Studio
Studio is Freedam's AI image editor. It creates edited versions of an image, such as social crops, background changes, branded composites and people edits, without exporting the file to another tool. The original asset is never changed, every edit of a session is recorded, and a result you save is attached to the asset as a rendering that goes through review.
Studio works on images only.
Who can use Studio
- You need the Use Studio permission. The User role has it by default, and Studio is included on every paid plan.
- You must be able to see the asset. Opening, running a skill and saving a result all check this, so a session stops working if the asset is later hidden from you.
- The access rule that applies to you must allow AI editing of the asset, and the asset's rights policy must not forbid modification. Otherwise Studio refuses to open the asset and shows a message explaining why, so you never start work that can't be completed.
- Saving a result does not need Edit asset metadata or Add and edit attachments. A person who can only view an asset can still save Studio results on it, when the three conditions above are met.
- You work only in the sessions you started.
- If you cannot see the Open in Studio button on an asset, ask your workspace administrator to add Use Studio to your role.
- People with the View other people's Studio work permission can see the sessions and results of other people, for the assets they can see themselves. The Studio audit screens in AI Operations also need Manage AI operations.
Opening Studio
On the asset detail page of an image, click Studio in the top bar. On narrow screens the button is in the More menu, as Open in Studio.
Studio opens full-screen. Click Exit (top-right) to return to the asset detail page. Your session and its versions are kept: if you open Studio on the same asset again within 24 hours, the session is restored. After 24 hours, Studio starts a new session.

Page layout
- Top bar: The Studio label, the original filename, and the Exit button.
- Left rail (desktop): Every tool, as an icon with its name. Hover a tool for a one-line description. Click a tool to open its settings panel next to the rail.
- Top canvas toolbar: Controls for the active tool only: the focus controls when cropping, Clear All when painting regions, Add Asset and the text controls when building a brand overlay.
- Canvas (main area): The image you are working on, with Zoom in, Zoom out and Reset zoom buttons. This is where you drag the crop box, paint regions, click detected people, and move overlay layers. Outside the Crop tool, a Starting from: selector in the bottom-right corner chooses the image the next edit starts from: Original or one of your versions.
- Bottom strip, Versions: Every result generated during this session, in the order it was produced. Hover a thumbnail for a larger preview. Click it to work from that version. The two buttons under each thumbnail save it or delete it.
- Mobile layout: On phones and narrow tablets the left rail is replaced by a bottom tool bar. Tap a tool to open a slide-up panel with its settings.
The twelve generation tools
Studio offers twelve built-in tools, selected in the left rail. Each one takes the current base version on the canvas, applies a transformation, and adds a new version to the bottom strip.
| Tool | What it expects | What you get back |
|---|---|---|
| Remove Background | Nothing to set | The subject cut out on a transparent canvas. |
| Crop | A preset, or a free box you drag on the canvas | A crop at the preset's size. If the box extends past the edge of the image, Studio fills the new area with AI-generated content (outpainting). |
| Resize | A new width and height in pixels | The image at that size. Up to 125% of the current size, the image is scaled directly. Above 125%, it goes to an AI upscaling model. |
| Add Background | A solid colour, or a scene preset and an optional description | The subject placed over the chosen background. The image must have transparent areas. |
| Brand Overlay | Logos, emblems and text layers from your Brand Guide, placed on the canvas | A branded composite that follows your Brand Guide rules on minimum size, maximum size and distance from the edges. |
| Style Transfer | A short description of the style you want | The image redrawn in that style (colour, texture, rendering), with the subject still recognizable. |
| Remove / Add | One or more regions painted on the canvas, and Remove or Add | The painted areas erased and filled from their surroundings, or replaced with what you describe. |
| Perspective | A camera angle: Top-down, Low angle, Centered subject or Shallow depth of field | The same scene rendered from the chosen angle. An optional instructions field lets you add detail. |
| Remove People | The people Studio detects in the image, of which you select some | The selected people removed and the background filled in, or changed according to instructions you type ("make them smile", "change hair colour to blonde"). |
| Daylight | Nothing to set | The same scene relit with soft, natural daylight. Composition and content stay identical. |
| Enhance | Nothing to set | A sharper, clearer image with finer detail, and minor exposure or colour issues corrected. |
| Denoise | Nothing to set | Grain and sensor noise reduced, with fine detail and texture kept. |
Most tools run from a Generate button at the bottom of their panel. Crop, Brand Overlay and Remove People have their own button: Apply Crop, Apply Brand Overlay, and Remove Selected or Apply Edit. While a tool runs, the button shows the current step, for example "Preparing image…", "Generating…" or "Saving result…". Each run adds a new version to the bottom strip. The original asset is never changed.
