# RawCut — full text
> Browser-based tools for reading, converting and rebuilding photographic edits. Everything runs on the visitor's own machine via the File API and WebGL: no upload, no account, no server-side processing.
This document contains the complete text of every tool description and guide on the site. The index-only version is at https://www.rawcut.art/llms.txt.
---
# Tools
## XMP → CUBE
URL: https://www.rawcut.art/tools/xmp-to-cube
Accepts: .xmp
Produces: .cube or .3dl
Convert a Lightroom .xmp preset into a .cube or .3dl 3D LUT for DaVinci Resolve, Premiere or Final Cut — at any grid size from 17³ to 65³. Baked through the same colour engine that renders the preview, so the export matches what you see.
- Reads Lightroom Classic, Lightroom CC and Camera Raw sidecars
- Any grid from 17³ to 65³, in Adobe .cube or Autodesk .3dl
- Tone curves, HSL, split toning and colour grading
- Runs entirely in the browser — nothing is uploaded
## CUBE Viewer
URL: https://www.rawcut.art/tools/cube-viewer
Accepts: .cube
Produces: Analysis + preview
Drop a .cube file to see what it actually does: the neutral-axis response curve, per-channel behaviour, hue shifts, and a live before/after on your own photo.
- Neutral-axis response curve — contrast and colour cast in one plot
- Per-channel R/G/B transfer curves
- Hue-shift map
- Live preview on a sample frame or your own photo
## Metadata Reader
URL: https://www.rawcut.art/tools/metadata-reader
Accepts: Photo or .xmp
Produces: Full metadata read-out
Drop a photo or a sidecar and read every field: camera, lens, the exposure triangle, white balance, tone curve points, the eight HSL bands, colour grading and masks.
- EXIF: camera, lens, aperture, shutter, ISO, focal length, GPS
- XMP: every Lightroom develop setting
- Tone curve and HSL visualisation
- No upload — parsed locally with the File API
## Image Compressor
URL: https://www.rawcut.art/tools/image-compressor
Accepts: Any image
Produces: WebP, JPEG or PNG
Compress a photo or convert it to WebP without uploading it. Shows the real saving, and says so honestly when a setting made the file bigger instead.
- WebP, JPEG and PNG output with a quality control
- Optional resize — the saving resizing buys dwarfs quality alone
- Reports the actual percentage saved, including when it grew
- Runs in a Web Worker, so large files never freeze the page
## Histogram Viewer
URL: https://www.rawcut.art/tools/histogram-viewer
Accepts: Any image
Produces: Histogram + clipping analysis
See the RGB and luminance distribution of any photo, plus exactly how much of the frame is crushed to pure black or blown to pure white — the two failures no edit can undo.
- RGB and luminance histograms
- Shadow and highlight clipping as a percentage of the frame
- Tonal distribution across shadows, midtones and highlights
- Analysed in the browser — nothing is uploaded
## Social Crop
URL: https://www.rawcut.art/tools/social-crop
Accepts: Any image
Produces: Platform-sized JPEG
Crop a photo to Instagram, YouTube, Pinterest or Open Graph dimensions with the subject where you want it, not stuck in the centre. Warns you when a crop would have to be upscaled.
- Ten presets at the sizes platforms actually display
- Click to move the crop off centre
- Warns when the source is too small and would be upscaled
- Cropped in the browser — nothing is uploaded
## Palette Extractor
URL: https://www.rawcut.art/tools/palette-extractor
Accepts: Any image
Produces: Hex, CSS, Tailwind or JSON
Extract a photo's dominant palette with hex codes, weighted by how much of the frame each colour actually occupies. Copy it out as hex, CSS custom properties, a Tailwind fragment or JSON.
- Median-cut quantisation — the same pass that builds Color-DNA strips
- Swatches sized by their real share of the frame
- Four to ten colours
- Copies as hex, CSS variables, Tailwind config or JSON
## EXIF Remover
URL: https://www.rawcut.art/tools/exif-remover
Accepts: Any image
Produces: Clean JPEG, PNG or WebP
See exactly what your photo is carrying — GPS coordinates, camera body, timestamps — then remove all of it. Unlike server-side strippers, the file never leaves your device, which rather defeats the purpose otherwise.
- Shows GPS location, camera, lens and timestamps before removing them
- Removes EXIF, XMP, IPTC and ICC in one pass
- Honours EXIF orientation so the photo is not rotated
- Runs entirely in the browser — nothing is uploaded
## Aspect Ratio
URL: https://www.rawcut.art/tools/aspect-ratio-calculator
Accepts: Frame dimensions
Produces: Bar and crop pixel sizes
Work out the exact letterbox bars, visible area and export size for any target ratio — 2.39:1 anamorphic, 16:9, 4:5, 9:16 and the rest. Shows the bars answer and the crop answer side by side, because they are not the same number.
- Eleven ratios from 2.76:1 down to 9:16
- Letterbox and pillarbox bar thickness in pixels
- Export size for cropping instead of adding bars
- Live proportional diagram of the framing
## CUBE → XMP
URL: https://www.rawcut.art/tools/cube-to-xmp
Accepts: .cube
Produces: .xmp
The reverse conversion. Tone curves are extracted exactly from the LUT's own response; colour and saturation are estimated. Shows you the round-trip error rather than hiding it.
- Master and per-channel tone curves extracted from the LUT
- Split-tone and saturation estimated from the neutral axis
- Live round-trip preview showing the approximation error
- Exports a Lightroom-compatible .xmp preset
## LUT Strength
URL: https://www.rawcut.art/tools/lut-strength
Accepts: .cube
Produces: .cube
Reduce or exaggerate a 3D LUT and export the result. Node opacity lives in your project file and cannot travel; this writes the strength into the LUT itself, so it survives to OBS, a camera monitor, or whoever you send it to.
- Any strength from 0 to 150%, blended toward the identity transform
- Live before/after on a sample frame at the chosen strength
- Above 100% pushes the look further than its author intended
- Runs entirely in the browser — nothing is uploaded
## LUT Resizer
URL: https://www.rawcut.art/tools/lut-resizer
Accepts: .cube
Produces: .cube
Convert a 3D LUT between grid sizes when software demands a specific one. Reports the largest channel shift the resample actually caused, measured on your LUT, rather than implying the conversion was free.
- 17³, 25³, 33³, 41³, 49³ and 65³ grids
- Trilinear resampling matched to the preview shader
- Measured worst-case channel error, shown before you download
- Runs entirely in the browser — nothing is uploaded
## 3DL ↔ CUBE
URL: https://www.rawcut.art/tools/3dl-to-cube
Accepts: .3dl or .cube
Produces: .cube or .3dl
Convert between the two 3D LUT formats. They store their entries in opposite channel orders, so a naive converter returns a file that opens fine and renders red and blue swapped — this one transposes, and is tested against a hand-written reference file.
- Both directions, chosen from the file you drop
- Transposes between blue-fastest and red-fastest entry order
- Infers the .3dl bit depth, and says which one it read
- 10-, 12- and 16-bit output for .3dl
## LRTEMPLATE → XMP
URL: https://www.rawcut.art/tools/lrtemplate-to-xmp
Accepts: .lrtemplate
Produces: .xmp
Lightroom Classic stopped accepting .lrtemplate in 2018 and offers no way back, so preset packs bought before then are simply refused. This converts them to .xmp — faithfully, not approximately, because both formats name their settings identically.
- Exact conversion — the two formats share their setting names
- Tone curves, HSL, split toning and colour grading all carried
- Shows what the preset contains before you download
- Runs entirely in the browser — nothing is uploaded
## Batch Converter
URL: https://www.rawcut.art/tools/batch-convert
Accepts: .xmp, .lrtemplate, .cube, .3dl
Produces: .zip
Drop a folder's worth of .xmp, .lrtemplate, .cube or .3dl files and get them back as one archive. No limit, no account, no upload — the same converters as the single-file tools, run in a loop on your own machine.
- Unlimited files — no cap and no paid tier
- Any grid from 17³ to 65³, to .cube or .3dl
- Real per-file progress, and it says how long before starting
- Names anything it could not read instead of silently dropping it
---
# Guides
## What Is a LUT? Look-Up Tables in Photo & Video Explained
URL: https://www.rawcut.art/learn/what-is-a-lut
Updated: 2026-07-10
Tool for this: Convert to or from a LUT — https://www.rawcut.art/tools/convert
A LUT — short for look-up table — is one of the simplest and most powerful ideas in color work: a table that says, for every possible input color, exactly which output color to show instead. Feed it a flat, washed-out frame and a good LUT returns the same frame with cinematic contrast, shifted hues and a signature mood — instantly and identically every time.
Because a LUT is just a table of numbers, it works everywhere: photo editors, video editors like Premiere Pro and DaVinci Resolve, cameras, monitors, even game engines. That portability is why colorists trade looks as LUT files rather than as software-specific presets.
### How a LUT actually works
Imagine a giant address book for colors. An input color — say a muted red — is looked up by its red, green and blue values, and the table answers with the color that should replace it, perhaps a deeper, warmer red. No math happens at application time; the transformation was baked into the table when the LUT was created. That's why LUTs apply in real time even on modest hardware.
Since storing every one of the ~16.7 million RGB colors would be wasteful, real LUTs store a coarse grid — commonly 17, 33 or 65 points per channel — and interpolate smoothly between the stored points. A 33-point 3D LUT holds 35,937 entries, which is plenty for a smooth, accurate look.
