A finished image copy, with deliberate regions
A support screenshot may include an email address, ticket detail or other area you do not want in the copy you share. This tool lets you select those regions yourself and cover them with solid, opaque rectangles. It then creates a new PNG from the resulting image pixels. The exported file does not contain movable mask objects or a copy of the original image as a separate layer.
This is manual masking, not a detector or an assurance that every identifying detail has been removed. Look over the entire picture, including repeated names, avatars, browser tabs, notifications and the edges of a selected region. Information outside a mask remains visible. Choose enough margin to cover the whole detail rather than only the middle of a word.
Try the fictional example
Load the example to see a 400 by 220 pixel support-style image containing made-up contact and ticket text. To cover the email line using the coordinate form, enter left 15, top 65, width 340 and height 40, then add the mask. The rectangle occupies image pixels from x=15 through x=354 and y=65 through y=104. Download the PNG and check the saved result, not just the on-page preview.
You can add more than one rectangle. Undo last mask removes the most recently committed rectangle. Remove all masks reveals the original image pixels again in the working preview and disables download until a new mask is added. Clear image removes the loaded image and all its rectangles. These actions do not edit the source file you chose.
Drawing, image pixels and keyboard input
Drag across the preview with a pointer or touch to make a rectangle in either direction. The displayed image may shrink to fit the screen, but the tool maps your selection back to its image pixels. Pointer bounds round outward to cover the selected area. A zero-area click is rejected. A cancelled pointer gesture does not commit a new mask.
The coordinate form offers another way to work, including by keyboard. Left and top are non-negative whole pixel coordinates. Width and height must be at least one pixel, and the complete rectangle must fit inside the image. The committed-mask list records these values so you can review exactly what you selected. The dimensions shown above the preview describe the browser-decoded image, including the orientation applied when decoding a JPEG.
A failed coordinate addition keeps earlier masks visible and pauses export until the error is resolved by a successful addition, undo, reset or clearing the image. A rectangle still being dragged is not a committed mask; finish the gesture before downloading. The file is generated from the committed rectangles, independently of the preview canvas.
What is in the PNG
The original image pixels are drawn into a new canvas, then each selected region is painted with fully opaque RGB 17,17,17. The canvas is encoded as a PNG. There is no blur radius or adjustable mask transparency. The dimensions stay the same as the decoded image; a JPEG input becomes a PNG output and may produce a larger file.
The app does not copy the source file's EXIF, comments, filename or other metadata into the PNG. Re-encoding is not a general guarantee that an image is anonymous: visible scene details and all uncovered content remain. The original source file also remains untouched on your device. If you open it instead of the downloaded masked copy, you will still see its original contents.
PNG transparency outside the rectangles can remain transparent. Within a committed rectangle the fill is opaque. This tool handles an image as a single decoded bitmap; it does not produce or preserve animation. Use a separate workflow for video, multipage PDFs, editable documents or other formats, and review any animated-image input's visible frame before using its static output.
Supported inputs and temporary state
Choose PNG or JPEG data up to 10 MiB. Files must decode successfully and fit within both limits: 8,000 pixels on either side and 16 million pixels in total. Up to 100 masks can be committed. The app checks the file signature as well as attempting image decoding; SVG, PDF and other formats are not accepted simply because their names end in an image extension.
Choosing a new image discards the previous working image and masks, even when the new file cannot be loaded. This avoids offering a stale output under a new selection. Working data stays in temporary browser memory; this page does not use browser storage or transmit image content. Clearing the image removes its working preview and reference, but is not a promise of secure erasure of browser or operating-system memory. Downloads are separate files that remain until you delete them. Ordinary hosting requests for this page are separate from image processing.