[{"content":"Darktable 5.6 shipped what might be its most impactful feature yet for photographers migrating from Lightroom: AI-powered object masking. No subscription, no cloud, just open-source AI running locally on your machine.\nWhat is AI Object Masking? The new masking system uses SAM 2.1 and SegNext to detect and isolate subjects automatically — click on something and Darktable figures out the rest.\nHow it compares to Lightroom Lightroom\u0026rsquo;s AI masking has been a major reason photographers hesitate to switch. This closes that gap significantly.\nWatch the full tutorial Key takeaways AI masking is available in Darktable 5.6 and above Works locally — no data leaves your machine Supports people, objects, skies, and backgrounds Free, forever ","permalink":"https://mariengb.com/posts/darktable-ai-masking/","summary":"\u003cp\u003eDarktable 5.6 shipped what might be its most impactful feature yet for photographers migrating from Lightroom: AI-powered object masking. No subscription, no cloud, just open-source AI running locally on your machine.\u003c/p\u003e\n\u003ch2 id=\"what-is-ai-object-masking\"\u003eWhat is AI Object Masking?\u003c/h2\u003e\n\u003cp\u003eThe new masking system uses SAM 2.1 and SegNext to detect and isolate subjects automatically — click on something and Darktable figures out the rest.\u003c/p\u003e\n\u003ch2 id=\"how-it-compares-to-lightroom\"\u003eHow it compares to Lightroom\u003c/h2\u003e\n\u003cp\u003eLightroom\u0026rsquo;s AI masking has been a major reason photographers hesitate to switch. This closes that gap significantly.\u003c/p\u003e","title":"Darktable 5.6 AI Object Masking: The Lightroom Killer Feature?"},{"content":"Watch the full tutorial What is Neural Restore in Darktable 5.6? Darktable 5.6 introduces Neural Restore, an AI-powered denoising system that works fundamentally differently from the traditional denoise modules. Rather than operating within the processing pipeline, it functions as a plugin that generates a brand new DNG file — clean from the sensor data up. If you\u0026rsquo;re coming from Lightroom, think of it as the equivalent of Lightroom\u0026rsquo;s AI Denoise button, which also generates a new DNG.\nThis makes it particularly powerful for heavily noisy images — think ISO 10,000 and above — where standard denoise modules struggle.\nThree Modes: Which One to Use? Neural Restore offers three modes, found on the left panel in both Lighttable and Darkroom:\nRaw Denoise — works at the very beginning of the pipeline, directly on your raw sensor data. This is the recommended starting point for new images. It generates a clean DNG that you then process normally with your exposure, AgX, cropping and other modules.\nDenoise — works at the end of the pipeline, on an already processed image. Useful if you want to apply AI denoising to an image you edited months ago without redoing all your work. It exports a TIFF file rather than a DNG.\nUpscale — covered in a dedicated tutorial, watch it here:\nFor most workflows, start with Raw Denoise on new images. Use Denoise only to salvage already-processed files.\nSetup: Activating the AI Models Before using Neural Restore, you need to activate it in Darktable\u0026rsquo;s settings:\nGo to Settings → scroll to the AI section near the bottom Download the Raw Denoise model (and optionally the Denoise model) Enable the models you downloaded Activate AI features globally Mac users: Darktable can struggle when left to choose its own acceleration. Manually select Apple Core ML — it works reliably for both AI masking and Neural Restore, and processing is now very fast (a few seconds at most).