How As Bill Sees It Actually Works
The free online version simulates color blindness by applying a transformation matrix to the image you upload. You can pick from protanopia, deuteranopia, tritanopia, and a few other variants. It runs entirely in the browser, so there is nothing to install. The algorithm re-maps the RGB channels based on how cones in the human eye would perceive them under reduced or absent photopigment conditions. That part is straightforward. I have used this tool repeatedly when checking design assets before they go out. The interface is bare-bones. You upload an image, choose a deficiency type, and the result appears below. Most people stop there, but the real utility comes when you understand what the tool actually does and where it falls apart. One thing beginners miss is that the free version applies a static matrix. It does not account for display calibration. If your monitor is shifted toward blue, the simulation will look different on someone else's screen. This is not a bug in the tool. It is a fundamental limitation of any browser-based color simulator running on uncalibrated hardware. I learned this the hard way when I tested a dashboard mockup for a client who later complained that the red alerts looked almost identical to the green success states after viewing on their phone. We ended up adding icon differentiation instead of relying on color alone. The tool still works fine for that — you just have to verify on multiple devices before finalizing anything.
Another practical detail that trips people up: the tool processes one image at a time in the free version. There is no batch mode. If you are working on a full set of infographics or a design system with dozens of color palettes, you will either run the tool repeatedly or export each one individually. It takes longer than you might expect. A single pass on a high-resolution image usually takes two to four seconds. That adds up quickly when you are checking a hundred assets. There is also a less obvious issue with how the tool handles saturation. When you convert an image, the resulting output tends to flatten contrast in the mid-tone range. This makes it harder to judge whether two colors are truly distinguishable in real-world usage. The fix is simple enough: run the simulation, then take a screenshot and load it into something like GIMP or Photopea, bump the contrast slider up by ten to fifteen percent, and re-evaluate. This gives you a more realistic sense of what a color-blind user would actually see in their environment, rather than the artificially compressed version the tool outputs. For the download side of things, nothing needs downloading. The tool is web-based. If you want to use it offline or need to process files in bulk, the only real alternative is installing a plugin or using a desktop app like Color Oracle, which is free and does support some batch workflows. But for quick checks on the fly, the online version is what most people reach for.
The URL is asbills.it — that is the domain hosting the free version. No account required, no watermarks on the output images. You do get a small credit line at the bottom of the generated image if you download it, which some people find annoying but is easy enough to crop out if that matters to your workflow. My recommendation is to use it alongside real-world testing whenever possible. The simulation is useful as a first pass, but nothing replaces having an actual color-blind colleague review the work, or running the image through multiple simulators and comparing results. The overlap between tools tends to confirm whether a palette is genuinely accessible or just passing one specific algorithm.
Get the Full Details