### 1D vs. 3D LUTs
A 1D LUT maps each channel independently: red in, red out. That's perfect for contrast curves and gamma corrections, but it can never say “shift blue-greens toward teal” because each channel is blind to the others.
A 3D LUT maps the full color coordinate — the combination of R, G and B — to a new color. That's what enables true color grading: hue shifts, split-toned shadows, film-stock emulation. When people talk about creative LUTs, they almost always mean 3D LUTs stored as .cube files.
### When to use a LUT (and when not to)
LUTs shine when you need a consistent look across many images or clips, when you move a look between different applications, or when you want real-time preview on set. They are the lingua franca of color.
Their limitation: a LUT is spatially blind. It transforms colors, not regions — it cannot brighten just a face or darken just a sky. That's what masks and local adjustments are for, which is why a full editing recipe is more expressive than a LUT alone.
### Creating your own LUT
The classic workflow: grade a reference image, then export the transformation as a .cube file. In RawCut you can do this in one click — build a look in the Sandbox (or borrow any community recipe), press “.cube”, and the exact WebGL-rendered transformation is baked into a standard 33-point LUT that loads in Premiere, Resolve or Final Cut.
- Bake a Lightroom preset into a .cube LUT: https://www.rawcut.art/tools/xmp-to-cube
- Every conversion between presets and LUTs: https://www.rawcut.art/tools/convert
### Frequently asked questions
**Is a LUT the same as a filter or preset?**
No. A filter or preset stores editing instructions (slider values), which stay editable. A LUT stores the finished color transformation as a table. LUTs are more portable; presets are more flexible.
**Do LUTs reduce image quality?**
A well-made LUT with 33+ grid points introduces no visible quality loss. Very small LUTs (8 or 16 points) can show banding in smooth gradients.
**Can I use a video LUT on photos?**
Yes — a .cube LUT is just a color mapping and works identically for stills and footage, as long as it was designed for the same kind of source (e.g. standard-contrast images rather than log footage).
---
## What Is a .cube File? The 3D LUT Format Explained
URL: https://www.rawcut.art/learn/what-is-a-cube-file
Updated: 2026-08-05
Tool for this: Open a .cube and see what it does — https://www.rawcut.art/tools/cube-viewer
The .cube file is the de-facto standard container for 3D LUTs, originally specified by Adobe and IRIDAS. If you've ever downloaded a “cinematic LUT pack”, you almost certainly received a folder of .cube files.
Despite carrying sophisticated color science, a .cube file is astonishingly simple: it's plain text you can open in any editor. This page covers what is inside one, the format's actual rules, what the size number means, and how it differs from the other LUT formats you'll run into.
### Inside a .cube file
A .cube file starts with a tiny header and then lists color values, three floating-point numbers (red, green, blue, each 0–1) per line:
TITLE names the look. LUT_3D_SIZE declares the grid resolution — 33 means 33 steps per channel, so 33³ = 35,937 data lines follow. The lines run through every grid coordinate with red varying fastest, then green, then blue. DOMAIN_MIN and DOMAIN_MAX (usually 0 and 1) declare the input range.
- TITLE "My Cinematic Look"
- LUT_3D_SIZE 33
- DOMAIN_MIN 0.0 0.0 0.0
- DOMAIN_MAX 1.0 1.0 1.0
- 0.023529 0.043137 0.070588 ← the new color for pure black
- … 35,936 more lines …
### The .cube file format, precisely
The format is loose enough that most files parse anywhere, but a few rules matter when you are writing or debugging one. Lines beginning with # are comments. Blank lines are ignored. Keywords are uppercase and each appears at most once. Whitespace between values can be spaces or tabs, and quantity does not matter.
Only LUT_3D_SIZE is truly required — everything else has a default. Omit DOMAIN_MIN and DOMAIN_MAX and readers assume 0 to 1, which is right for almost every creative LUT. A file may legally declare a domain outside that range, which is how some log-footage transforms accept values above 1.0, but support for it is inconsistent enough that most tools quietly clamp.
The one thing you cannot get wrong is the ordering. Entries run red fastest, then green, then blue — so the second line is the output for one step of red with green and blue at zero, not one step of blue. Reverse it and the file still loads, still has the right number of lines, and renders with the red and blue axes swapped. It is the single most common way a hand-written LUT converter fails.
- LUT_3D_SIZE n — required; n between 2 and 65
- TITLE "…" — optional, ignored by most renderers
- DOMAIN_MIN / DOMAIN_MAX — optional, default 0 and 1
- LUT_1D_SIZE — a different, one-dimensional format; not interchangeable
- # — comment to end of line
### What the size number means, and why 33
A 3D LUT does not store an answer for every colour — there are 16.7 million in 8-bit alone. It stores answers on a regular lattice and interpolates between them. LUT_3D_SIZE is how many steps that lattice has along each axis, so the entry count is the cube of it.
33 became the default because it is what Adobe's documentation used in its examples and what most export dialogs offer first. It is also a sensible compromise: fine enough that creative looks show no banding, small enough that the file stays around a megabyte. 17 turns up in hardware — LUT boxes, on-camera monitors, game engines — where memory is tight. 65 is for transforms with abrupt behaviour, such as log-to-Rec.709 conversions or hard hue keys, where the space between 33 points is enough to lose the shape.
Larger is not better in itself. Resampling a 33³ LUT up to 65³ produces a bigger file describing exactly the same transform, because the new points are computed from the old ones rather than recovered from anywhere.
- 17³ = 4,913 entries — roughly 120 KB
- 33³ = 35,937 entries — roughly 900 KB
- 65³ = 274,625 entries — roughly 7 MB
- Resize a LUT between 17, 33 and 65 points: https://www.rawcut.art/tools/lut-resizer
### Where .cube files work
Practically everywhere color is graded: Adobe Premiere Pro and After Effects (Lumetri), DaVinci Resolve, Final Cut Pro, Photoshop (Color Lookup adjustment layer), Affinity Photo, OBS for live streams, and many camera monitors for on-set preview.
Notably absent: Lightroom, which uses its own profile and preset formats instead — one reason it's useful to have a look in both .xmp and .cube form.
### .cube, .3dl and .look — the other formats
A LUT you download might not be a .cube. The three you are most likely to meet hold comparable data and differ in how they write it down.
.3dl is Autodesk's, from Lustre and still used across Flame. It stores integers rather than decimals, does not record which bit depth those integers are on, and — the detail that catches people out — orders its entries with blue varying fastest, the exact opposite of .cube. Converting between the two requires transposing, not copying.
.look is Adobe SpeedGrade's, and it is not a lookup table at all: it is an XML document describing grading operations, closer in spirit to a preset than to a LUT. It cannot be converted to .cube by re-reading numbers; it has to be evaluated.
There is no .cube variant worth worrying about beyond this. The format has stayed stable for over a decade, which is most of why it won.
### Making one, and changing one
You can write a .cube by hand for anything simple, and people do — a two-point file that inverts an image is eight lines. For real looks, the practical routes are exporting from a grading application, or converting a preset you already have.
Modifying an existing LUT is where the plain-text format pays off. Reducing a LUT's strength, changing its grid size, or converting it to another format are all just arithmetic on the numbers, and none of them require the application that made it.
- Make a .cube from a Lightroom preset: https://www.rawcut.art/tools/xmp-to-cube
- Inspect what a .cube actually does: https://www.rawcut.art/tools/cube-viewer
### Frequently asked questions
**How do I open a .cube file?**
To inspect it, any text editor works — it is plain text, and the header tells you the size immediately. To use it, load it into your editor's LUT slot: Lumetri's Creative tab in Premiere, a Color Lookup adjustment layer in Photoshop, or a LUT node in Resolve. To see what it does to a photograph before committing, drop it into the RawCut CUBE viewer.
**What exactly is the .cube file format?**
A plain-text description of a three-dimensional colour lookup table. A short header declares the grid size and input range, then one line per lattice point gives the output colour as three decimals between 0 and 1. Entries are ordered with red varying fastest, then green, then blue. There is no binary section, no compression and no metadata beyond the title.
**Why does my .cube file have 35,937 lines?**
Because it is a 33-point LUT: 33 × 33 × 33 = 35,937, one line per lattice point. A 17-point file has 4,913 and a 65-point file has 274,625. If the line count is not a perfect cube of the declared LUT_3D_SIZE, the file is truncated or corrupt and most readers will reject it.
**Are .cube files safe to download?**
Yes — they are pure text data with no executable content. The worst a bad LUT can do is make your footage look ugly.
**Can I convert a Lightroom preset to a .cube LUT?**
Not in Lightroom itself. RawCut's XMP to CUBE tool does it: it renders the preset's global colour transformation through the same engine that draws the preview and bakes the result into a standard .cube. What cannot come across is anything spatial — masks, vignettes and grain have no representation in a LUT.
**Can I change a .cube file's size or strength?**
Yes, and neither needs the software that produced it. Strength is a blend between each entry and the input colour it came from, so a 50% version is arithmetic on the existing numbers. Size is a resample of the lattice. RawCut has a tool for each.
**Is a .cube file the same as a preset?**
No. A LUT is a finished transformation — a table of answers with no record of how they were reached. A preset is a set of instructions that remain editable. That is why a preset can be opened and learned from, and a LUT can be applied in software that has no idea what a shadow slider is.