\nUsing Raw Denoise: Step by Step Open your image in Darkroom Find the Neural Restore plugin on the left panel Click Generate Preview — Darktable renders a small preview area rather than the full image Move the peeker to inspect different areas of the image (if it stops responding, restart Darktable — a known bug that will be fixed) Adjust the Strength slider (0–100%) Set your output folder for the new DNG (default is same folder as original) Click Process — a new DNG appears in your Lighttable in about 5 seconds Finding the Right Strength There is no universal number — it depends on your image and your taste. Here\u0026rsquo;s a practical guide:\n100% — maximum denoising, but risks a plastic, detail-losing look especially on organic textures 80% — a good starting point, strong noise reduction with acceptable detail retention 60–70% — better detail preservation if 80% feels too aggressive The classic denoising trade-off applies here too: the more noise you remove, the more detail you risk losing. Generate a few test DNG files at different strengths and compare them directly in Lighttable — it\u0026rsquo;s easier than judging from the small preview.\nOne key advantage of Raw Denoise: because it works directly on sensor data before any processing, you can later lift shadows aggressively without noise creeping back in — something that\u0026rsquo;s much harder to achieve when denoising later in the pipeline.\nDarktable AI Denoise vs Lightroom AI Denoise: My Take Both tools produce excellent results at high ISO. Lightroom\u0026rsquo;s AI Denoise is polished and well integrated. But Neural Restore in Darktable 5.6 is genuinely competitive — and it\u0026rsquo;s completely free, open source, runs locally, and requires no subscription.\nThe Darktable team themselves recommend a sensible workflow: use the classic denoise modules first for moderately noisy images, and reach for Neural Restore only when standard denoising isn\u0026rsquo;t enough. That\u0026rsquo;s good advice — Neural Restore is a precision tool for difficult files, not a replacement for the full denoise workflow.\nThe model is still improving, and the speed improvements since early 2026 previews are significant. It\u0026rsquo;s worth experimenting with on your own images to find what works for your shooting style.\n","permalink":"https://mariengb.com/posts/darktable-neural-restore-denoise-and-upscale-tutorial/","summary":"\u003ch2 id=\"watch-the-full-tutorial\"\u003eWatch the full tutorial\u003c/h2\u003e\n\u003cdiv style=\"position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden;\"\u003e\n      \u003ciframe allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen\" loading=\"eager\" referrerpolicy=\"strict-origin-when-cross-origin\" src=\"https://www.youtube.com/embed/pqGB_ND7mLI?autoplay=0\u0026amp;controls=1\u0026amp;end=0\u0026amp;loop=0\u0026amp;mute=0\u0026amp;start=0\" style=\"position: absolute; top: 0; left: 0; width: 100%; height: 100%; border:0;\" title=\"YouTube video\"\u003e\u003c/iframe\u003e\n    \u003c/div\u003e\n\n\u003ch2 id=\"what-is-neural-restore-in-darktable-56\"\u003eWhat is Neural Restore in Darktable 5.6?\u003c/h2\u003e\n\u003cp\u003eDarktable 5.6 introduces Neural Restore, an AI-powered denoising system that works fundamentally differently from the traditional denoise modules. Rather than operating within the processing pipeline, it functions as a plugin that generates a brand new DNG file — clean from the sensor data up. If you\u0026rsquo;re coming from Lightroom, think of it as the equivalent of Lightroom\u0026rsquo;s AI Denoise button, which also generates a new DNG.\u003c/p\u003e","title":"Darktable Neural Restore Denoise and Upscale Tutorial"},{"content":"Watch the full tutorial What is AI Upscaling in Darktable 5.6? Darktable 5.6 completes its AI trilogy — alongside AI masking and AI denoising — with AI upscaling. This feature goes directly head-to-head with Lightroom\u0026rsquo;s Super Resolution, letting you increase the megapixel count of your image with a single click, entirely free and open source.\nLike Neural Restore Denoise, upscaling is part of the Neural Restore plugin (not a traditional pipeline module). If you haven\u0026rsquo;t seen the denoise video yet, watch it first — the setup is identical.\nHow it Works Upscaling generates a new TIFF file from your original — it\u0026rsquo;s not non-destructive like standard Darktable modules. Your original file stays untouched, and the new upscaled TIFF is automatically imported into your film roll. This is the same approach Lightroom Super Resolution uses.