The real decision this camera hands you is upstream of gamma, in the Color Space menu covered earlier in this guide.

Does this guide apply to the original EOS R6 or the newer EOS R6 Mark III?

Partly, and it's worth thirty seconds to confirm which body your footage actually came from before you assume the settings transfer directly.

The original Canon EOS R6, the predecessor to the Mark II, also shares the HDMI-to-Rec.2020 limitation. Rodrigo Polo's testing specifically names both cameras together: the EOS R5 and R6 "are unable to output C-Log 3 with Cinema Gamut through HDMI," as documented in his writeup on the fix for this behavior. Where the original R6 differs from the Mark II covered throughout this guide is that it doesn't offer the same three-way BT.709/BT.2020/Cinema Gamut color space menu; check your specific unit's manual before assuming that choice exists on an original R6.

Canon announced the EOS R6 Mark III on November 6, 2025, adding Canon Log 2 alongside Canon Log 3, a higher-resolution 32.5 megapixel sensor, and internal RAW recording at up to 7K, per Canon's own newsroom announcement of the Mark III. Don't assume this guide's recording-time table carries over to it, either. The Mark III's higher resolution sensor and heavier codec options give it its own thermal profile, distinct from the Mark II's: PetaPixel's testing of the R6 Mark III recorded a shutdown at 28 minutes 30 seconds in 4K 60p "Fine," while 4K 30p "Fine" ran for 1 hour 42 minutes before the test ended, a heat pattern shaped by the newer sensor and codec, not one you can extrapolate from the Mark II figures earlier in this guide. If your footage is tagged as coming from an R6 Mark III rather than a Mark II, confirm which Log curve you actually shot before applying any color setting from this guide, since C-Log2 needs a different Color Space Transform Input Gamma than C-Log3 does, and confirm recording times separately rather than assuming the Mark II's numbers apply.

If your clip's metadata reads EOS R6 Mark II and Canon Log 3, every setting in this guide applies directly. Confirm the exact camera and Log curve first, the same way you'd confirm which in-camera color space was selected, before troubleshooting a color mismatch that might actually be a camera-model mismatch instead.

Can the Canon R6 Mark II record RAW video, and does Resolve import it?

Not internally, and that's a real difference from the R5, which does.

The R6 Mark II has no internal RAW recording mode of any kind; its internal options are Canon Log 3, HDR PQ, and standard picture profiles, all through the H.264 codec covered earlier. What it does offer is external RAW through its HDMI port, up to 6K ProRes RAW to a compatible recorder, per Atomos's own compatible cameras listing for the R6 Mark II, which notes that recording ProRes RAW from this camera requires updating the connected Atomos device to a specific AtomOS firmware version before it will recognize the signal correctly.

DaVinci Resolve imports ProRes RAW natively, with no plugin required, and treats it through its own RAW debayer pipeline rather than as a standard Color Space Transform input. That's the same distinction covered on the R5's HDMI ProRes RAW path: a Ninja-recorded ProRes RAW .mov opens in Resolve's Camera RAW palette, not as a clip needing an Input Color Space and Input Gamma pairing the way C-Log3 footage does. If your R6 Mark II project is entirely internal C-Log3 footage, this section doesn't apply to you at all. If you added an external recorder for a specific shot, know that it's a different import path from everything else in this guide.

The Canon R6 Mark II has no internal RAW mode, every RAW option on this camera runs through an external recorder connected over HDMI. Confirm which path a given clip took, internal H.264 Log or external ProRes RAW, before assuming the Color Space Transform settings above apply to it.

Why won't some Canon R6 Mark II clips import into DaVinci Resolve at all?

If a clip won't come online or won't generate a thumbnail, that's a different failure mode from a wrong-looking grade, and on this specific camera it's less likely to be the codec itself than it would be on an R5.