---
## What Is an XMP File? Lightroom Sidecars & Presets Explained
URL: https://www.rawcut.art/learn/what-is-an-xmp-file
Updated: 2026-07-10
Tool for this: Open an .xmp and read what it contains — https://www.rawcut.art/studio
An .xmp file is a small text file that travels alongside a photo and carries everything about it that isn't pixels: editing instructions, ratings, keywords, copyright. XMP (Extensible Metadata Platform) is an Adobe-created, XML-based open standard.
For photographers, XMP matters most in two forms: the sidecar file that stores your Lightroom/Camera Raw edits next to a RAW file, and the preset file that packages a reusable look.
### Why sidecars exist: non-destructive editing
RAW files are treated as read-only negatives — an editor never alters them. Instead, every slider you move is written as an instruction: “Exposure +0.5, Shadows +30, this tone curve, that mask.” Those instructions live in a catalog or in an .xmp sidecar named like the photo (IMG_4021.CR3 → IMG_4021.xmp).
The payoff: edits are reversible forever, and sharing the tiny sidecar shares the entire edit. Anyone with the same RAW can reproduce — or remix — your grade exactly.
### What's actually inside
Open an .xmp in a text editor and you'll find XML with namespaced attributes. The interesting ones for editing carry the crs: prefix (Camera Raw Settings): crs:Exposure2012, crs:Contrast2012, tone curve point lists, per-color HSL adjustments, split toning and color grading values, and — since Lightroom's masking update — complete descriptions of radial, gradient, brush and AI masks with their local corrections.
This is exactly what RawCut parses in your browser: drop an .xmp next to a photo and the full recipe — every slider, curve and mask — becomes visible, learnable and exportable.
### XMP presets vs. sidecars
A preset .xmp is structurally the same file without a photo attached — just the settings you chose to include. Lightroom stores them in its Settings folder; importing one gives you a reusable starting look. Because presets store instructions rather than baked colors, they stay fully editable — the key difference from a LUT.
- Convert an .xmp into a LUT for video software: https://www.rawcut.art/tools/xmp-to-cube
### Frequently asked questions
**Can I delete .xmp files?**
If you delete a sidecar, the RAW file survives but the edits described in it are lost (unless they're also in a Lightroom catalog). Keep sidecars with their RAWs when archiving.
**How do I get Lightroom to write .xmp sidecars?**
Catalog Settings → Metadata → “Automatically write changes into XMP”, or select photos and press Ctrl/Cmd+S to write them on demand.
**Does an .xmp file contain the image itself?**
No — only metadata and instructions. That's why sidecars are a few kilobytes while RAW files are tens of megabytes.
---
## What Is a RAW Photo? RAW vs. JPEG Explained Simply
URL: https://www.rawcut.art/learn/what-is-a-raw-photo
Updated: 2026-07-10
Tool for this: Compare RAW vs. edit live — https://www.rawcut.art/studio
A RAW file is your camera's sensor reading, saved with (almost) no processing: no white balance, no contrast curve, no sharpening, no compression to 8-bit color. It's the digital negative — flat and unglamorous, but holding everything the sensor saw.
A JPEG, by contrast, is a finished print: the camera has already chosen the look and thrown away the data it didn't need. That single difference explains almost every practical trade-off between the two.
### Why RAW gives you editing superpowers
Bit depth: RAW files store 12–14 bits per channel — thousands of tonal steps where a JPEG keeps 256. Pull the shadows of a JPEG up two stops and you get banding and mud; do it to a RAW and detail simply appears.
White balance in RAW is a label, not a commitment — you can change 'tungsten' to 'daylight' after the fact with zero cost. Highlight recovery, deep shadow lifts, and aggressive color grading all survive because the data is still there.
### The costs
RAW files are 3–6× larger, need an editing step before sharing, and every camera brand has its own flavor — Canon .CR3, Sony .ARW, Nikon .NEF, Fujifilm .RAF — plus Adobe's universal .DNG. Straight out of camera they look flatter than JPEGs, because nothing has been done to them yet. That flatness is the point: it's headroom.
### RAW + XMP: the editable pair
Because RAW editing is non-destructive, your edits live as instructions in a catalog or an .xmp sidecar. Share the RAW plus its sidecar and someone else can open your exact edit, study it, and grade their own version — the workflow RawCut is built around: see the before, read the recipe, remix the look.
### When to shoot RAW (and when JPEG is fine)
Shoot RAW when light is difficult (sunsets, backlight, night), when you plan to edit seriously, or when the shot matters. JPEG remains sensible for high-volume casual shooting where storage and speed win. Many cameras offer RAW+JPEG so you don't have to choose in the moment.
### Frequently asked questions
**Can I open RAW files in my browser?**
Browsers can't decode RAW sensor data directly, but most RAW files embed a full-size JPEG preview. RawCut extracts that preview automatically when you drop a RAW into the Inspector.
**Is DNG better than my camera's native RAW?**
DNG is an open, documented format — great for archival and compatibility. Native formats are equally editable today; conversion is a workflow preference, not a quality change.
**Why do my RAW photos look worse than the JPEGs?**
They don't have less quality — they have less processing. The camera shows JPEGs with contrast, saturation and sharpening already applied. A RAW is waiting for you to make those choices.
---
## How to Edit a Photo Step by Step: The 7-Stage Workflow
URL: https://www.rawcut.art/learn/how-to-edit-a-photo-step-by-step
Updated: 2026-07-10
Tool for this: Watch a real edit build up — https://www.rawcut.art/community
Great edits aren't a pile of random slider moves — they follow an order. Professionals work from global to local and from tone to color, because each stage builds on a stable foundation laid by the previous one.
This is the same seven-stage sequence RawCut's Recipe Replay uses when it plays back a community edit layer by layer. Learn it once and every editor — Lightroom, Capture One, darktable — will feel familiar.
### 1 · White balance first
Set temperature and tint before anything else — every color judgement you make later depends on it. Aim for neutral first (whites look white), then bias deliberately: warmer for golden-hour intimacy, cooler for a clinical or moody feel.
### 2 · Exposure & tone
Set overall brightness (exposure), then shape the tonal extremes: recover highlights, open shadows, and anchor true white and black points with the whites/blacks sliders. Goal: no clipped detail you care about, and a histogram that uses the full range without slamming its edges.
### 3 · Presence: texture, clarity, dehaze, vibrance
These mid-level controls add (or soften) local contrast and color intensity. Use vibrance before saturation — it protects skin tones by boosting muted colors more than already-rich ones. Restraint here separates polished edits from over-cooked ones.
### 4 · The tone curve
The curve is fine tonal sculpting: a gentle S adds contrast; lifting the black point creates the faded 'matte film' look; per-channel red/green/blue curves are the classic way to color-cast shadows and highlights — the foundation of most film emulations.
### 5 · HSL / color mixer
Now steer individual color families: shift hues (orange skin tones, teal skies), tame or boost saturation per color, and lighten or darken by color (darker blues = dramatic skies). This is where a look's palette really forms.
### 6 · Color grading / split toning
Tint the shadows, midtones and highlights independently — the signature move of cinematic color. Teal shadows with warm highlights is the famous example, but subtler combinations define most professional looks.
### 7 · Local adjustments last
Finish with masks: brighten the subject, darken the sky with a gradient, dodge a face with a brush, let AI select the subject. Local work comes last because every global change would otherwise shift what your masks were balancing against.
### Frequently asked questions
**Do I have to follow this exact order?**
It's a guideline, not a law — but tone-before-color and global-before-local will save you from constantly redoing earlier work.
**How do I learn from other people's edits?**
Study real recipes, not just finished images. RawCut shows every post's full editing recipe and can replay it stage by stage so you see what each layer contributed.
**How much editing is too much?**
A useful test: step away for ten minutes, then look at the before/after slider. If the edit fights the light that was actually there, pull it back.
---
## LUT vs. Preset: What's the Difference and Which Should You Use?
URL: https://www.rawcut.art/learn/lut-vs-preset
Updated: 2026-07-10
Tool for this: Get both from one look — https://www.rawcut.art/studio
“LUT” and “preset” get used interchangeably in tutorials and marketplaces, but they are fundamentally different objects — and choosing the wrong one for the job causes real frustration.
The one-line version: a preset is a recipe (editable instructions), a LUT is a finished sauce (a baked transformation).
### What each one stores
A preset (like a Lightroom .xmp) stores slider positions: exposure +0.3, shadows +25, this curve, these HSL shifts, even masks. Apply it and all those controls remain live — you can see them, adjust them, learn from them.
A LUT (.cube) stores only the end result as a color table: this input color becomes that output color. Compact, fast, universal — and opaque. You can't open a LUT and see 'how much contrast' it adds; that information was consumed in the baking.
### Strengths, honestly compared
Each format wins in its natural habitat:
- Editability: preset ✓ — every slider stays adjustable; LUT ✗
- Works in video editors: LUT ✓ — Premiere, Resolve, FCP; preset ✗
- Local adjustments/masks: preset ✓; LUT ✗ (color-only, spatially blind)
- Real-time on set / in OBS: LUT ✓
- Learning how a look was built: preset ✓ — it is the documentation
- Adapting to different exposures: preset (semi) — a LUT applies identically even when the shot needs different treatment
### Use both: recipe → LUT
The professional pattern is a pipeline, not a choice: develop the look as an editable recipe, then export a LUT for the places that need one (video timelines, streaming, on-set monitors). RawCut automates the pair — every look can leave as an .xmp preset and as a baked .cube LUT of the same grade.