Your rail may differ from this list. Administrators can add custom skills built for your workspace, and can hide or reorder any tool. Brand Overlay is shown only when your Brand Guide contains logos or emblems.
Custom skills
Next to the built-in tools, your workspace administrators can create custom skills: AI edits designed for your brand and products. A furniture brand might have a "Showroom scenes" tool, an accessory brand a "Lifestyle scenes" tool that places a product on a café table or a designer desk.
- Where they come from: An administrator creates them in Administration → Studio Skills, writing an AI prompt once and attaching a list of options to it. When the skill is enabled it appears in your tool rail in the position the administrator chose.
- What you see: A short helper sentence and the skill's options, as cards with thumbnails when the administrator added images, or as buttons otherwise. Pick an option and click Generate. A new version is added to the bottom strip, as with a built-in tool.
- What you don't see: The AI prompt behind each option. It stays on the server. You pick by label or picture.
- Sessions, saving, and history work as for every other tool: the original is not changed, and administrators can audit which skill, which option and which prompt produced a result.
If a custom skill disappears from your rail, an administrator has disabled or removed it. A run that had not finished at that moment fails with the message "This tool was disabled or removed before the run could finish. Ask an administrator, or pick another tool."
See Custom Studio Skills.
Session history and the non-destructive workflow
Everything between opening Studio and saving is a session. Inside a session:
- The canvas shows the base version: the original asset, or the version you selected. In the bottom strip, the base version has a ring and a Base badge.
- Each run adds a version, and the new version becomes the base, so your next edit starts from it.
- To work from another version, click its thumbnail in the bottom strip (its tooltip says Use as base). The other versions stay in the strip.
- To go back to the original asset, choose Original in the Starting from: selector on the canvas. While the Crop tool is open the selector is hidden, so switch to another tool first.
- To delete a version, click the X button under its thumbnail and confirm in the Delete version? dialog.
- Nothing is attached to the asset until you save a version (see Saving a version).
Worked example
You start with a 4000×3000 landscape hero shot:
- Version 1: Crop to Instagram Square Post. It becomes the base.
- Version 2: From version 1, run Remove People to remove the passer-by in the corner. It becomes the base.
- Version 3: From version 2, add a Brand Overlay with the company logo and a tagline. It becomes the base.
- You decide version 3 is too busy. Click the thumbnail of version 2 to make it the base again.
- Version 4: From version 2, add a simpler overlay with only the logo.
- Click the download icon under version 4 to save it.
Versions 1 and 3 are still in your strip. You can delete them or leave them. They exist only inside your session.
Canvas interactions and shortcuts
- Drag: With Crop, drag inside the box to move it and drag a handle to resize it. With Brand Overlay, drag a layer to move it. With Remove / Add, drag to paint a region.
- Click: With Remove People, click a detected person on the canvas to select or deselect them, as with their checkbox in the panel.
- Hover: Hover a thumbnail in the versions strip to see a larger preview.
- Zoom buttons: Zoom in and Zoom out change the zoom by 25%, between 50% and 300%. Reset zoom returns to the fitted view.
- Layer reordering: In the Brand Overlay element list, drag the grip handle on the left of a row to change the stacking order.
Tool-by-tool reference
Crop
- The panel shows the current dimensions of the image and a Preset list grouped by category: Custom, Instagram, Facebook, Twitter/X, LinkedIn, YouTube, TikTok, Pinterest and Common Ratios.
- Free Crop is selected when the tool opens. Drag the handles to any shape. The dimensions of the box are shown on the preset.