Because the R6 Mark II records its standard Log footage in H.264 rather than H.265, it sidesteps the specific decode wall that trips up H.265-based cameras in Resolve's free version. theatre of noise's breakdown of Resolve's codec limitations clarifies that Windows lacks native HEVC (H.265) support out of the box, which is why H.265 clips show "Media off-line" with working audio until the separate Microsoft "HEVC Video Extensions" package is installed, a gap Filmmaking Elements' guide to fixing HEVC media offline errors walks through in more detail. H.264, the R6 Mark II's codec, doesn't share that specific Windows extension requirement, so most R6 Mark II clips that fail to import trace back to a different cause.

Check these instead, roughly in order: a missing or renamed .THM thumbnail sidecar file from a card that wasn't copied correctly, a partial or interrupted card-to-drive transfer that corrupted the file mid-copy, or a project timeline resolution set for standard 4K trying to preview a 6K oversampled clip before Resolve has generated the right proxy. Our general DaVinci Resolve media offline guide covers the broader set of causes, moved files, missing render caches, and drive permission issues, if none of the R6 Mark II-specific checks above resolve it.

A Canon R6 Mark II clip that won't come online in Resolve is less often the free-version H.265 wall that traps R5 footage, and more often a corrupted transfer or a missing sidecar file. Confirm the file itself copied correctly before assuming Resolve's codec support is the problem.

Do you need proxies to edit R6 Mark II 4K or external RAW footage smoothly?

Often, yes, and for a different reason depending on whether you're cutting standard internal footage or externally recorded RAW.

The R6 Mark II's 6K oversampled 4K mode leans on long-GOP H.264 decoding, which is real GPU and CPU work on a lower-powered laptop even though the files themselves are relatively compact on disk. External 6K ProRes RAW pulled in over HDMI is a heavier load again, since every frame needs a debayer pass before Resolve can display it at all, the same computational cost RAW footage carries on any camera. Both situations respond to the same fix: build lower-resolution stand-ins before you start cutting.

Right-click a clip or bin in the Media Pool and choose Generate Optimized Media to transcode a local proxy copy in the background, at a resolution and codec set under Project Settings, Master Settings, Optimized Media and Render Cache. Pixflow's proxy workflow guide recommends leaving Proxy Media Resolution on "Choose Automatically" for most 4K and 6K sources, which sizes proxies proportional to your timeline resolution automatically, dropping to Quarter or Eighth Resolution manually if playback still stutters. Codec choice follows platform: ProRes on a Mac, DNxHR on Windows, per Beginners Approach's breakdown of Resolve's proxy resolution and codec options, with DNxHR LB for the smallest files and DNxHR HQX when storage isn't a constraint.

Proxies fix playback performance, not color. Building optimized media from R6 Mark II footage doesn't change or skip the Color Space Transform steps covered earlier; Resolve relinks to full-resolution originals for the final render regardless of what you cut with, so the Input Color Space you set still has to match what was actually recorded.

Do you need DaVinci Resolve Studio for any of this?

No, not for the core C-Log3 workflow this page covers. The Color Space Transform effect, Resolve Color Management's automatic tagging, and native ProRes RAW decoding all work fully in the free version, with no functional restriction on the color math itself.

Per Blackmagic's own free version product page, the free tier covers professional color grading tools without a Studio license. Where Studio genuinely enters the picture, per Blackmagic's Studio product page, is Temporal or Spatial Noise Reduction on a noisy high-ISO clip, Dolby Vision metadata authoring or HDR10+ analysis for a specific delivery target, or multi-GPU rendering if a 6K RAW timeline is bringing a single-GPU machine to a crawl.

Getting Canon R6 Mark II C-Log3 footage to grade correctly is a color management problem, and color management is not a feature Blackmagic gates behind the Studio license. Don't let a general free-versus-Studio comparison talk you into a $295 upgrade just to make R6 Mark II footage look right, unless you've actually hit one of the specific noise reduction or HDR delivery walls above. Our full DaVinci Resolve Studio price breakdown covers exactly what the license changes if you're weighing it for other reasons.

A step-by-step example: taking one C-Log3 clip from flat to Rec.709

Here's the full sequence in order, from a fresh R6 Mark II clip in the Media Pool to a delivered Rec.709 grade.

