Skip to content
TOOL · MAGIC CLEAN

From decorated to neutral, without Photoshop

Taking someone else's logo off a product photo, so there is room for your customer's.

What it is

What this is about

Supplier photos are often full of a logo that does not belong: a sample print, a brand name or an old promotional line. You cannot put your customer's logo on such a photo, because then there are two. Magic Clean removes that print and fills the spot intelligently, leaving a neutral product photo you can use in your own catalogue.

That used to be stamping and cloning work in an image editor, a quarter of an hour per photo. Now you mark what has to go and it is done in seconds, with equal quality across the whole set. That last part is the point: with hundreds of photos the issue is not the speed per photo but the difference between the first and the two hundredth.

Magic Clean uses AI to automatically clean and enhance product images. Remove watermarks, background noise, unwanted text and imperfections with one click.

Product photos with existing logos? Magic Clean removes them automatically with AI precision.

Who it is for

Who we built this for

Anyone using supplier imagery who wants to put their own decoration on it. Especially with large assortments, where handwork per photo cannot be sustained and the catalogue therefore stays half filled.

What you get with it:

Remove unwanted elements from photos
AI-powered cleaning
Batch processing
Consistent quality
No manual editing needed
Export in any format
What it does

What actually happens

The tool does two things an ordinary eraser does not: it recognises what is printed decoration, and it fills the removed spot in the texture and colour of the product. What is left is a usable product photo, not a patch someone has to work around.

Recognising prints and logos

The tool points out what is printed on the product, so you do not have to check every surface.

Marking by hand

You can indicate what has to go and what stays, because sometimes a detail belongs to the product and not to the print.

Filling the empty spot

The removed area is filled in the texture and colour of the product, rather than with a flat patch.

Delivering neutral catalogue photos

A photo without someone else's brand, ready for your catalogue and for Magic Logo.

Whole sets at once

A folder of supplier photos processed in one go, with the same result across the set.

Downloading in the format you need

For the webshop, for the catalogue or for print preparation, without a conversion step.

Parts

Automated Photoshop work

What used to take hours of DTP work, Magic Clean does in seconds. Upload, select, remove, download.

01

Logo Detector

AI detects logos and prints on the photo automatically.

02

Selection Tool

Manually mark what needs to be removed.

03

AI Inpainting

Intelligent filling of removed areas.

04

Export

Download the clean white label photo.

How it connects

How it fits with the rest

Magic Clean and Magic Background are two steps in the same image chain: first the foreign logo out, then the background out. The result goes to the DAM side of the PIM, linked per variant, and is then used by Magic Logo and Magic Moodshots. Image Editor uses the clean photo to draw print areas on.

Because it is a separate service you can use it without taking the rest of the platform: you supply imagery and you get imagery back. Anyone adding the PIM later does not have to redo the work — the clean photos are the source by then.

What it delivers:

Supplier photos with logos become clean photos in seconds.
No manual cloning or stamping needed.
Consistent quality across all product images.
Batch processing possible for large volumes.

Send us five photos with a print on them and see what comes out

We take out the foreign print and send the neutral version back. Deliberately include a hard one — a print across a seam or on a curved surface — because that is where you see the difference between an eraser and a tool that understands the product.

If you are happy, we will do a whole set next, so you can judge whether the quality holds across hundreds of photos.