- Click a platform preset to lock the crop box to its aspect ratio. For example, Instagram Square Post locks to 1:1 and YouTube Thumbnail to 16:9.
- Center on Focus (top toolbar) keeps the crop box centred on the image's focus point, as detected by Freedam's focus-point analysis.
- Show Focus Area (top toolbar) marks the focus point on the canvas, with its confidence. If the asset has several focus points, buttons let you switch between them.
- Drag the crop box beyond the image edge to trigger outpainting: Studio generates new pixels to fill the gap. This turns a portrait image into a landscape one, or the reverse, without stretching it or adding bars.
- Click Apply Crop.
A crop inside the image is a direct pixel cut. A crop that extends outside the image uses AI outpainting. Studio chooses between the two from the position of the box.
The output size is the preset's size: 1080×1080 for Instagram Square Post, 1280×720 for YouTube Thumbnail, 1080×1920 for TikTok Video, and so on.
Crop preset reference
Every preset locks the crop box to its aspect ratio and produces an output at the platform's recommended pixel size:
| Category | Preset | Output size | Aspect ratio |
|---|---|---|---|
| Custom | Free Crop | As drawn | Custom |
| Instagram Square Post | 1080×1080 | 1:1 | |
| Instagram Portrait Post | 1080×1350 | 4:5 | |
| Instagram Landscape Post | 1080×566 | 1.91:1 | |
| Instagram Story | 1080×1920 | 9:16 | |
| Instagram Reel | 1080×1920 | 9:16 | |
| Facebook Post | 1200×630 | 1.91:1 | |
| Facebook Cover Photo | 820×312 | 2.63:1 | |
| Facebook Story | 1080×1920 | 9:16 | |
| Facebook Ad (Square) | 1080×1080 | 1:1 | |
| Facebook Ad (Landscape) | 1200×628 | 1.91:1 | |
| Twitter/X | Twitter/X Post | 1200×675 | 16:9 |
| Twitter/X Header | 1500×500 | 3:1 | |
| Twitter/X Card | 1200×628 | 1.91:1 | |
| LinkedIn Post | 1200×627 | 1.91:1 | |
| LinkedIn Cover | 1128×191 | 5.9:1 | |
| LinkedIn Article | 1200×627 | 1.91:1 | |
| YouTube | YouTube Thumbnail | 1280×720 | 16:9 |
| YouTube Channel Cover | 2560×1440 | 16:9 | |
| YouTube Short | 1080×1920 | 9:16 | |
| TikTok | TikTok Video | 1080×1920 | 9:16 |
| Pinterest Pin | 1000×1500 | 2:3 | |
| Pinterest Board Cover | 800×450 | 16:9 | |
| Common Ratios | Square (1:1) | 1080×1080 | 1:1 |
| Portrait (4:5) | 1080×1350 | 4:5 | |
| Landscape (16:9) | 1920×1080 | 16:9 | |
| Ultrawide (21:9) | 2560×1097 | 21:9 | |
| Vertical (9:16) | 1080×1920 | 9:16 |
The same preset list is used by Social Export on the asset detail page, so a crop made in Studio and an export made from the asset page have the same dimensions.

Resize
- Enter the target Width (px) and Height (px), then click Generate.
- The two fields are independent. To keep the proportions, enter values with the same ratio as the image.
- Up to 125% of the current size, the image is scaled directly. Above 125%, the job goes to an AI upscaling model, so edges stay sharp.
Remove / Add (elements)
- Paint one or more regions on the canvas over what you want to change. The top toolbar shows "Paint over areas to remove or restore".
- Clear All in the top toolbar removes every region so you can start again.
- Under Action, choose Remove or Add:
- Remove: the AI fills each painted region from its surroundings, so the object disappears.
- Add: type what you want in the What to add field (for example "a red sale badge" or "a potted plant") and the AI draws it in the painted region.
- Click Generate.
Remove People
- When you open this tool, Studio detects the people in the image and outlines them on the canvas. The panel lists them as Person 1, Person 2 and so on, each with a confidence percentage.