\nOne key advantage over Lightroom: Darktable offers both 2x and 4x upscaling, whereas Lightroom is limited to 2x. A 24MP file from a Nikon D750 for example becomes a 97MP file at 4x — though bear in mind that comes with significantly larger file sizes and longer processing times.\nSetup The setup is identical to AI Denoise:\nGo to Settings → AI section Download the Upscale model Enable the model Activate AI features globally One thing worth noting: upscaling is available in both Darkroom and Lighttable. The Lighttable option is particularly useful because you can batch process multiple images at once — a meaningful advantage over working image by image in Darkroom.\nUsing the Upscale Function Open your image in Darkroom (or select multiple in Lighttable for batch) Find Neural Restore on the left panel, select the Upscale tab Click Generate Preview — hover over the preview to magnify and inspect detail Choose your scale: 2x or 4x Set your output TIFF depth and colour profile Choose your output folder (default is same as original) Hit Process On an Apple M3 Pro MacBook Pro processing via GPU, a 24MP → 97MP upscale takes roughly 2 minutes. Plan accordingly — run a batch and step away while it processes.\nImportant: Wide Gamut Warning This is the most critical thing to know before using upscaling: the upscale model does not handle wide gamut colours. Any colours outside the sRGB range — think vivid sunsets, bright skies, saturated subjects — will be clipped, not compressed. Hard clipping, not a graceful rolloff.\nThe practical advice: work on your image first, map your colours, and bring everything within a safe gamut before upscaling. Apply AgX or your tone mapper, check your highlights, then upscale the finished result rather than the raw file.\nWhen Should You Actually Use It? Upscaling is a precision tool, not an everyday one. The most useful cases:\nPrinting — pushing a smaller file to A3 or larger where your original megapixel count falls short Old or scanned images — breathing new life into archive photos from older cameras or film scans Cropped images — recovering resolution after a heavy crop For most modern cameras shooting at 24MP+, you\u0026rsquo;re already fine printing A3 without upscaling. Use it when you have a specific reason, not as a routine step.\nDarktable Upscale vs Lightroom Super Resolution Darktable Lightroom Max upscale 4x 2x Cost Free Subscription required Output TIFF DNG Wide gamut Clips Handles gracefully Batch processing ✅ ✅ Speed (24MP→97MP) ~2 min Similar The wide gamut handling is currently Lightroom\u0026rsquo;s edge. Everything else — especially the 4x option and the price — favours Darktable. As the model matures, that gamut limitation will likely improve.\nWhat\u0026rsquo;s Next This wraps up the Darktable 5.6 AI series. The next video covers the UX and UI improvements in 5.6 — smaller changes that add up to a meaningfully smoother workflow. Stay tuned.\n","permalink":"https://mariengb.com/posts/darktable-neural-restore-upscale-tutorial/","summary":"\u003ch2 id=\"watch-the-full-tutorial\"\u003eWatch the full tutorial\u003c/h2\u003e\n\u003cdiv style=\"position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden;\"\u003e\n      \u003ciframe allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen\" loading=\"eager\" referrerpolicy=\"strict-origin-when-cross-origin\" src=\"https://www.youtube.com/embed/Tar6N3hqZD4?autoplay=0\u0026amp;controls=1\u0026amp;end=0\u0026amp;loop=0\u0026amp;mute=0\u0026amp;start=0\" style=\"position: absolute; top: 0; left: 0; width: 100%; height: 100%; border:0;\" title=\"YouTube video\"\u003e\u003c/iframe\u003e\n    \u003c/div\u003e\n\n\u003ch2 id=\"what-is-ai-upscaling-in-darktable-56\"\u003eWhat is AI Upscaling in Darktable 5.6?\u003c/h2\u003e\n\u003cp\u003eDarktable 5.6 completes its AI trilogy — alongside AI masking and AI denoising — with AI upscaling. This feature goes directly head-to-head with Lightroom\u0026rsquo;s Super Resolution, letting you increase the megapixel count of your image with a single click, entirely free and open source.\u003c/p\u003e","title":"Darktable Neural Restore Upscale Tutorial"}]