1. Confirm the source. Check Clip Attributes or your on-set notes for which color space was selected, Cinema Gamut, BT.2020, or BT.709, whether it was recorded internally or over HDMI, and whether the ISO was 800 or above. This determines your Input Color Space in the next step.
2. Build optimized media if playback is stuttering. For 6K oversampled 4K or external ProRes RAW footage, generate proxy media at Quarter or Automatic resolution before you touch color.
3. Node 1, Color Space Transform in. Set Input Gamma to Canon Log 3. Set Input Color Space to match the recorded setting: Canon Cinema Gamut, Rec.2020, or Rec.709, using Rec.2020 for anything recorded over standard HDMI regardless of the camera's menu. Set Output Color Space to DaVinci Wide Gamut, Output Gamma to DaVinci Intermediate.
4. Node 2, Noise Reduction, if the shot needs it. If the clip was shot at a high ISO and shows visible grain once out of Log, add Temporal or Spatial Noise Reduction here, right after the Color Space Transform, before any contrast moves.
5. Node 3, primary correction. With the clip in DaVinci Wide Gamut, set white balance from a neutral reference if available, then bring black point and white point in using the Lift and Gain wheels, watching the scopes for clipping.
6. Node 4, creative grade. Contrast shaping, color balance, secondary qualifiers, and power windows as the shot needs. Once in a normal working space, R6 Mark II footage grades like footage from any other camera.
7. Node 5, Color Space Transform out. Input Color Space DaVinci Wide Gamut, Input Gamma DaVinci Intermediate, Output Color Space Rec.709, Output Gamma Rec.709 or Rec.709-A.
8. Check against the troubleshooting table below before calling the grade final, especially if this clip is intercut with footage recorded under a different in-camera color space.

Five nodes handle the mechanism end to end, with Noise Reduction as the one optional insertion depending on ISO. The step most specific to this camera is step one, confirming which of three possible color spaces the clip actually carries, since guessing wrong there produces a mismatch no amount of node 4 grading will fix cleanly.

A second worked example: taking an external ProRes RAW clip from HDMI to Rec.709

The five-node sequence above is for internal C-Log3 footage. An external ProRes RAW clip pulled in over HDMI takes a different path from the first node onward, and it's worth walking through once so the two workflows don't get mixed on the same timeline.

1. Confirm the file is actually RAW. In the Media Pool, a ProRes RAW clip opens its own Camera RAW tab in the Clip Attributes or Color page inspector, rather than exposing Input Color Space and Input Gamma the way a standard C-Log3 MP4 does. If a clip from an external recorder doesn't show that tab, it was likely recorded as standard ProRes video over HDMI, not RAW, and belongs on the standard C-Log3 path instead, tagged Rec.2020 as covered earlier in this guide.
2. Set the debayer resolution. In the Camera RAW panel, choose a debayer quality for the timeline, Half or Quarter Resolution for smooth scrubbing while editing, separately from the one used for final render, Full Resolution or Premium if the GPU can handle it. This is the RAW-specific equivalent of the proxy workflow covered above, and it doesn't require generating separate optimized media files.
3. Set the RAW decode settings to match how it was shot. ProRes RAW from an external Atomos recorder carries color science metadata from the camera at the point of capture. Confirm ISO, white balance, and color space in the Camera RAW panel match your on-set notes rather than leaving them on the recorder's defaults, the same discipline as confirming Input Color Space on a standard C-Log3 clip.
4. Grade from the decoded RAW, not through a second Color Space Transform. Once the Camera RAW panel is set, the clip already sits in a working color space Resolve understands; stacking a Color Space Transform node on top of it the way you would a standard Log clip double-converts the image and produces a result that won't match your internally recorded C-Log3 clips from the same shoot.
5. Match it back to your C-Log3 sequence. With both sources sitting in DaVinci Wide Gamut, RAW decoded through Camera RAW, C-Log3 converted through the CST-in node, grade both against the same reference clip or scopes target so a cut between an internal C-Log3 shot and an external ProRes RAW shot of the same scene doesn't show a visible seam.

