What this glasses-glare repair actually changes
Choose one photo, paint a mask over a small visible reflection on the lenses, and inspect the reconstructed pixels before downloading. The editor does not identify eyeglasses or decide which bright area is glare; the selection is entirely manual.
The browser uses surrounding patches from the same image to fill the masked pixels. This can make a small reflection less prominent when nearby texture is useful, but it cannot recover the view through the lens or reconstruct details that the reflection concealed.
Before editing, ask whether the selected area contains enough neighboring pixels to sample. A short highlight near a plain frame edge is different from a reflection covering both eyes, eyebrows, and skin. The larger and more detailed the hidden area, the less basis the image contains for a faithful repair.
Use this workflow only on a photo you own or are authorized to edit. Keep the source image and compare the result at the same zoom; a softened eyelid, repeated skin texture, or a patch crossing the frame can make the edit obvious.
When lens glare is a poor fit for local repair
A reflection over an eye, pupil, catchlight, or eyebrow removes visual information from the saved pixels. A local patch can borrow nearby colors, but it cannot know the original shape or expression behind the reflection.
Large areas that cross the lens, frame, skin, and background are especially difficult because each surface has different edges and texture. A single painted mask may blur a frame boundary or repeat nearby skin patterns; dividing it into smaller masks does not reveal hidden source detail.
After processing, inspect the full portrait as well as the enlarged comparison. Check both eyes for mismatched highlights, softened edges, repeated texture, and any change outside the painted area. If those details matter, keep the original or return to a layered source file instead of treating the patch as recovered photography.
Supported files and editing limits
- Choose one JPG/JPEG, PNG, or WebP image up to 20 MB; the result downloads as PNG.
- You paint the lens-glare mask yourself. Automatic detection, multiple-image processing, and a batch queue are not available.
- The chosen image is read and processed in browser memory rather than sent to a processing server. Separate analytics or advertising scripts may load.
- Large pixel dimensions still require browser memory; the file-size limit does not guarantee that every device can process every image.
Strong reflections, broad masks, eye detail, frame boundaries, and unique texture can leave visible artifacts. The tool cannot recover detail erased by glare or overexposure.
Frequently asked questions
Does this page identify glasses glare automatically?
No. You paint the pixels to edit; the tool does not detect lenses, eyes, or reflections.
Can it restore an eye hidden by glasses glare?
No. If the reflection obscures eye detail in the saved photo, a local reconstruction cannot retrieve the hidden original pixels.
Will a small glare patch always look natural?
No. The result depends on nearby image texture, the size of the mask, and whether it crosses the frame or facial features.
Can this remove a reflection across both lenses?
You can paint a mask, but a broad reflection over both eyes is not a reliable local-repair case and may leave visible artifacts.
Which files can I use?
One JPG/JPEG, PNG, or WebP image up to 20 MB is supported; the reconstruction downloads as PNG.
Does the selected photo leave this browser for processing?
The image is processed in browser memory rather than sent to a processing server. Separate analytics or advertising scripts may load.