Image Generation, In Practice

The craft of getting a specific, finished image out of a model on a deadline. Choosing between hosted tools and open weights on cost, control and licensing. Preset platforms like Higgsfield and what they take away. Reference images and denoise strength, inpainting and outpainting, pose and depth conditioning, character consistency, upscalers that invent detail, the finishing pass in a raster editor, and what you must tell a client.

1. 1The generator you pick is a licensing decision9 minThe licence on a model's weights governs whether you may use it commercially, and it is a completely separate question from who owns the picture it made.
2. 2A preset is somebody else's workflow, frozen8 minRecipe-driven tools trade control for speed, and the bill arrives at the first revision, when the thing the client wants changed is the thing the preset already decided.
3. 3Stop describing it and show it the picture8 minDenoise strength decides how much of your reference survives; the useful range is roughly 0.4 to 0.6, and everything above 0.75 throws the reference away.
4. 4The finished image is never one generation9 minGenerate a base plate, then repair it region by region; the "inpaint at full resolution" setting is what gives a small detail enough pixels to be drawn correctly.
5. 5The model is guessing your composition9 minConditioning makes composition an input rather than an outcome, and letting the control end around 60% of the steps is what stops the result looking traced.
6. 6The same face twice is the hard part9 minOne-photo identity adapters get you a family resemblance; a specific real person needs a LoRA trained on varied photographs, or their actual photographed face composited in an editor.
7. 7Upscaling either restores detail or invents it8 minDiffusion-based upscalers re-generate rather than sharpen, so never point one at text, a logo, a product detail or a face you need to stay recognisable.
8. 8Nothing goes to a client straight from the model8 minCurves for contrast, a colour pass to match the set, fine grain to fake a sensor, and real type over any generated words — that is the ten minutes that separates output from a deliverable.
9. 9You are paying to make decisions you could make cheaply8 minChoose the frame on cheap low-step renders and spend full quality only on the winner; after about six full-quality attempts the remaining distance is editing work, not generation.
10. 10What you can sell, and what you must disclose10 minBeing licensed to use a model, owning its output, and having to disclose that output is generated are three separate questions, and only the ownership one is genuinely unsettled.