From 5370676783220ce191a7f75eb85b3365ed2887a8 Mon Sep 17 00:00:00 2001 From: Cursor Agent Date: Wed, 18 Nov 2026 15:20:04 +0000 Subject: [PATCH] Link English tool pages to the matching blog guides. Add body-only sentences so the AI, preset, LUT, batch and collage pages point at the existing English guides, and link the .cube post back to photo presets. Co-authored-by: Joan Comadran --- src/content/blog/en/what-is-a-cube-lut.mdx | 2 +- src/content/tools/en/background-remover.md | 2 +- src/content/tools/en/batch.md | 2 +- src/content/tools/en/collage.md | 2 +- src/content/tools/en/lut.md | 2 +- src/content/tools/en/presets.md | 2 ++ src/content/tools/en/remove-objects.md | 4 ++-- src/content/tools/en/upscaler.md | 2 +- 8 files changed, 10 insertions(+), 8 deletions(-) diff --git a/src/content/blog/en/what-is-a-cube-lut.mdx b/src/content/blog/en/what-is-a-cube-lut.mdx index 6ac4eae..873f85b 100644 --- a/src/content/blog/en/what-is-a-cube-lut.mdx +++ b/src/content/blog/en/what-is-a-cube-lut.mdx @@ -65,7 +65,7 @@ They solve different problems, and Lightframes imports both. | What it cannot carry | Crop, masks, lens corrections, a camera profile | Camera profiles, lens corrections and masks are not imported | | When to use it | A look you bought or exported as a `.cube`, including a film grade you also use on video | A preset whose recipe you want as sliders, including packs sold as `.xmp` | -A LUT is a finished colour map. An XMP preset is a list of instructions. If the pack gives you both, the `.cube` applies the table and the `.xmp` applies the slider recipe; they are not the same pipeline. The steps for `.xmp` files, and which settings carry over, are in [using XMP presets without Lightroom](/xmp-presets-without-lightroom). +A LUT is a finished colour map. An XMP preset is a list of instructions. If the pack gives you both, the `.cube` applies the table and the `.xmp` applies the slider recipe; they are not the same pipeline. The steps for `.xmp` files, and which settings carry over, are in [using XMP presets without Lightroom](/xmp-presets-without-lightroom). Saving either file so you can apply it again is on [photo presets](/tools/photo-presets). Lightroom and Camera Raw are trademarks of Adobe. Lightframes is independent and is not affiliated with Adobe. diff --git a/src/content/tools/en/background-remover.md b/src/content/tools/en/background-remover.md index 56005a3..fda23fb 100644 --- a/src/content/tools/en/background-remover.md +++ b/src/content/tools/en/background-remover.md @@ -54,7 +54,7 @@ The subject is detected by an AI model that runs in your browser, on your device 3. Lightframes cleans the mask, dropping small floating specks the model may leave in busy backgrounds while keeping a soft margin so hair and fuzzy edges survive. 4. The mask is computed once per photo and reused for the preview, the export and any crop, so changing the framing doesn't run the model again. -After that first download, background removal works offline. The edit is non-destructive: the original file is untouched, and switching back to **Original** restores the photo. +After that first download, background removal works offline. The edit is non-destructive: the original file is untouched, and switching back to **Original** restores the photo. Which model finds the subject, and where that file is loaded from, is in [run AI image models in the browser](/blog/run-ai-image-models-in-the-browser). ## Background modes and options diff --git a/src/content/tools/en/batch.md b/src/content/tools/en/batch.md index d572310..288e140 100644 --- a/src/content/tools/en/batch.md +++ b/src/content/tools/en/batch.md @@ -81,7 +81,7 @@ If nothing has been copied, Paste edits uses the active photo as the source. Sel ### Presets on a selection -Applying a preset from the Presets tab affects every selected photo, and the Intensity slider (0–100%) adjusts all of them together. See [photo presets](/tools/photo-presets) for saving and importing looks. +Applying a preset from the Presets tab affects every selected photo, and the Intensity slider (0–100%) adjusts all of them together. See [photo presets](/tools/photo-presets) for saving and importing looks. A