### Frequently asked questions
**Why does the same LUT look wrong on some shots?**
A LUT assumes its input resembles the footage it was designed on. Under- or over-exposed shots feed it different colors, so it answers with a different look. Normalize exposure and white balance first.
**Can a LUT contain masks or grain?**
No. A LUT transforms colors one at a time with no knowledge of position or texture. Vignettes, grain and dodging live in presets or manual edits.
**Which should I sell or share as a creator?**
Both, ideally: presets for photographers who want to learn and tweak, LUTs for videographers and streamers. They serve different audiences with the same look.
---
## What Is EXIF Data? Photo Metadata Explained (and Why It Matters)
URL: https://www.rawcut.art/learn/what-is-exif-data
Updated: 2026-07-10
Tool for this: Read any photo's EXIF instantly — https://www.rawcut.art/tools/metadata-reader
Every digital photo carries a hidden logbook: EXIF metadata, written by the camera at the moment of exposure. It records the gear and settings — camera model, lens, aperture, shutter speed, ISO, focal length — and often the time and GPS position too.
For learning photography, EXIF is gold: it answers 'how was this shot taken?' with data instead of guesses.
### What EXIF typically contains
The most useful fields, roughly in order of interest:
- Camera make & model, and the exact lens used
- Exposure triangle: aperture (f-stop), shutter speed, ISO
- Focal length (and 35mm-equivalent, revealing crop factor)
- Date & time of capture
- GPS coordinates, if location tagging was on
- Flash, metering and exposure-compensation settings
### Reading EXIF to learn
Seeing that a dreamy portrait was f/1.4 at 1/250s, or that a silky waterfall was a 2-second exposure at ISO 100, turns inspiration into a reproducible plan. This is why RawCut reads EXIF automatically from every upload and organizes posts into gear hubs — you can browse everything shot on your exact camera or lens and study the numbers behind each look.
### Privacy: what you're sharing without noticing
GPS coordinates in EXIF can reveal your home or a subject's location. Big social platforms strip metadata on upload, but files you send directly (email, cloud drives, some messengers) usually keep it. Before publishing originals, check whether location data should be removed — most editors offer 'export without location' options.
### Frequently asked questions
**How do I view a photo's EXIF data?**
On desktop: file properties/inspector, or any photo editor's Info panel. In RawCut, drop the photo into the Inspector — the EXIF dashboard appears instantly, parsed in your browser.
**Does editing destroy EXIF?**
Normally no — editors carry it into exports unless you strip it deliberately. Screenshots, however, contain none of the original EXIF.
**Can EXIF be faked?**
Yes, metadata is editable with the right tools, so treat it as helpful context rather than forensic proof.
---
## The Teal & Orange Look: Why Hollywood Loves It and How to Create It
URL: https://www.rawcut.art/learn/teal-and-orange-look
Updated: 2026-07-10
Tool for this: Try the look on your photo — https://www.rawcut.art/studio
Watch any blockbuster and you'll see it: warm, golden skin tones set against cool teal shadows and backgrounds. Teal & orange is the most recognizable color grade in modern cinema — and one of the most requested looks in photography.
It isn't a fashion accident. The look exploits a real property of color perception, which is why it survives every trend cycle.
### The color theory
Teal and orange sit opposite each other on the color wheel — they are complementary. Complementary pairs create the strongest simultaneous contrast the eye can perceive, so subjects graded warm literally pop from cool surroundings.
The crucial detail: human skin, regardless of ethnicity, lives in the orange range of the spectrum. Push everything else toward teal and you get maximum separation while skin stays natural — the whole trick in one sentence.
### Building the look, stage by stage
A restrained teal & orange grade uses several small moves rather than one big one:
- White balance: keep skin honest — don't just crank warmth globally
- Tone curve: lift the blue channel's shadows (adds teal), ease its highlights (adds warmth)
- HSL: shift blues/aquas toward teal; nudge orange luminance up so faces glow
- Split toning / color grading: teal (~hue 200–215) into shadows, warm gold (~hue 40–50) into highlights
- Desaturate competing greens and yellows so the two-color story stays clean
### Why it often goes wrong
The failure mode is intensity: radioactive skin, ink-black teal shadows, and every color hammered into two buckets. Strong looks are built from many gentle moves — check the recipes of well-graded community shots and you'll consistently find split-toning saturations around 10–25, not 80.
### Frequently asked questions
**Does teal & orange work for landscapes without people?**
Yes — golden light plus cool shadows reads as 'cinematic dawn/dusk'. Without skin to protect you can push the warm tones further.
**Is there a one-click teal & orange preset?**
Plenty exist — including looks you can borrow on RawCut — but exposure and white balance differ per photo, so expect to fine-tune after applying (RawCut's Auto-fit does this adjustment for you).
**LUT or preset for this look?**
Learn it as a preset/recipe first (you can see the moves), then export a LUT for video work. Same look, two containers.
---
## How to Install and Use LUTs in DaVinci Resolve (Step by Step)
URL: https://www.rawcut.art/learn/how-to-install-luts-in-davinci-resolve
Updated: 2026-08-05
Tool for this: Convert a preset into a LUT — https://www.rawcut.art/tools/xmp-to-cube
Installing a LUT in DaVinci Resolve takes about thirty seconds. Getting it to look the way it looked in the vendor's preview takes understanding one thing: a LUT is a function, and a function is only correct for the input it was written for.
This guide covers the mechanics first — where the folder lives on each platform, how to make Resolve rescan it — then the part that actually causes the frustration, which is input colour space, and finally the three things people ask next: applying a look to every clip, exporting your own, and what the free version can do.
### Where to put .cube files
Resolve reads LUTs from a fixed directory tree. You can drop files directly in the root, but creating a subfolder per pack keeps the browser navigable once you have more than a handful.
After copying files in, Resolve does not notice automatically. Open Project Settings → Color Management and press Update Lists, or restart the application. If a LUT does not appear after that, the file is almost certainly malformed rather than misplaced.
- macOS — /Library/Application Support/Blackmagic Design/DaVinci Resolve/LUT
- Windows — C:\ProgramData\Blackmagic Design\DaVinci Resolve\Support\LUT
- Linux — /opt/resolve/LUT (or /home/resolve/LUT on some installs)
- Check whether a .cube file actually parses: https://www.rawcut.art/tools/cube-viewer
### Applying a LUT on a node vs. in Color Management
There are two places a LUT can live, and they mean different things. Right-clicking a node and choosing a LUT applies it at that point in your grade — it is one operation among many, and you can grade before and after it.
Setting a LUT in Project Settings → Color Management applies it to everything as an output transform, after your grade. That is the correct home for a display or conversion LUT, and the wrong home for a creative look you want to grade underneath.
The practical rule: creative looks go on a node near the end of the chain. Technical transforms — log to Rec.709, camera to working space — belong in Color Management, or on a dedicated node before any creative work.
### Why your LUT looks wrong
This is the real question behind most LUT problems. A creative LUT is authored against a specific input: Rec.709, or a log profile like S-Log3, V-Log or C-Log3. Feed it something else and the maths is still applied faithfully — to the wrong numbers.
Log footage through a Rec.709 look comes out flat and milky, because the LUT expected contrast that was never there. Rec.709 footage through a log-designed look comes out crushed and over-saturated, for the mirror-image reason.
The fix is ordering, not strength. Convert to the space the LUT expects first — with a Color Space Transform node or the project's colour management — and apply the creative LUT after that. Reaching for the mix slider to tame a look that is being fed the wrong input treats the symptom.
### Dialling in strength
Once the input is right, the Key palette's Output Gain controls how much of the node's effect survives — the standard way to run a look at 60% rather than 100%.
A LUT applied at full strength on top of an already-graded image is usually too much, because both are doing contrast. Grade to a neutral, correct starting point first; the look then has room to be a look rather than a second correction.
Output Gain is the right answer inside Resolve, with one limitation worth knowing: it lives in the project file. Send the .cube to a colleague, load it on a camera monitor or point OBS at it and the LUT arrives at full strength again, because the number that tamed it was never in the file. Baking the reduction into the LUT itself is the version that travels.
- Bake a LUT to 40% so the strength travels with it: https://www.rawcut.art/tools/lut-strength
### How to add a LUT to all clips at once
Applying a look clip by clip is the slow way, and there are three faster ones depending on what you actually mean.
If every clip should get the same grade, grade one properly and then right-click it and choose Apply Grade to All Clips — or select the rest and use Ripple Node Changes, which pushes edits from the current clip out to the whole selection. Both copy the grade, so later changes need repeating.
If the look belongs to the sequence rather than the shots, use a Timeline-level node instead. Switch the node editor from Clip to Timeline in the Color page and apply the LUT there: it sits under everything, applies once, and stays editable in one place. This is the right home for a delivery or display transform.
### Creating and saving your own LUT in Resolve
Any grade you build can leave as a .cube. Right-click the clip's thumbnail in the Color page and choose Generate 3D LUT, pick a size (33 Point is the default and the right answer), and Resolve writes the file. Save it into the LUT folder listed above and it appears in the browser alongside everything else after an Update Lists.