A ProRes RAW clip's color space is set through the Camera RAW panel at the point of debayer, not through a second Color Space Transform node stacked on top of it. Treating an external RAW clip like a Log clip that needs its own CST-in node is the most common way this specific workflow goes wrong, and it produces a double-converted image that's hard to diagnose without knowing to check for it.

Why does C-Log3 footage still look wrong after you've applied the Color Space Transform?

Work through these in order before assuming the footage itself is somehow bad.

First, and most specific to this camera, an Input Color Space mismatch. Confirm you're not applying Cinema Gamut to a clip actually shot with BT.2020 or BT.709 selected, or applying Cinema Gamut to anything recorded over HDMI, which always converts to Rec.2020 regardless of the camera menu. This is the most common cause of "some clips from the same shoot look right and others don't."

Second, log-on-log. If a clip looks oversaturated or hue-shifted in a way that doesn't respond normally to color wheels, check whether a LUT is stacked on top of a Color Space Transform doing the same conversion, or whether Resolve Color Management is automatically tagging a clip that also has a manual node applying its own transform. This is also the failure mode described above for external ProRes RAW: a Camera RAW panel doing its own decode, plus a Color Space Transform node stacked on top of it.

Third, timeline color space mismatch. If your project's overall Color Management doesn't match your output node's target, Rec.709 versus Rec.2100 being the common version of this, footage can look correct in isolation and wrong against your monitoring setup or on export.

Fourth, and easy to miss: confirm the clip actually decoded at full resolution, rather than grading a proxy that never relinked to full-resolution media before export.

Fifth, and specific to motion rather than color: if the mismatch looks like a wobble or lean on vertical lines during a fast pan rather than a color shift, that's rolling shutter, not something a Color Space Transform, RCM, or a LUT will ever fix, since it's a sensor timing artifact rather than a color pipeline issue.

When the menus alone don't get you there fast enough

Everything in this guide is a settings problem with a settings answer, but finding the right menu on a deadline is its own kind of friction, especially the first time you're checking whether a clip was shot Cinema Gamut, BT.2020, or BT.709 before you can even start the Color Space Transform.

TryUncle is the on-screen assistant for DaVinci Resolve on macOS. Ask in plain words, and Uncle points at the exact control on your screen. If you're on a Mac asking out loud "why does my C-Log3 footage look grey," that's a faster route to the Color Space Transform panel than searching this page a second time mid-session. It's a paid app, currently in founder pricing, and macOS only, so it won't help a Windows editor working through Resolve's menus. Check TryUncle for current pricing before deciding if it fits your workflow.

It's worth knowing the category has other tools built on a different premise. Sottocut, PremiereCopilot, heyeddie.ai, and cutagent.ai automate parts of an edit or answer chat questions about your footage, rather than watching your live screen and pointing at controls. If your goal is automating a cut, those solve a different problem. If your goal is finding the right Color Space Transform setting faster on your own screen, on-screen guidance is the more direct tool for that specific job.

What's the full settings cheat sheet at a glance?

Before the table, the one-line version. Confirm which color space was actually selected in-camera before assuming Cinema Gamut. Tag HDMI-recorded footage as Rec.2020 regardless of that camera menu setting. Stay at or above ISO 800. Reach for a manual Color Space Transform node the moment a shoot mixes color space settings or cameras, and lean on Resolve Color Management only when you're confident every clip shares one setting. Plan long uncropped 4K takes around the roughly 40-minute heat limit, not just the card capacity. And if a clip won't come online at all, check the file transfer before touching color, since that's rarely a codec wall on this camera the way it is on H.265-based bodies.

Confirm which color space you actually selected before you build a Color Space Transform, keep ISO at 800 or above, and tag HDMI-recorded footage as Rec.2020 no matter what the camera menu shows. That's the entire workflow. Most R6 Mark II footage that "looks wrong" in Resolve isn't broken. It's carrying one of three possible color space tags, and now you know which questions to ask before you pick one.