film-style grade you can paste across the set, including which groups to copy, is in [make photos look like film](/blog/make-photos-look-like-film). ## Selecting and sorting photos diff --git a/src/content/tools/en/collage.md b/src/content/tools/en/collage.md index ed538da..ebd0296 100644 --- a/src/content/tools/en/collage.md +++ b/src/content/tools/en/collage.md @@ -92,7 +92,7 @@ If you open more than nine photos, Lightframes starts you off with grids of four - **Use thin borders for a modern look**, or turn borders off for a tight mosaic. White or cream borders give a classic print look; black suits night and city shots. - **Mix orientations sensibly.** Put portrait photos in tall cells and landscapes in wide ones to avoid heavy crops. - **Give the first slide a strong lead photo** in multi-slide posts, since it's the one shown in the feed. -- **Want one image running across slides?** That's the [Instagram carousel maker](/tools/instagram-carousel-maker), in the same mode. +- **Want one image running across slides?** That's the [Instagram carousel maker](/tools/instagram-carousel-maker), in the same mode. How to split one panorama into those slides is in [make a seamless Instagram carousel from a panorama](/blog/seamless-instagram-carousel-panorama). ## Limitations diff --git a/src/content/tools/en/lut.md b/src/content/tools/en/lut.md index 99d8b29..08f4372 100644 --- a/src/content/tools/en/lut.md +++ b/src/content/tools/en/lut.md @@ -46,7 +46,7 @@ updated: 2026-11-08 ## What applying a LUT in Lightframes means -A LUT (look-up table) is a file that maps every input colour to an output colour, and applying one gives a photo a ready-made colour grade in one step. Lightframes reads 3D LUTs in the common `.cube` format, the same files often sold or shared as film emulations and cinematic looks, and lets you apply them to still photos, including camera RAW files. It is for photographers who already own LUTs, video creators who want matching stills, and anyone who wants a consistent grade across a set. The editor is free, open source and runs in the browser. For what `LUT_3D_SIZE` and the RGB table mean, and how a `.cube` differs from a Lightroom preset, read [what a .cube LUT is and how to use it on photos](/blog/what-is-a-cube-lut). +A LUT (look-up table) is a file that maps every input colour to an output colour, and applying one gives a photo a ready-made colour grade in one step. Lightframes reads 3D LUTs in the common `.cube` format, the same files often sold or shared as film emulations and cinematic looks, and lets you apply them to still photos, including camera RAW files. It is for photographers who already own LUTs, video creators who want matching stills, and anyone who wants a consistent grade across a set. The editor is free, open source and runs in the browser. For what `LUT_3D_SIZE` and the RGB table mean, and how a `.cube` differs from a Lightroom preset, read [what a .cube LUT is and how to use it on photos](/blog/what-is-a-cube-lut). For a film look built from the tone curve, colour grading and grain, with or without a `.cube`, see [make photos look like film](/blog/make-photos-look-like-film). ## How it works, and why it stays on your device diff --git a/src/content/tools/en/presets.md b/src/content/tools/en/presets.md index 8272000..4316419 100644 --- a/src/content/tools/en/presets.md +++ b/src/content/tools/en/presets.md @@ -89,6 +89,8 @@ These are the same develop settings stored in presets exported from Lightroom or Lightroom and Camera Raw are trademarks of Adobe. Lightframes is an independent open-source project and is not affiliated with or endorsed by Adobe. +A film-style grade you build with the sliders can be saved as a preset. The order of those sliders is in [make photos look like film](/blog/make-photos-look-like-film). A `.cube` file is a colour table rather than a list of sliders. What `LUT_3D_SIZE` and the RGB samples mean is in [what a .cube LUT is](/blog/what-is-a-cube-lut). + ## Tips - Build presets on a well-exposed photo. A preset saved from a photo that needed +2 EV will push every other photo up too. diff --git a/src/content/tools/en/remove-objects.md b/src/content/tools/en/remove-objects.md