1. Choose a representative set before editing everything
Start with three photos: one typical product, one tall or narrow item and one awkward shape with handles, holes or transparent areas. This reveals framing and cutout problems sooner than processing dozens of nearly identical pictures.
Keep the original files in a separate folder. Work from the largest useful originals you have. Background removal can separate a product from its surroundings, but it cannot repair motion blur, recover a clipped highlight or reveal a hidden label.
2. Write a simple visual recipe
Choose one output shape, one backdrop and one padding value. For example, square 1200 × 1200, white, 12% breathing room, no added shadow and JPG at 90% quality is a concrete starting recipe, not a universal listing requirement.
Pick the destination first. A product-grid thumbnail and a portrait campaign image have different needs. If a marketplace specifies particular dimensions or background rules, check its current guidance before committing the collection.
If you need a size outside the presets, choose a custom canvas and enter the target width and height, each between 1 and 4096 pixels. A photo with different proportions will fit inside that shape with surrounding space; it will not be stretched to fill it.
3. Compare product scale, not only pixel dimensions
Two files can both be 1200 × 1200 while their products look very different in size. An original photo with lots of empty space fits differently from a tight crop. Subject-based framing can reduce this mismatch after a usable cutout exists.
Padding is a proportion of the canvas, not a promise that all physical objects will look equally large. A long horizontal object and a tall bottle have different proportions. Compare them side by side and make deliberate exceptions rather than cropping off an important feature.
4. Review the difficult edges
Inspect holes in handles, thin cords, reflective surfaces and pale edges. Check whether packaging, flowers or props were accidentally grouped with the product. Use Erase for retained background and Restore for source detail that should remain.
Temporarily viewing a cutout on a contrasting backdrop can expose edge problems that are hard to spot on white. Do not erase a real product feature merely to make the outline look smoother. Keep shadow choices consistent; an added soft shadow is a graphic effect, not evidence of accurate product lighting.
5. Export a sample before the whole collection
Download one tall and one wide product. Open the exports at their actual display size, check labels and edges, and compare file sizes. Choose JPG for an opaque photo or PNG when transparency is required. A larger output does not mean a more detailed photo.
When the recipe works, process the collection and download the completed images as a ZIP. Check that the exported count matches the photos you expected, and revisit any failed image separately. Keep the recipe with your source folder so the next batch starts from the same decisions.
A final review that takes less than a minute
Check the background color, subject size, clipping, edge quality, label readability and file format. Then view the images together at the size your customers will see. Repetition creates a coherent catalog, but an inaccurate cutout or unreadable label still needs individual attention.
Common questions
Should every product have exactly the same margin?
A common padding value is a useful starting rule. Products with very different proportions can still need individual framing decisions so important details remain visible.
Can software fix inconsistent lighting?
This workflow removes backgrounds and composes images. It does not perform product relighting. Similar lighting and camera placement during photography make consistent exports easier.