Lore
AI Photo Editing
AI Photo Editing, as this book documents it, is a working knowledge base for turning a phone photo into something that reads as professionally made — not a single tutorial but several parallel routes: the phone-only path (Photos app, Lensa, Lightroom Mobile), the desktop Photoshop path (non-destructive foundations, then either AI-assisted or fully manual compositing), a separate local/node-based path through Flux and ComfyUI, and a closing grab-bag of tools and prompting techniques that don't fit anywhere else. The arc is cumulative in places and parallel in others: mobile-editing-baseline and lightroom-mobile-masking-and-light build a straight line of increasing control without ever leaving a phone; photoshop-non-destructive-foundations is the hinge everything Photoshop-related depends on, after which generative-fill-compositing-toolkit and manual-compositing-craft split into two answers to the same compositing problem — one AI-driven, one entirely by hand; flux-comfyui-model-tuning then leaves Photoshop altogether for a numbers-driven local toolchain; and beyond-photoshop-tools-and-prompting closes by naming what else exists outside all of that, explicitly and admittedly not exhaustively.
This is for someone who already has photos to edit and wants to know which tool and which technique actually earns a premium result, not a beginner course in any one app. Coverage is honestly uneven: photoshop-non-destructive-foundations draws on PHLEARN PRO trailers that name building blocks without teaching them step by step, manual-compositing-craft rests on a single worked panorama example, and flux-comfyui-model-tuning carries numbers its own source calls empirical or unconfirmed. And the subject's own stated goal names tools this book never reaches — the ChatGPT image model and Lovart appear in no chapter, a gap beyond-photoshop-tools-and-prompting says plainly at its close.
Как читать
Read front to back if you're building the whole picture: mobile-editing-baseline into lightroom-mobile-masking-and-light covers everything achievable on a phone, photoshop-non-destructive-foundations is the required desktop foundation before either generative-fill-compositing-toolkit or manual-compositing-craft, and flux-comfyui-model-tuning and beyond-photoshop-tools-and-prompting close out the local and outside-Photoshop tracks. If you already edit comfortably on mobile, skip straight to Photoshop Foundations: Non-Destructive Editing & Selection; if you already know Photoshop's non-destructive plumbing, go directly to whichever compositing route matches the job — Generative Fill & AI Compositing Toolkit for an AI-assisted background swap, Manual Compositing Craft: Light, Color & Depth for full hand control — and treat Flux & ComfyUI: Local Model Tuning as its own standalone track you can read independently of everything Photoshop.
Главы
- Mobile Baseline: iPhone Photos App & Lensa — The starting point for anyone beginning from a phone camera roll: one-tap Photos-app tools, manual sliders, and Portrait Mode's lighting styles, capped by handing the face off to Lensa. It opens the book because every later chapter assumes a photo has already passed through some version of this stage.
- Lightroom Mobile: Masking, Sliders & Simulated Light — Where phone editing graduates from one-tap adjustment into deliberate masking and simulated light, using Lightroom Mobile's Select-Subject tool, Effects sliders, and a fabricated radial-gradient sun. It follows the Photos/Lensa baseline as the next tier of control still available without touching a desktop.
- Photoshop Foundations: Non-Destructive Editing & Selection — The desktop entry point: smart objects, Camera Raw as a filter, and the selection-to-mask pipeline that every later Photoshop chapter — AI-assisted or manual — is built on top of. It comes here because nothing downstream in Photoshop makes sense without this plumbing first.
- Generative Fill & AI Compositing Toolkit — The AI-driven route through Photoshop compositing — Generative Fill, Harmonize, Generative Upscale, and the finishing stack that closes the gap they leave short of professional. It builds directly on the previous chapter's selection-to-mask pipeline, now aimed at a generated background instead of a real one.
- Manual Compositing Craft: Light, Color & Depth — The non-AI answer to the same compositing problem: a fully manual, depth-ordered build with hand-matched light and color at every layer. It sits right after the AI toolkit chapter as the other way to reach physical believability, no model call involved.
- Flux & ComfyUI: Local Model Tuning — A self-contained detour into local, node-based generation — LoRAs, the Model Sampling Flux node, and sweep workflows for tuning an image by numbers rather than by masking or prompting. It follows the Photoshop chapters because it's a genuinely different toolchain, not a continuation of them.
- Beyond Photoshop: Other Tools & Prompting Techniques — The closing chapter, collecting what's tied to other generators — Midjourney's character consistency, Firefly's reference tricks — and to prompting itself, for jobs neither Photoshop nor Flux is the right tool for. It comes last because it's deliberately narrow and says outright which tools it never reaches.
Открытые вопросы
- generative-fill-compositing-toolkit and manual-compositing-craft solve the same problem — a believable composite — through opposite mechanisms (AI generation plus a finishing stack, versus fully hand-matched light and color), but nothing across the book says when to choose one path over the other or how to combine them on a single job.
- The book's non-AI Photoshop material is thin in a way its AI material isn't: photoshop-non-destructive-foundations and manual-compositing-craft both trace back to PHLEARN PRO sources (unwalked trailers, a single worked panorama course), while generative-fill-compositing-toolkit and flux-comfyui-model-tuning draw on more varied, hands-on sources — so the manual side of the craft is documented in visibly less depth than the AI side.
- The subject's stated goal names six tool categories, but the chapters together reach only four of them (mobile editors, Photoshop generative AI, Flux/ComfyUI, and partially Midjourney); the ChatGPT image model and Lovart are never covered by any chapter, a gap beyond-photoshop-tools-and-prompting names directly rather than papering over.
- Matching a subject's light and color to its surroundings is taught three separate times with three different mechanisms — Harmonize's one-click AI pass in generative-fill-compositing-toolkit, eyedropper-sampled manual matching in manual-compositing-craft, and masked Effects/Mix adjustments in lightroom-mobile-masking-and-light — with no chapter saying how the three relate or whether a single image might need more than one.
- Selection and masking methods multiply without ever being ranked against each other: Lightroom Mobile's Select-Subject and Mix panel, Photoshop's Select-Subject/Remove-Background/Select-and-Mask pipeline, and Generative Fill's choice between a precise mask and a whole-image conversational prompt are each taught inside their own chapter, with nothing saying which to reach for first when more than one is technically available on the same photo.