Imagild · practical answers
Choose a workflow by the result you need. These answers separate published photo methods from newer workbench previews, and explain where AI output still needs your review.
Reviewed September 26, 2026 · Availability and prices can changeBrowse the Image Skill Market, compare the original and result for a method, then try it with a photo you have permission to use. A Skill is a repeatable visual direction, not a promise of an identical outcome for every input.
Compare photo SkillsChoose a Skill when you want a new visual interpretation, such as a poster or illustrated travel memory. Choose the editor for controlled exposure, crop, perspective or local corrections. Inspect faces, hands, text and important objects after any AI operation.
Open the photo editorThe standard Skill flow lets you choose a photo and method before sign-in. An account and verification are required when you generate. Check the current workbench for access rules on other tools.
See the first-image flowStart with a clear, authorised source photo, request minimal changes to facial structure, and compare the result with the original. An instruction to preserve identity reduces ambiguity but cannot guarantee exact facial fidelity; reject results that alter the person.
Open controlled editingLook for a source photo that resembles your subject and framing, then compare the outline, pose, hands, buildings, small objects and any text. The example shows what a method can do; it does not guarantee the same result from a different photo.
Compare eight source/result pairsYes: treat the generated image as a draft, compare it with the source, and use the photo workbench for restrained crop, exposure or local corrections where appropriate. Keep the original photo so you can reject an edit that changes an essential detail.
See the editor workflowYes, as digital postcard-style or decorative stamp-style artwork. Neither workflow mails or manufactures a physical product, and generated lettering or borders should be checked before printing.
Photo-to-postcard guideYou can develop photo-led digital artwork for either format. Add final copy in a controlled layout tool when text must be exact, and check usage rights, crop, resolution and any distributor or printer requirements separately.
Photo-to-poster guideThe marketplace includes visual interpretations in these directions. A zine-style image is not an editable multi-page publication; pixel-art interpretation is not a fixed-grid game sprite; digital Riso texture is not press-ready colour separation.
Explore visual methodsMatch the example’s subject and composition to yours. Landscape, portrait and small-object examples do not transfer equally. Compare the source/result pair and inspect whether the method keeps the features you care about.
See real comparisonsUse a poster Skill for an illustrated visual direction; use the editor for controlled corrections to the original photo. If the event name, date or sponsor mark must be exact, finish that typography in a layout tool after generation and proof the exported dimensions before publishing.
Read the poster workflowNo. These Skills produce digital interpretations. Pixel-art-inspired output is not automatically a fixed-grid game asset, and Riso-looking output is not separated into press-ready inks. Confirm dimensions, colors and any manufacturing requirements with the target workflow.
Compare pixel and print stylesPortrait directions are being developed in the new workbench. Check the tool itself for current generation availability before buying a plan for a particular preset. Use consented photos; for children, a guardian should control upload and sharing. An AI travel scene is not a record of real travel.
Check portrait studio statusThe fan-merch workbench previews several digital formats. Check whether generation is enabled for the selected format; physical printing is separate. Do not use a celebrity likeness, logo or copyrighted image for distribution without the necessary permissions.
Check fan-merch toolsA visual analyser can describe visible subjects, action, light, framing and style to propose a new prompt. It cannot recover the original hidden prompt. A still image alone cannot prove motion or camera movement; verify current tool availability before relying on it.
Check prompt-analysis toolDo not assume so. A longer film generally needs an approved script, multiple shots, continuity checks, selective regeneration and assembly. The microfilm workspace is under development; check its current generation and editing steps before ordering.
Check microfilm workspaceApprove a script and shot list first. Identify which source photos each scene may use, keep a consistent wardrobe and location record, then review scene order, transitions and rights to music or likenesses. Generate and inspect individual shots before assembling a longer film; do not promise seamless continuity from one prompt.
Review the film workflowA photo supports observations about visible subject, framing, lighting and color. A video can additionally show action sequence, camera movement and timing. Either output is an interpretation, not a recovery of the original hidden prompt; mark unseen details as uncertain.
Check prompt-analysis availabilityNew verified accounts receive 50 points once, not daily. A successful 1K image costs 10 points; 2K and 4K cost 15 and 20. Do not assume preview video workflows are included; check current tool availability and pricing.
Read current pricingPoints are reserved when a generation starts. If the job fails without a delivered result, the reserved points are released. A delivered image costs the amount shown for its resolution; inspect balance and job details if something looks wrong.
Review jobsNo. Generation produces digital visual results. Verify the downloaded dimensions, edit exact text separately, obtain source and likeness permissions, and proof physical production independently. AI can change details even when asked to preserve them.
Read image-rights guidance