index 0bd5534..ba214f0 100644 --- a/src/content/tools/en/remove-objects.md +++ b/src/content/tools/en/remove-objects.md @@ -45,7 +45,7 @@ updated: 2026-10-20 The object remover in Lightframes lets you erase things you don't want in a photo by painting over them. It's built for the everyday clean-up jobs: a tourist walking through your landscape, a bin or a sign at the edge of the frame, a power line across the sky, skin blemishes in a portrait, or the dust spots that a dirty sensor leaves in the same place on every shot. -It sits in the same editor you use for exposure, colour and presets, so you can retouch and develop a photo in one place. If you shoot RAW, you can open the file directly; spot removal works on RAW files from the [RAW photo editor](/tools/raw-photo-editor) just as it does on JPEGs. +It sits in the same editor you use for exposure, colour and presets, so you can retouch and develop a photo in one place. If you shoot RAW, you can open the file directly; spot removal works on RAW files from the [RAW photo editor](/tools/raw-photo-editor) just as it does on JPEGs. The step-by-step, including when to use Heal, Clone or Remove, is in [remove objects from photos online](/blog/remove-objects-from-photos-online). ## How it works, and what happens to your photo @@ -61,7 +61,7 @@ The three modes use different techniques: For Heal and Clone, Lightframes picks a source area automatically by looking for the best match around the spot. If it chooses badly, drag the green outline to another place or press **New source** (shortcut `/`) to try the next candidate. -Remove needs the model the first time you select it. It downloads once from Hugging Face (about 28 MB), is checked against a fixed checksum, and is then kept in your browser's storage, so later sessions start straight away, even offline. The model runs on your device's processor in a background worker, so the editor stays responsive while it fills a spot. Your photo is not part of that download. +Remove needs the model the first time you select it. It downloads once from Hugging Face (about 28 MB), is checked against a fixed checksum, and is then kept in your browser's storage, so later sessions start straight away, even offline. The model runs on your device's processor in a background worker, so the editor stays responsive while it fills a spot. Your photo is not part of that download. Which model Remove loads, and where that file comes from, is in [run AI image models in the browser](/blog/run-ai-image-models-in-the-browser). ## Controls diff --git a/src/content/tools/en/upscaler.md b/src/content/tools/en/upscaler.md index f7924f3..a166c24 100644 --- a/src/content/tools/en/upscaler.md +++ b/src/content/tools/en/upscaler.md @@ -48,7 +48,7 @@ In Lightframes, upscaling is part of exporting. You develop the photo as usual ( ## How it works, and why nothing is uploaded -Lightframes uses **Real-ESRGAN (general x4v3)**, an open super-resolution model under the BSD-3-Clause licence, and runs it inside your browser. The photo is never sent anywhere: the only thing downloaded is the model file itself, about 5 MB, the first time you pick 2× or 4×. It is checked against a fixed checksum and kept in your browser's storage, so later exports start straight away and work offline. +Lightframes uses **Real-ESRGAN (general x4v3)**, an open super-resolution model under the BSD-3-Clause licence, and runs it inside your browser. The photo is never sent anywhere: the only thing downloaded is the model file itself, about 5 MB, the first time you pick 2× or 4×. It is checked against a fixed checksum and kept in your browser's storage, so later exports start straight away and work offline. Which model enlarges the photo, and where that file is loaded from, is in [run AI image models in the browser](/blog/run-ai-image-models-in-the-browser). The model runs on your graphics card through WebGPU when your browser supports it, and falls back to the processor otherwise. It works on 128 px tiles that overlap slightly, keeping only the middle of each tile, so tile edges don't show as seams. Each tile's 4× result is resampled straight to the factor you asked for, which keeps memory use down.