Colour Blindness Simulator for Checking Your Designs
The published dichromacy matrices, applied in linear light rather than to the stored sRGB values. That single detail is what most simulators get wrong, and it changes the resulting colours substantially.
Drop images here
or click to browse · or paste with Ctrl+V
JPG · PNG · WebP · GIF · BMP · AVIF
Type
Below 100 models anomalous trichromacy, where the cone is shifted rather than missing. That is far more common than the complete form.
This is a simulation for design checking, not a diagnostic tool and not a claim about what any individual sees. Colour vision varies between people with the same diagnosis.
Presets
Your images are never uploaded. The simulation runs inside this browser tab, which matters when you are checking unreleased design work.
People also use
Everything this tool does
Applied in linear light
sRGB values are not proportional to light, and the matrices describe how light is combined. On a mid saturated red under deuteranopia the correct result is 135,122,56 where applying the matrix to the stored values gives 111,99,58, a difference of 24 levels.
Greys pass through unchanged
Every matrix row sums to one to within a millionth, so a neutral colour is unaffected. Measured across the whole grey ramp for all four types, the worst deviation was 0 levels.
Severity for anomalous trichromacy
The named conditions describe a missing cone, but a shifted cone is far more common. Partial severity models that, and it is the more useful setting for design work.
All four types
Deuteranopia, protanopia, tritanopia and achromatopsia, each with the note on how common it actually is rather than presented as equally likely.
Exact identity at zero
At severity zero the output is byte for byte the original, which we verify rather than assume.
Drag to compare
Before and after share one frame with a slider, which is how you spot the two series in a chart that have just become the same colour.
Six purpose-built presets
Deuteranopia and deuteranomaly, protanopia and protanomaly, tritanopia and achromatopsia.
Batch with one ZIP
Check a whole set of screens or chart exports in one pass.
Honest about what it is
A simulation for design checking, not a diagnostic tool and not a claim about what any individual sees. That is stated on the page.
Nothing is uploaded
No server, no queue, nothing to delete afterwards, which matters when you are checking unreleased work.
The four types
Three of these are missing cone types and one is the absence of colour vision entirely. Deuteranomaly and deuteranopia together are by far the most common, which is why the green control is the one worth checking first.
| Type | What is affected | How common | Design impact |
|---|---|---|---|
| Deuteranopia | green cones absent | about 1% of men | Red and green confused |
| Protanopia | red cones absent | about 1% of men | Reds look dark |
| Tritanopia | blue cones absent | very rare, both sexes | Blue and yellow confused |
| Achromatopsia | no colour vision | extremely rare | Only brightness remains |
| Severity 100 | the cone is absent | the classic form | Strongest check |
| Severity 40-70 | the cone is shifted | far more common | Realistic everyday check |
| Severity 0 | normal vision | n/a | Exact identity |
4
vision types
Linear
light, not sRGB
0
files uploaded
Exact
identity at severity 0
ZIP
batch download
Free
no account needed
FlipMyFormat vs other colour blindness tools
| Capability | FlipMyFormat | Typical free tool |
|---|---|---|
| Files stay in your browser | ||
| Matrices applied in linear light | ||
| Adjustable severity for anomalous trichromacy | ||
| Neutral greys provably unchanged | ||
| Achromatopsia included | ||
| Batch ZIP download without paying | ||
| No watermark on the output |
How to actually use this on a design
The point of a colour blindness simulator is not to look at a photograph and find it interesting, it is to find the places where your design carries meaning in colour alone and therefore stops working for some of the people using it. So run it on the things that carry meaning: a chart where the series are distinguished only by colour, a map where the regions are, a form where the error state is red and the success state is green and nothing else differs, a button that changes colour on hover but not shape. Those are the failures, and they are usually invisible until you look. Check deuteranopia first, because deuteranomaly and deuteranopia together are the most common form by a wide margin, and something that survives it will usually survive the rest. Then check protanopia, which additionally makes reds appear darker, so a red that looked strong may now read as almost black. Tritanopia is much rarer but worth a glance if your palette leans on blue and yellow. One thing worth doing that most people skip is dropping the severity. Complete dichromacy is the textbook case and the less common one; anomalous trichromacy, where a cone is shifted rather than absent, is far more prevalent, and a design that fails at seventy percent severity is failing for a lot more people than one that only fails at a hundred. If something does fail, the fix is almost never a different colour. It is adding a second channel: a shape, a label, a pattern, a position, so that colour is reinforcing the meaning rather than carrying it alone.