One caveat that catches people: Generate 3D LUT bakes the node graph as a colour transformation only. Anything spatial — power windows, tracked masks, blurs, grain — is not in the output, because a LUT has no concept of where a pixel is. If the grade depends on a window, the exported LUT will not look like the grade.
Coming from stills instead, you can convert a Lightroom preset directly. RawCut bakes the .xmp through a real colour pipeline and writes a 33³ .cube, so the LUT matches what the preset does rather than approximating it.
- Convert a Lightroom .xmp preset into a .cube LUT: https://www.rawcut.art/tools/xmp-to-cube
### Do LUTs work in the free version of DaVinci Resolve?
Yes, completely. LUT import, node-level application, Timeline nodes, Color Management and Generate 3D LUT are all in the free version, and nothing in this guide requires Resolve Studio.
What Studio adds is elsewhere: the Neural Engine features, temporal noise reduction, HDR grading tools, more than UHD output and hardware decoding options. None of them change how LUTs behave. A creative look that works in Studio works identically in the free build.
### Frequently asked questions
**How do I adjust LUT strength in DaVinci Resolve?**
Resolve has no strength slider on the LUT itself. Apply the LUT on its own node, then lower that node's Key Output Gain in the Key palette: at 0.5 the look is mixed half strength over the ungraded image. The alternative is a Layer Mixer node with the LUT on the upper layer and its opacity dialled back. Both approaches keep the original grade intact underneath.
**How do I add a LUT to all clips at once in DaVinci Resolve?**
Grade one clip, then right-click it and choose Apply Grade to All Clips, or select every clip in the timeline and use Ripple Node Changes. For a look that should sit under everything on the timeline, put the LUT on a Timeline-level node in the Color page instead, which applies once to the entire sequence rather than per clip.
**What is the difference between importing and applying a LUT?**
Importing copies the .cube file into Resolve's LUT folder so it appears in the LUT browser; it changes nothing on screen. Applying assigns an imported LUT to a specific node, clip or timeline. A LUT that is imported but never applied is the usual reason people report that 'nothing happened'.
**Why doesn't my LUT show up in DaVinci Resolve?**
Either Resolve has not rescanned the folder — press Update Lists in Project Settings → Color Management, or restart — or the file is not a valid 3D LUT. Resolve silently ignores malformed files, which is why a bad .cube looks identical to a missing one. RawCut's CUBE Viewer parses the file in the browser and shows you its response curve, so you can tell the two apart in seconds.
**Should I apply the LUT before or after grading?**
Creative looks go after your primary correction, near the end of the node chain, so you are grading a neutral image rather than fighting the look. Technical conversion LUTs go first, or in Color Management. Getting this order backwards is the second most common cause of a LUT looking wrong.
**Can I use Lightroom presets in DaVinci Resolve?**
Not directly — Resolve does not read .xmp files. Convert the preset into a .cube LUT first and Resolve will treat it like any other LUT. Global settings such as exposure, tone curves, HSL and colour grading survive the conversion; masks and local adjustments cannot, because a LUT has no spatial component.
**Does the free version of Resolve support LUTs?**
Yes. LUT import, node-level application and Color Management are all in the free version. Nothing in this guide requires Resolve Studio.
**What LUT size should I use?**
33×33×33 is the standard and is what most software and hardware expect. 17³ is smaller and slightly less precise in steep gradients; 65³ is more precise but rarely distinguishable in practice and slower to load on monitors and cameras.
---
## Colour Grading vs. Colour Correction: The Difference That Matters
URL: https://www.rawcut.art/learn/color-grading-vs-color-correction
Updated: 2026-07-26
Tool for this: See the two passes separated — https://www.rawcut.art/studio
The two terms get used interchangeably, and the confusion is expensive: it is the reason a preset that looked incredible on the photo it was built for looks wrong on everything else.
Correction and grading are different jobs with different success criteria. Correction has a right answer. Grading does not.
### Correction: making the image accurate
Colour correction is the technical pass. White balance so neutrals are neutral, exposure so the image sits where it should, contrast so the tonal range fills the frame without clipping, and any per-shot fixes for lens or sensor behaviour.
Correction is objective in a way grading is not. A grey card should read grey. Skin should sit in a plausible hue range. Blown highlights are a defect, not a choice. Two competent people correcting the same frame will land in nearly the same place.
### Grading: making the image feel like something
Grading is the creative pass that happens after. Cooling shadows and warming highlights, pulling saturation out of everything except one hue, lifting blacks for a matte finish, pushing greens toward olive so foliage stops shouting.
There is no correct answer here, only intent. The same corrected frame can plausibly become a warm nostalgic still or a cold clinical one, and neither is wrong.
### Why the order is not negotiable
Grading assumes a known starting point. A teal-and-orange grade works by pushing shadows cool and highlights warm — but 'shadows' and 'highlights' are defined by where the tones currently sit. Apply that grade to an under-exposed, magenta-cast frame and it operates on the wrong regions entirely.
This is the whole reason presets travel badly. A preset is a grade, and it silently assumes the correction has already happened. Applied to a raw, uncorrected file it produces something unpredictable, and the usual response — nudging the preset's own sliders — is treating the symptom.
Correct first, then grade. On a corrected frame the same preset behaves consistently across a whole shoot.
- 1. White balance — neutrals neutral
- 2. Exposure — brightness where it belongs
- 3. Contrast and tone — highlights and shadows recovered
- 4. — correction ends, grading begins —
- 5. HSL — per-hue shaping
- 6. Split toning / colour grading wheels
- 7. Local adjustments and finishing
### How to tell which one you are doing
A useful test: could someone else look at your change and call it objectively wrong? If yes, you are correcting. If the only available verdict is 'I don't like it', you are grading.
Another: does the adjustment need to change per photo in a set? Corrections almost always do, because exposure and white balance vary shot to shot. A grade should be able to stay fixed across a set — that consistency is what makes it a look rather than a fix.
### Frequently asked questions
**Is colour correction or colour grading more important?**
Correction, without much doubt. A well-corrected image with no grade looks clean and professional. A heavily graded image on top of a bad correction looks broken, and the grade usually gets blamed for it.
**Can one preset do both?**
Most presets try, which is exactly why they are inconsistent. A preset that includes an exposure or white balance offset is baking in a correction for one specific photo. Presets that stick to grading moves — HSL, split toning, curves — travel far better.
**Do I need to correct RAW files if they already look flat?**
Yes. Flat is not the same as neutral. A log or flat RAW rendering still has a white balance and an exposure that need setting; flatness only describes the contrast curve.
**Where do tone curves fit?**
Both, depending on intent. A gentle S-curve restoring contrast is correction. A lifted-black matte curve is a grade. The same tool, two jobs — which is one reason the boundary confuses people.
---
## What Is a Camera Profile? DCP, Adobe Standard and Camera Matching
URL: https://www.rawcut.art/learn/what-is-a-camera-profile
Updated: 2026-07-26
Tool for this: Read a sidecar's profile — https://www.rawcut.art/tools/metadata-reader
A RAW file contains sensor readings, not colours. Something has to decide that a particular set of readings means 'skin' rather than 'brick', and that something is the camera profile.
It is the most consequential setting most photographers never change — and it sits above every adjustment in the edit, which is why changing it late can undo work you have already done.
### What a profile actually contains
A camera profile is a colour transform: a matrix mapping the sensor's native response to a standard colour space, usually with a lookup table for hue and saturation refinements the matrix alone cannot express, plus a tone curve.
Adobe's format is the DCP (DNG Camera Profile). Every supported camera ships with several, because there is no single objectively correct mapping — a sensor's response does not match human vision, and closing that gap involves judgement.
### Adobe Colour, Adobe Standard and Camera Matching
Adobe Colour is the modern default: a moderately contrasty, slightly warm rendering designed to look good across every supported body. Adobe Standard is the older, flatter, more neutral option, and is still the better starting point for heavy grading precisely because it imposes less.
Camera Matching profiles are Adobe's reconstruction of your manufacturer's own JPEG rendering — Camera Standard, Camera Vivid, Camera Portrait on Canon and Nikon, or the film-simulation names on Fujifilm. If your RAW file looks worse than the preview did on the camera's screen, this is why: the camera showed you a JPEG rendered with its own profile, and Lightroom opened the RAW with Adobe's.
- Adobe Colour — contrasty, warm, the current default
- Adobe Standard — flat and neutral, best base for grading
- Camera Matching — mimics the manufacturer's JPEG look
- Creative profiles — looks, not colour science
### Profiles are not presets
This distinction matters more than it sounds. A profile operates on the raw sensor data before any slider is applied; a preset operates on the result. Changing profile therefore shifts the ground under every adjustment you have already made.
The practical consequence: choose the profile first, before correcting or grading. Switching from Adobe Colour to Adobe Standard halfway through an edit will change how your existing white balance and HSL moves read, and you will end up redoing them.
This is also why profile choice does not travel in a preset in the way people expect. A preset can specify a profile, but if the recipient's camera has no matching profile the setting is silently ignored — a common reason a shared preset looks different on someone else's files.
### Which one should you use
For heavy grading, Adobe Standard. Its flatter starting point leaves more room and imposes fewer decisions you then have to undo.