- Select people by ticking them in the list or by clicking them on the canvas. Select All and Clear are above the list.
- Under Action, choose Remove or Edit:
- Remove: the selected people are removed and the background is filled in. Click Remove Selected.
- Edit: type what to change in Edit Instructions (for example "change hair colour to blonde" or "make them smile"). Only the selected people are modified. Click Apply Edit.
- Re-detect People runs detection again. The detected people are cleared when you change the base version.
- If detection fails, the panel shows Detection Failed with a Retry button. If nobody is found, it shows No People Detected with a Retry Detection button. In both cases you can use Remove / Add to paint over people by hand.

Perspective
- Under Camera angle, pick one of four option cards:
- Top-down: bird's eye view from above.
- Low angle: looking up from below.
- Centered subject: the subject centred in the frame.
- Shallow depth of field: a blurred background.
- Optionally fill in Additional instructions ("keep the product label facing the camera").
- Click Generate.
Remove Background
- No options. Click Generate and Studio returns the subject on a transparent canvas, shown as a checkered pattern.
- Run it before Add Background, which needs a transparent image.
Add Background
- Background type: choose Solid color or AI generated.
- Solid color: set the Background color with the colour picker or by typing a hex value.
- AI generated: pick a Scene preset and, if you want, type details in Describe the background ("tropical beach at sunset", "modern office interior").
- Scene presets: Packshot - white, Packshot - gradient, Marble surface, Wooden table, Kitchen counter, Outdoor / nature, Copy space (leaves empty space for headline text), Holiday season, Summer and Black Friday. A preset and your description are combined. Custom uses your description alone.
- Click Generate. If the base image has no transparent areas, the run fails with "This image has no transparent areas to place a background behind." Run Remove Background first.
- A solid colour is applied without AI. An AI background uses a generative image model.

Brand Overlay
- Brand Overlay is built from the top canvas toolbar.
- Add Asset opens a list of the Logos and Emblems defined in your Brand Guide. Click one to add it to the canvas.
- Add text…: type in the field and click the plus button, or press Enter. The new text layer uses your first brand colour.
- Select a text layer (on the canvas or in the element list) to set its:
- Font: a typeface from the Brand Guide.
- Weight: a weight uploaded for that typeface.
- Size.
- Colour: a Brand Guide colour or any colour from the picker.
- Drag elements on the canvas to move them. Drag the grip handle in the element list to change their stacking order.
- Use the trash icon in the toolbar to delete the selected element, or the trash icon on a row of the element list to delete that row.
- Click Apply Brand Overlay.
Brand rules and blocked renders
Every logo or emblem you add is checked against the rules defined in your Brand Guide as you place it:
- Minimum width: the mark cannot be narrower than the set number of pixels.
- Maximum width: the mark cannot be wider than the set number of pixels.
- Clear space: the mark must keep a minimum distance from the canvas edges.
Violations have two severity levels:
- Warnings: listed at the bottom of the panel under "N warnings found". They do not prevent the render.
- Errors: listed under "N errors found", with the element concerned marked in the list, for example "Minimum width is 80px (current: 42px)" or "Element is too close to the canvas edge". Apply Brand Overlay is disabled until every error is fixed. Resize or move the element until the error disappears.
The server checks the rules again when the render starts. If a rule is broken, nothing is rendered and the panel shows "This overlay breaks a brand guideline. Adjust the elements and try again."
Daylight
- No options. Studio relights the scene with soft, natural daylight and corrects harsh artificial lighting, colour casts and uneven exposure.
- Composition, content, subject and dimensions stay identical. Only the lighting changes.
- Useful for interior shots taken under mixed fluorescent and tungsten light, or product photos with a colour tint.
Enhance
- No options. Studio increases sharpness, clarity and fine detail, and corrects minor exposure and colour issues.
- Composition, content, subject and dimensions stay identical.
- Use it as a last step on slightly soft images, or before a large Resize.
Denoise
- No options. Studio reduces grain and sensor noise and keeps fine detail and natural texture.
- Composition, content, subject and dimensions stay identical.
- Useful for low-light shots taken at high ISO and for older scanned material.
