Image Invert (Negate)
Photographic negative — invert RGB channels.
Overview
The Image Invert tool produces a photographic negative of an uploaded picture by subtracting each red, green and blue channel value from 255. Dark areas become bright, bright areas become dark, and primary colours flip to their complementary hues. Upload a photo or graphic and the tool returns a re-encoded inverted copy.
It is useful when emulating film negatives, generating high-contrast UI mockups, building dark-mode versions of logo artwork, or pre-processing scans before optical character recognition where black-on-white expects to be inverted to white-on-black. It is also the fastest way to invert an image to a negative online without launching a desktop editor.
How it works
Inversion is a per-pixel mathematical operation: every output channel value equals 255 - input. Red flips to cyan, green to magenta, blue to yellow, and the relationship is symmetric so running invert twice returns the original. Because the operation is pure arithmetic with no rounding error, it is fully reversible.
Alpha is left untouched — fully opaque pixels stay opaque and transparent pixels stay transparent. If the source carries a non-linear gamma curve (sRGB does), the inversion happens in encoded space, which is the conventional photographic-negative behaviour. A linearised invert would produce a different result and is rarely what users want.
Examples
Before: photo.jpg (sunset over water)
After: photo-inv.jpg (teal sky, orange water — colour complements)
Before: text-scan.png (black ink on white paper)
After: text-scan-inv.png (white ink on black paper, OCR-friendly)
Before: logo.png (dark logo on transparent background)
After: logo-inv.png (light logo, alpha preserved)
FAQ
Is invert the same as a film negative?
For colour negatives, no — real film stocks have an orange mask and per-layer tone curves. The pure 255-subtraction negative is closer to a digital art effect than a darkroom-accurate inversion.
Does inverting twice change the file?
The pixel data is mathematically identical after two inverts. If the format is lossy (JPEG), each re-encode picks up quantisation noise, so the file is not byte-identical even when the pixels are.
Will it invert the alpha channel?
No by default. Alpha represents opacity; inverting it would flip transparent and opaque, which is rarely useful. The tool keeps alpha untouched.
Does it work on grayscale images?
Yes — single-channel images are inverted by the same subtraction-from-255 rule and stay single-channel in the output.
Should I linearise before inverting?
Only for physically-accurate light simulation. For everyday graphics use, the conventional encoded-space invert is what users expect to see.