For fast, natural-looking edits, Adobe Colour or the Camera Matching profile for your body — especially with Fujifilm, whose film simulations are genuinely good and hard to reproduce with sliders.
For consistency across a shoot, any of them, as long as it is the same one. Mixed profiles across a set is a subtle but real cause of edits that will not match no matter how carefully you copy settings between frames.
### Frequently asked questions
**What's the difference between a camera profile and a preset?**
A profile interprets raw sensor data into colour before any editing; a preset is a set of editing adjustments applied afterwards. The profile sits underneath everything, which is why changing it late in an edit shifts the meaning of adjustments you have already made.
**Why does my RAW look worse than the camera's preview?**
The preview was a JPEG rendered with your manufacturer's profile. Your editor opened the RAW with Adobe's. Switching to the Camera Matching profile for your body usually gets you back to what you saw on the screen.
**Does a profile get saved in the .xmp sidecar?**
The profile's name and identifier do, so the sidecar records which profile was used. The profile data itself does not — it lives in the editor's installation. If the recipient does not have that profile, the setting is ignored.
**Do camera profiles apply to JPEGs?**
No, or at least not meaningfully. A JPEG has already been rendered to a standard colour space by the camera, so the interpretation step is done. Profiles matter for RAW files, where that decision is still open.
**Are creative profiles the same as camera profiles?**
They share a container but do different jobs. Creative profiles apply a look on top of a correct colour interpretation — they are effectively presets in profile clothing, and they can be dialled back with an amount slider, which true camera profiles cannot.
---
## How to Match Colours Between Two Photos (Without Guessing)
URL: https://www.rawcut.art/learn/how-to-match-colors-between-photos
Updated: 2026-07-26
Tool for this: Measure instead of guessing — https://www.rawcut.art/studio
Matching two photos by eye is unreliable, because eyes adapt. Look at a warm frame for thirty seconds and it stops looking warm — which is precisely when you make the wrong call.
The fix is to work in a fixed order and lean on measurements at each step. The order matters because each stage changes what the next one sees.
### Pick the reference first
Decide which frame is correct and do not touch it. Matching two images to each other simultaneously means chasing a moving target, and you will converge on something neither of you wanted.
Usually the reference is whichever frame has the best light. If neither does, correct one properly first and make that the reference.
### Match exposure before anything else
Find a region present in both frames that should be the same brightness — a patch of sky, a wall, a subject's forehead — and match its luminance. Everything downstream depends on this, because contrast and colour perception both shift with brightness.
Use a measurement rather than judgement. A pixel readout or a histogram tells you the two patches read 148 and 172 far more reliably than your eye, which is busy adapting.
### Then white balance
Find something that should be neutral in both frames and make it neutral in both. If R, G and B read within a few points of each other on a grey patch, the white balance is right regardless of how the frame feels.
If there is no neutral object, use skin, which occupies a narrow and well-documented hue range. Skin that reads too magenta or too green is a white balance error, not a grading choice.
### Then contrast, then hue
With exposure and white balance matched, compare the two histograms. If one has crushed blacks and the other does not, that is a black-point difference, not a colour one — fix it with blacks and shadows rather than by chasing colour.
Only now address hue. At this point most of the perceived mismatch will already be gone, and what remains is usually one or two HSL bands — often greens, which vary enormously between lighting conditions and are the most common residual offender.
- 1. Exposure — match a common region's luminance
- 2. White balance — make neutrals neutral in both
- 3. Black point and contrast — match the histogram shape
- 4. HSL — fix the one or two bands still visibly off
- 5. Only then, a shared grade on top
### Matching a whole set
For a shoot, correct every frame to neutral independently, then apply one shared grade across all of them. That is the reverse of the instinct — which is to find a look on one frame and paste it everywhere — and it is why the instinct produces sets that almost match.
The shared grade is only consistent if what it lands on is consistent. Correction is per-frame; grading is per-set.
### Frequently asked questions
**Why do my photos look matched on my screen but not on my phone?**
Screen calibration, and often screen white point. If your monitor runs warm you will correct against it and everything will read cool everywhere else. Checking on a second device is a rough but genuinely useful sanity check.
**Can I just copy settings from one photo to another?**
Only for grading moves. Copying exposure and white balance across frames shot in different light produces exactly the mismatch you were trying to fix — those are per-frame corrections. Copy the HSL, curves and split toning; leave the corrections alone.
**What if the two photos were shot in completely different light?**
Accept a partial match. Golden hour and overcast midday cannot be made identical without a grade so heavy it damages both. Match exposure and neutrals, get the skin tones consistent, and let the ambient difference stand — it usually reads as intentional.
**Is there an automatic way to match two photos?**
Automatic matching solves the global part — exposure, white balance, overall saturation — by measuring both images and fitting the difference. It is a good starting point, and it is measurement rather than guesswork, but per-hue work usually still needs a person.
---
## What Is a CR3 File? Canon's RAW Format Explained
URL: https://www.rawcut.art/learn/what-is-a-cr3-file
Updated: 2026-08-05
Tool for this: Read a CR3's metadata — https://www.rawcut.art/tools/metadata-reader
A .CR3 file is the RAW image a modern Canon camera writes: unprocessed sensor data plus the settings the camera was using, saved instead of — or alongside — a JPEG.
It replaced the older .CR2 format in 2018, and the replacement was not cosmetic. CR3 is built on a different container and a different compression codec, which is why older software can open a CR2 from 2010 and refuse a CR3 from a newer body.
### What changed from CR2 to CR3
CR2 was a TIFF file underneath — the same structure as a layered image, with the RAW data as one of its parts. CR3 abandoned that for the ISO Base Media File Format, the container family that also holds MP4 and HEIF. Inside it, the image data is compressed with Canon's CRX codec rather than the older lossless scheme.
This is why support arrived slowly. A CR2 reader could lean on existing TIFF handling; a CR3 reader had to be written more or less from scratch, and for several months after the format appeared the only software that could open one was Canon's own.
Practically, CR3 means smaller files at the same quality, faster writes, and a format that no longer resembles anything else on your drive.
### RAW and C-RAW: the choice that matters
Canon bodies writing CR3 offer two settings, and the menu does not make the difference obvious. Plain RAW is lossless — the sensor data is compressed the way a ZIP compresses text, and decompresses to exactly what was captured. C-RAW is lossy, producing files roughly 40% smaller by discarding information it judges you will not see.
For most photographs, most of the time, C-RAW is genuinely hard to distinguish. Where it shows is under heavy manipulation: large shadow recoveries, aggressive local contrast, and big crops are where the discarded data would have been useful. Its artefacts appear in smooth gradients and around high-contrast edges.
The decision is about headroom, not about how the file looks straight out of camera. If a shoot is going to be pushed hard in post, or archived for a decade of unknown future use, the extra 40% is cheap insurance. For high-volume work that is graded lightly, C-RAW is a reasonable default.
### What is inside the file
Beyond the sensor data, a CR3 carries a full-size JPEG preview — which is what your camera's playback screen and your operating system's thumbnailer are actually showing you — along with EXIF metadata, the lens correction data Canon's own software uses, and the Picture Style that was active.
That last point causes a recurring surprise: the preview has the Picture Style baked in, the RAW data does not. Opening the file in a third-party converter shows you a flatter image than the camera showed you, and nothing is wrong. You are seeing the data rather than Canon's rendering of it.
Bodies with Dual Pixel RAW enabled store a second, slightly offset image alongside the first, roughly doubling file size in exchange for limited after-the-fact focus and bokeh adjustment in Canon's software.
### Opening a CR3
Current versions of Lightroom, Camera Raw, Capture One, DxO PhotoLab, Affinity Photo and DaVinci Resolve all read CR3, as does Canon's free Digital Photo Professional. Older releases do not, and no update to the file will help — the reader is what is missing.
The usual workaround for stubborn software is Adobe's free DNG Converter, which rewrites the CR3 as a DNG that almost everything can read. This is lossless for plain RAW files and preserves the original data.
### Frequently asked questions
**Is CR3 better than CR2?**
As a container, yes: smaller files for the same image quality and faster writing. As image data, they are equivalent at the same bit depth — a CR3 shot in lossless RAW holds what a CR2 would have held. The practical downside of CR3 is compatibility with older software, not quality.
**Should I shoot RAW or C-RAW?**
C-RAW when storage or buffer depth matters and the images will be graded normally; RAW when they will be pushed hard, heavily cropped, or archived. C-RAW discards data permanently, so the choice cannot be revisited later. If you are unsure on a particular shoot, the safe answer is RAW.
**Why does my CR3 look flat compared to the camera screen?**
The camera screen shows the embedded JPEG preview, which has the active Picture Style applied. The RAW data has no style applied at all — that is the point of it. Any converter will start you from the unrendered data, and applying a matching camera profile gets you back to the look you saw.
**How do I convert a CR3 to JPEG?**
Any RAW converter will export one: Lightroom, Camera Raw, Capture One, or Canon's free Digital Photo Professional. Avoid tools that merely extract the embedded preview — that gives you the camera's JPEG rather than a conversion of the RAW data, at whatever size the preview happens to be.
**Can I recover a CR3 that software says is corrupt?**
Sometimes. Because a CR3 carries a full-size embedded JPEG, a file whose RAW payload is damaged may still yield a usable image if the preview survived. That is a salvage operation, not a repair — the RAW data itself is not recoverable once lost.