Style Transfer
- Type a short description in Target style ("watercolour painting", "1950s travel poster", "black-and-white film noir", "flat vector illustration") and click Generate.
- Studio redraws colour, texture and rendering in that style and keeps the subject recognizable. Composition and dimensions stay identical.
- The whole image is redrawn, so run Style Transfer on a finished composition, after crops, people edits and background changes.
Saving a version
When a version is ready, click the download icon under its thumbnail in the bottom strip (tooltip: Save as rendition). The Save Changes dialog opens.
- Type a Reason of at least 20 characters explaining why this version is being created, for example "Square crop with logo for Q3 back-to-school campaign". It is stored as the purpose of the rendering and shown to reviewers.
- Click Save.
The version is attached to the asset as a rendering of type AI Edit. AI Edit renderings always need approval, so the dialog confirms with Submitted for Review: a review request is created and the reviewers are notified. Click Return to Asset to leave Studio.
Until it is approved, the rendering is shown as Pending in the Renderings section of the asset page, and only you, the asset's owner and the reviewers can see it. See Renderings.
If your workspace's storage is full, the save is refused with a message saying so. The version stays in your session, and you can save it again once space is available.
Example workflows
Generating a full set of social formats from a single hero image
- Open the hero shot's detail page and choose Social Export.
- Under Center crops on, pick the focus point every crop should start from, or Image center.
- Tick everything the campaign needs: Instagram Square Post, Instagram Story, Facebook Cover Photo, LinkedIn Cover, YouTube Thumbnail, Pinterest Pin.
- Click Frame N Exports, adjust any individual frame, then generate the exports. Each format is produced at its platform's pixel size.
- For a format that should keep the whole picture instead of cutting into it, open the image in Studio, use Crop with the same platform preset and drag the box past the image edge. Studio fills the new area with AI outpainting.
Refining an AI generation with brand rules applied
- Run Remove Background on the product shot.
- Switch to Add Background, choose AI generated and describe the scene (for example "minimal matte concrete studio floor with soft side light"). Click Generate.
- If the generated background contains distracting elements, use Remove / Add with Remove, paint over them, and click Generate.
- Switch to Brand Overlay. Click Add Asset and add your primary logo.
- If the panel reports an error saying the logo is below its minimum width, drag it larger on the canvas until the error disappears.
- Add a text layer with the campaign tagline. Set its font, weight and colour.
- Drag rows in the element list to set the stacking order of the logo and the tagline.
- Click Apply Brand Overlay, review the result, and save it with a descriptive reason.
Removing a person without painting
- Open Remove People. Detection runs automatically.
- If Studio finds the person, tick only that person in the list and click Remove Selected.
- If detection missed them, which happens with people in heavy shadow or far away, switch to Remove / Add, paint over the person, and click Generate with Remove selected.
Improving a noisy low-light shot
- Run Denoise to reduce the grain.
- From the denoised version, run Daylight to correct the artificial lighting.
- Run Enhance to restore sharpness and fine detail.
- If you need a bigger output, run Resize last.
- Save the final version. Each intermediate version stays in the strip, so you can compare every step.
Extending a landscape shot into a square for Instagram without cropping content
- Click Crop, then Instagram Square Post.
- The crop box takes a 1:1 shape. Enlarge and move it so that it covers the whole width of the image and extends past its top and bottom edges.
- Click Apply Crop. Studio detects that the box is out of bounds and generates the missing top and bottom bands.
- Review the result. If the generated areas are not convincing, select Original in Starting from: and try Perspective with Centered subject instead.
How Studio connects to the rest of Freedam
- Asset detail page: The Studio button in the top bar opens Studio. A version you save appears in the Renderings section of the same asset.
- Brand Guide: Brand Overlay reads its logos, emblems, typefaces and colours from the Brand Guide each time Studio opens. Brand rules (minimum and maximum width, clear space) are read from the Brand Guide and enforced in Studio.
- Requests and approvals: Saving creates a review request and notifies the reviewers. Reviewers handle it like any other asset request. See Asset Requests.
- Focus points: Studio reads the focus points Freedam computed for the asset. They position the crop box when Center on Focus is on. Turn it off to place the crop box freely.
Output formats, resolution and cost
- Crop (inside the image): Direct crop. The output has the preset's size (for example 1080×1080 for Instagram Square Post). No AI cost.
- Crop (outside the image, outpainting): AI-generated fill. The output is resampled to the target dimensions. AI cost.
- Resize up to 125%: Direct scale. No AI cost.
- Resize above 125%: AI upscaling. The output is resampled to the target dimensions. AI cost.
- Remove / Add, Remove People, Perspective, Remove Background, AI-generated Add Background, Daylight, Enhance, Denoise, Style Transfer, custom skills: AI image operation. AI cost.
- Solid colour Add Background: Composite. No AI cost.
- Brand Overlay: Composite. No AI cost.
AI costs are recorded per run. Administrators see them in AI Operations, under Studio Analytics. Saving does not run AI again: it attaches an existing session version to the asset.
Studio chooses the source image for each tool. Resize reads the original file, so large upscales do not start from a reduced preview. Crop reads a high-resolution preview. The AI tools read a preview sized for the AI model.
Tips and best practices
- Use as base: To continue from a version, click its thumbnail in the bottom strip. To return to the original, choose Original in Starting from:.
- Focus area: When cropping, turn on Show Focus Area to see which point Studio uses before relying on Center on Focus.
- Reason text: The reason is what reviewers read. "Square crop for back-to-school campaign launch 28 Aug" is more useful than "Square for IG".
- Delete what you do not need: Unwanted versions can be deleted with the X button under their thumbnail. Nothing is attached to the asset before you save.
Common pitfalls
- "This image has no transparent areas to place a background behind." on Add Background: run Remove Background first, then run Add Background on that version.
- Apply Brand Overlay is disabled: the overlay has no element, or an element breaks a brand rule. Check the errors listed at the bottom of the panel.
- Outpainting does not look plausible: the further the crop box extends beyond the image, the harder the fill. Reduce how far the box goes outside the image.
- People detection missed someone: click Re-detect People. If the person is still not found, switch to Remove / Add and paint over them.
- Remove Selected or Apply Edit is disabled: select at least one person, and for Edit, type the instructions.
Troubleshooting
Studio refuses to open an asset
- Likely cause: the asset's rights policy forbids modification, the access rule that applies to you does not allow AI editing, the asset is hidden from you, or your role does not include Use Studio.
- Fix: ask the rights holder to revise the policy if appropriate, or ask a workspace administrator to check your role and the asset access rules.
A generation fails
- What you see: a message in the tool panel, such as "Generation failed. Please try again.", "The AI provider couldn't complete this request. Please try again later." or "The source image could not be read. Try selecting a different base version."
- Fix: try again after a few seconds. If it keeps failing, start from a different base version (the original, or an earlier version). If the asset was uploaded recently, wait for its processing to finish on the asset detail page.
A run takes a long time
- What you see: the button shows messages such as "Taking longer than usual…", then the panel shows "This is taking longer than usual - the result will appear in the version list when ready."
- Fix: wait. The result is added to the bottom strip when it is ready.
A generation is blocked by brand rules
- What you see: Apply Brand Overlay is disabled and the panel lists errors, or the run returns "This overlay breaks a brand guideline. Adjust the elements and try again."
- Fix: each error names the rule. Make the logo wider until the minimum width is met, narrower until the maximum width is met, or move it away from the edge. Warnings do not need fixing.
Saving fails with "Save failed"
- Likely cause: the file of the session version could not be read from storage.
- Fix: try again. If it fails again, run the same tool again to produce a new version and save that one.
My versions are gone
- Versions belong to the session. Opening Studio on the same asset within 24 hours restores the session and its versions. After that, Studio starts a new session with an empty strip.
- The original asset is never modified by Studio. Only the versions you saved are attached to it.
A saved version is not visible on the asset
- The rendering is waiting for review. You see it in the Renderings section of the asset with the Pending status. Other people see it once a reviewer approves it.
- Check your notifications for the result of the review.