**Which cameras produce CR3 files?**
Canon bodies with a DIGIC 8 processor or newer, beginning with the EOS M50 in 2018 and covering the EOS R line and recent DSLRs. Anything older writes CR2.
---
## What Is an ARW File? Sony's RAW Format Explained
URL: https://www.rawcut.art/learn/what-is-an-arw-file
Updated: 2026-08-05
Tool for this: Inspect an ARW — https://www.rawcut.art/tools/metadata-reader
An .ARW file is Sony Alpha RAW — the unprocessed sensor data from a Sony camera, stored in a structure derived from TIFF/EP, the same standard several other manufacturers build on.
As a container it is unremarkable. What makes ARW worth reading about is that Sony gives you more compression choices than most manufacturers, and that the camera sometimes changes bit depth without telling you.
### Three kinds of ARW
Recent Sony bodies offer uncompressed, lossless compressed and compressed RAW, and they are genuinely different files.
Uncompressed is the largest and simplest: every sample written out at full precision. Lossless compressed applies reversible compression, producing files roughly 40–60% smaller that decode to exactly the same data — there is no image-quality reason to prefer uncompressed over it, and newer bodies offer size variants of it too.
Compressed RAW is the lossy option, and it is the one with a reputation. Sony's scheme groups neighbouring pixels and stores them with a shared exponent and reduced precision, which is efficient and almost always invisible — but around very high-contrast edges it can produce faint banding or posterisation that survives into the final image. Astrophotographers and anyone recovering deep shadows next to bright highlights are the people who notice.
- Uncompressed — largest, no compression artefacts, no advantage over lossless
- Lossless compressed — the sensible default on bodies that offer it
- Compressed — smallest, lossy, occasional edge artefacts under heavy recovery
### The 14-bit to 12-bit drop
This is the ARW behaviour that surprises people, because nothing on screen announces it. On many Sony bodies, certain modes reduce RAW bit depth from 14 to 12: continuous shooting at high frame rates, silent/electronic shutter on some models, long-exposure noise reduction, and bulb mode among them. Which modes trigger it varies by body and firmware.
Twelve bits is 4,096 levels per channel against 14 bits' 16,384. In a normally exposed frame you will not see the difference. In a deep shadow lift of several stops — exactly what you do with an underexposed frame — the coarser steps show up as banding in smooth areas.
The practical response is not to avoid those modes but to know when you are in them, and to expose so that shadows do not need rescuing when you are.
### Working with ARW files
Support is broad: Lightroom, Camera Raw, Capture One (which has a particularly close relationship with Sony files), DxO PhotoLab, Affinity Photo, RawTherapee and darktable all read ARW. Sony's own Imaging Edge is free and handles the newest bodies first.
The recurring compatibility problem is not the format but the body. Each new sensor needs its own profile and decoding parameters, so a camera released last month may not open in software that is otherwise current. Waiting for a point release, or converting to DNG in the meantime, are the two options.
ARW files carry the usual EXIF, plus Sony's Creative Style and any in-camera corrections as metadata that third-party converters may or may not choose to apply.
### Frequently asked questions
**Which ARW compression should I use?**
Lossless compressed, on any body that offers it. It is meaningfully smaller than uncompressed and decodes to identical data, so uncompressed buys nothing. Choose lossy compressed only when card space or buffer depth genuinely constrains the shoot.
**Is Sony's compressed RAW actually a problem?**
Rarely, and only in specific conditions: very high-contrast edges combined with heavy exposure recovery. Ordinary photographs graded normally show nothing. The reason it has a reputation is that the situations where it does show — night skies, backlit subjects pulled up several stops — are the ones people scrutinise most closely.
**How do I tell whether a file is 12-bit or 14-bit?**
The camera does not display it, but a metadata reader will show the bit depth recorded in the file. If it reads 12 and you expected 14, check whether you were in continuous drive, silent shutter, bulb or long-exposure NR — those are the usual causes.
**Why won't my software open a new Sony camera's ARW files?**
Because ARW is not one format but a family, and each new sensor needs decoding support added. Update your converter; if the body is very new, use Adobe's free DNG Converter as a stopgap, which produces a DNG almost anything can read.
**Are ARW and A7R IV/A1 files the same format?**
Same extension, different internals. Sony has revised the ARW structure several times, and newer bodies also introduced additional lossless-compressed size options. This is why extension alone never guarantees a file will open.
---
## What Is a NEF File? Nikon's RAW Format Explained
URL: https://www.rawcut.art/learn/what-is-a-nef-file
Updated: 2026-08-05
Tool for this: Inspect a NEF — https://www.rawcut.art/tools/metadata-reader
A .NEF file is Nikon Electronic Format: the sensor data from a Nikon camera together with the settings and metadata that describe how it was captured.
It is one of the older RAW formats still in active use, built on TIFF, and Nikon has extended it rather than replaced it. That continuity is a real advantage — but it means a single extension now covers files with quite different compression inside.
### Compression: lossless, lossy, and none
Nikon's menu has historically offered three settings, worded confusingly. Uncompressed writes everything. Lossless compressed reduces size reversibly, decoding to exactly the original data. Compressed — sometimes labelled simply that — is lossy: it applies a curve that allocates fewer levels to the highlights, where human vision is least sensitive to steps.
The lossy option is more subtle than its name suggests. Because it discards precision specifically where the eye is least able to detect it, the visible cost is small even under moderate manipulation. It becomes visible in the same place all lossy RAW does: extreme highlight or shadow recovery, where the missing levels turn a smooth gradient into steps.
As with every manufacturer, lossless compressed is the sensible default. It is strictly smaller than uncompressed with no cost at all.
### High Efficiency: HE and HE*
Recent Nikon mirrorless bodies, beginning with the Z 9, added two further options that behave differently from anything above. HE* aims to match lossless-compressed quality at around a third of the size; HE goes smaller still.
These are not simply better compression of the same data — they use a fundamentally different, more computationally expensive scheme, which is why they appeared only once the processing hardware could sustain it. In practice HE* is close enough to lossless that most photographers cannot separate them in a blind comparison, and the storage saving is substantial.
The trade is compatibility. High Efficiency files need explicit support, and software that reads ordinary NEFs happily may not read these.
### 12-bit and 14-bit
Nikon exposes bit depth as an explicit menu choice rather than changing it behind your back. Fourteen bits gives 16,384 levels per channel against twelve bits' 4,096, and the difference matters in exactly one place: how far you can lift shadows before the steps between levels become visible as banding.
For images that will be graded gently, 12-bit is genuinely sufficient and writes faster. For anything with deep shadows you intend to open up, or for high-dynamic-range scenes, 14-bit is worth the size.
### Picture Controls, and why files look flat elsewhere
A NEF stores the active Picture Control — Standard, Neutral, Vivid, Flat, and any custom ones — as metadata rather than as applied processing. Nikon's own software reads it and renders accordingly, so a NEF opened in NX Studio looks like the camera's screen.
Third-party converters mostly do not, and start you from their own baseline rendering. Nothing has been lost; the setting is still in the file. Applying a camera-matching profile in Lightroom or Capture One restores the intended starting point.
One genuine trap: NRW is a different Nikon RAW format used by some compact cameras. It shares neither structure nor tooling with NEF despite the family resemblance in the name.
### Frequently asked questions
**Should I shoot 12-bit or 14-bit NEF?**
14-bit whenever shadow recovery is likely — landscape, interiors, anything high-contrast. 12-bit is fine for well-exposed work that will be graded lightly, and writes faster, which matters for burst shooting. The difference never shows in a normally exposed midtone; it shows in a shadow lifted three stops.
**What is the difference between HE and HE* on a Nikon Z?**
Both are High Efficiency modes using a more advanced compression scheme than the older options. HE* prioritises quality and targets parity with lossless compressed at roughly a third of the size; HE prioritises size and goes further. Both need software that explicitly supports them.
**Why does my NEF look flat in Lightroom but not in Nikon's software?**
Nikon's NX Studio reads the Picture Control stored in the file and renders with it applied. Lightroom starts from its own baseline instead. The setting is still there — choosing a Camera Matching profile in the Profile browser gets you back to the camera's rendering.
**Is lossy compressed NEF worth avoiding?**
Not on principle. Nikon's lossy scheme removes precision from the highlights, where it is hardest to see, and the practical cost is small. Use lossless compressed when you can — it costs nothing — but a lossy NEF is not a compromised file in ordinary use.
**Can I convert a NEF to DNG?**
Yes, with Adobe's free DNG Converter. The conversion is lossless for the image data and produces a file that far more software can read. Nikon-specific metadata such as Picture Control settings may not carry across in a form other applications will use.
---
## What Is a RAF File? Fujifilm RAW and X-Trans Explained
URL: https://www.rawcut.art/learn/what-is-a-raf-file
Updated: 2026-08-05
Tool for this: Read a RAF's metadata — https://www.rawcut.art/tools/metadata-reader
A .RAF file is Fujifilm RAW — sensor data plus settings from an X-series or GFX camera.
Unlike most RAW formats, the interesting part of a RAF is not the container. It is that many Fujifilm cameras use a colour filter array that almost nothing else uses, which means the software reading the file has to do something genuinely different with it. That single fact explains most of what people notice about Fujifilm files.
### X-Trans, and why it complicates things
Almost every digital camera places colour filters over its sensor in a Bayer pattern: a repeating 2×2 tile of red, two greens and blue. Decades of research have gone into reconstructing a full-colour image from it, and every RAW converter is very good at it.
Fujifilm's X-Trans sensors use a 6×6 tile instead, with a less regular arrangement of the same three colours. The intent was to break up the repetition that causes moiré, allowing Fujifilm to omit the optical low-pass filter and gain a little sharpness.
The consequence is that standard demosaicing algorithms do not apply. Every converter has to implement X-Trans handling separately, and they differ in quality. This is the origin of the long-running complaint about 'watercolour' or 'wormy' rendering of fine detail — foliage especially — in some software: it is a demosaicing artefact, not something wrong with the file.
Not every Fujifilm camera is X-Trans. The GFX medium-format line uses conventional Bayer sensors, as do some entry-level X bodies, and those files behave like any other RAW.
### Which software handles RAF well
This matters more for Fujifilm than for other brands, and it is worth testing rather than assuming. Capture One has long been regarded as strong on X-Trans. Iridient Developer built its reputation on it. RawTherapee and darktable both offer X-Trans-specific demosaic algorithms you can choose explicitly.
Adobe's handling has improved substantially over the years and its Enhance Details feature was aimed squarely at this problem. The old criticism is less true than it was, but the fact that it needed a dedicated feature tells you the underlying difficulty is real.
If fine natural detail in a RAF looks smeared or oddly textured, trying a second converter on the same file is a genuinely useful diagnostic — and one that would tell you nothing on a Bayer file.
### Film simulations live in the file
Fujifilm's film simulations — Provia, Velvia, Classic Chrome, Acros and the rest — are a large part of why people shoot the brand, and a RAF records which one was active as metadata rather than baking it in.
Fujifilm's own software and a few third-party converters apply it. Others do not, which is why a RAF can open looking noticeably plainer than the camera's screen showed. The embedded JPEG preview has the simulation applied; the RAW data does not.
This is also why in-camera RAW conversion remains genuinely useful on Fujifilm bodies in a way it is not elsewhere: reprocessing a RAF in-camera gives you the manufacturer's own rendering of a simulation, which no external converter reproduces exactly.
### Compression and file size
Recent bodies offer uncompressed, lossless compressed and lossy compressed RAF. As with every manufacturer, lossless compressed is the default worth using — smaller than uncompressed with no cost.
RAF files from higher-resolution bodies are large, and the GFX line larger still. Nothing about the format is unusual here; the sensors are simply big.
### Frequently asked questions
**Why does fine detail in my RAF look smeared?**
Almost certainly X-Trans demosaicing rather than the file. Foliage, hair and fine texture are where the difference between converters shows most. Opening the same file in a second converter — Capture One, Iridient, or RawTherapee with an X-Trans-specific algorithm — will tell you in a minute whether that is the cause.
**Do film simulations survive into Lightroom?**
The setting is recorded in the file, but Lightroom does not apply Fujifilm's rendering of it by default. Adobe ships camera-matching profiles named after the simulations which get close. For the manufacturer's exact rendering, reprocess the RAF in-camera or use Fujifilm's own software.
**Are all Fujifilm cameras X-Trans?**
No. The GFX medium-format line uses conventional Bayer sensors, and some entry-level X-series bodies do too. Those files demosaic like any other RAW and none of the X-Trans considerations apply.
**Should I convert RAF to DNG?**
It is an option for compatibility, and Adobe's free converter handles X-Trans files. Be aware that the DNG still contains X-Trans mosaic data, so it does not sidestep the demosaicing question — it only changes which software can open the file.
**What is a .RAF file's relationship to the JPEG my camera also wrote?**
The JPEG is Fujifilm's own rendering of the same capture, with the film simulation, sharpening and noise reduction applied. The RAF is the data before any of that. On Fujifilm in particular the JPEG is genuinely good enough that many photographers keep both deliberately.
---
## What Is a DNG File? Adobe's Open RAW Format Explained
URL: https://www.rawcut.art/learn/what-is-a-dng-file
Updated: 2026-08-05
Tool for this: Inspect a DNG — https://www.rawcut.art/tools/metadata-reader
A .DNG file is a Digital Negative: Adobe's attempt at a RAW format that is not tied to one manufacturer. Its specification is published, which is the whole point — a CR3 or an ARW can only be read by software someone chose to write a decoder for, and that choice may not outlive the camera.
The confusing part is that two very different things share the extension. Some cameras write DNG natively. Other DNGs are conversions of a manufacturer's RAW file. They behave differently and the distinction matters.
### Native DNG and converted DNG
A native DNG is what the camera wrote. Leica, Pentax, Ricoh, Sigma, DJI drones and Apple's ProRAW all produce them directly, and for those cameras DNG simply is the RAW format — there is no other version of the file.
A converted DNG is what Adobe's free DNG Converter produces from a NEF, CR3, ARW or similar. The sensor data comes across losslessly, so nothing is degraded, but this is a translation rather than the original document.
The practical difference is metadata. Manufacturer-specific information — Nikon's Picture Controls, Fujifilm's film simulations, some lens-correction data, various maker notes — has no standard place to live in a DNG. Adobe's converter preserves what it can, but software downstream may not know how to use it.
### Why anyone converts
Three reasons, in descending order of how often they apply.
Compatibility: a camera released last month may not open in your converter for weeks. DNG conversion is the standard stopgap, because almost everything reads DNG.
Self-contained edits. This is the one photographers feel day to day. Edits to a proprietary RAW live in a separate sidecar file — an .xmp next to the .NEF — and if the two are separated, the edits are gone. A DNG stores them inside the file itself, so moving one file moves everything.
Longevity. An open, published specification is a better bet across decades than an undocumented one, which is the archival argument institutions make.
### Mosaic, linear and lossy DNG
A normal DNG holds mosaic data — the raw sensor readings before demosaicing, exactly like the file it came from. This is the form that keeps all the flexibility of RAW.
A linear DNG has already been demosaiced: the colour reconstruction is done and baked in. Some converters produce these, and some cameras with unusual sensor layouts write them. They remain high-bit-depth and editable, but the demosaicing decision has been made for you and cannot be revisited.
There is also a lossy DNG option, which applies JPEG-style compression to the image data for dramatically smaller files. It is a real compromise and worth choosing deliberately rather than by accident — it is the one setting in the DNG Converter that discards information.
### Embedding the original
The DNG Converter can embed the complete original RAW file inside the DNG it produces. The result is larger than either file alone, but it means the conversion is fully reversible: the original can be extracted later, byte for byte.
For anyone converting an archive, this removes most of the risk from the decision. You keep the open format's advantages without betting that nothing was lost in translation.
### Frequently asked questions
**Does converting to DNG lose quality?**
No, unless you choose the lossy DNG option. The sensor data is transferred with lossless compression and produces identical results. What can be lost is manufacturer-specific metadata that has no standard representation in DNG.
**Should I convert my RAW files to DNG?**
It depends on what you value. Convert if you want edits stored inside the file rather than in sidecars, or if you are archiving for the long term and want an open format. Do not convert if you rely on manufacturer software or on features that read maker-specific metadata. Embedding the original removes most of the downside if you are unsure.
**Is DNG smaller than a manufacturer's RAW?**
Usually somewhat, because Adobe's lossless compression is often more efficient than what a camera applies in real time. The saving is modest — typically 10 to 20% — and disappears if you embed the original, which makes the file larger than both.
**What is the difference between DNG and a .dcp camera profile?**
A DNG is an image file. A DCP is a colour profile that tells a converter how to interpret one — the matrices and curves that turn sensor readings into meaningful colour. They share lineage in Adobe's DNG specification but do entirely different jobs.
**Why does my iPhone ProRAW file open as a DNG?**
Because it is one. ProRAW is Apple's use of the DNG specification, with computational-photography results written into the format. It is a native DNG, not a conversion, and includes more processing than a conventional RAW does.
**Can I convert a DNG back to CR3 or NEF?**
Only if the original was embedded at conversion time, in which case it can be extracted intact. Otherwise no — the manufacturer's container cannot be reconstructed from the DNG, even though the image data is all there.
---
## What Is a .3dl File? Autodesk's 3D LUT Format Explained
URL: https://www.rawcut.art/learn/what-is-a-3dl-file
Updated: 2026-08-05
Tool for this: Convert a .3dl now — https://www.rawcut.art/tools/3dl-to-cube
A .3dl file is a three-dimensional colour lookup table in the format Autodesk introduced with Lustre and still uses across Flame. Like a .cube, it maps input colours to output colours; unlike a .cube, it writes them as integers and leaves several things implicit.
If you have been handed a .3dl and your software wants a .cube, the conversion is straightforward — but there is one detail that silently ruins it, and it is worth knowing before you trust a converter.
### What is inside a .3dl
The file is plain text. It may open with a header — 3DMESH on one line, then Mesh followed by two numbers — or with nothing at all. Then comes a single line of many integers, the shaper ramp, whose length tells you the grid size. Then one line per lattice point, three integers each.
Compared with .cube, three things are missing or different. The values are integers rather than decimals between 0 and 1. There is no field naming the grid size explicitly; you infer it from the ramp or from the entry count. And there is no statement of what scale the integers are on.
- 3DMESH — optional header keyword
- Mesh