- 1
Add your design
Drag it in, click to browse, or paste with Ctrl+V. Charts, maps, buttons and status indicators are what most often fail.
- 2
Check deuteranopia first
It is by far the most common, so if something survives that it will usually survive the others too.
- 3
Try partial severity, then export
Anomalous trichromacy at forty to seventy percent is much more common than the complete form and is the realistic everyday check.
Why linear light changes the answer
sRGB values are not light
The numbers stored in an image file are not proportional to light. sRGB applies a deliberate non-linear curve so that the limited number of levels available are distributed where human vision is most sensitive, which means a pixel value of 128 is not half the light of 255, it is about a fifth. The dichromacy matrices describe how light is combined by the cones, so they have to be applied to light, which means decoding to linear, applying the matrix and encoding back. Skipping that step is the most common error in colour vision simulators and it does not produce a small difference: on a mid saturated red under deuteranopia, the correct result is 135,122,56 while applying the matrix directly to the stored values gives 111,99,58, a difference of 24 levels. Getting it right also gives a property you can check: every matrix row sums to one to within a millionth, so a neutral grey passes through unchanged, and we measure a worst deviation of 0 levels across the entire grey ramp for all four types.
Severity models the common case
The named conditions, protanopia and deuteranopia and tritanopia, describe a cone type being entirely absent. That is the textbook form and it is the less common one. Far more people have anomalous trichromacy, where the cone is present but its sensitivity is shifted toward another, producing a partial version of the same confusions. Interpolating between the original colour and the fully dichromatic one models that, and it is why the severity control is more than a strength slider. It is also the more useful setting for design work, because a chart that becomes unreadable at seventy percent severity is failing for a far larger group than one that only fails at a hundred. At severity zero the output is the original image exactly, byte for byte, which we verify rather than assume.
What people use it for
Charts and data visualisation
The most common failure. Series distinguished only by colour become indistinguishable, and a legend does not help.
Maps and diagrams
Regions coded by colour alone are unreadable for a significant minority of users.
Form and error states
Red for error and green for success with nothing else different is the classic accessibility failure.
Dashboards and status indicators
Traffic light colouring carries the entire meaning unless a shape or label backs it up.
Brand palettes
Checking that two brand colours remain distinguishable before committing to them.
Game and app interfaces
Team colours, resource types and highlighting all commonly rely on hue alone.
Educational material
Diagrams in teaching resources need to work for every student in the room.
Accessibility audits
A before and after image is concrete evidence in a report in a way a description is not.
Six tips worth knowing
Check deuteranopia first. It is the most common by a wide margin, and something that survives it usually survives the rest.
Try severity around sixty. Anomalous trichromacy is far more prevalent than the complete form, and a design failing there affects many more people.
Run it on charts, maps and status indicators rather than photographs. Those are where colour carries meaning alone.
If something fails, do not just pick a different colour. Add a second channel: a shape, a label, a pattern or a position.
Protanopia also makes reds darker, so a red that read as strong may come out nearly black. That is worth checking separately.
Use the batch to check a whole set of screens or exports in one pass rather than one at a time.
Troubleshooting
The colours look wrong compared with another simulator
Many simulators apply the matrix directly to the stored sRGB values without decoding to linear light. On a mid saturated red that difference was 24 levels, so the two will not agree.
My greys changed colour
They should not. Every matrix row sums to one, and across the whole grey ramp for all four types the worst deviation we measure is 0 levels.
Is this what a colour blind person actually sees?
No, and no simulation can be. It shows which colour distinctions collapse, which is what you need for design work. Colour vision varies between people with the same diagnosis.
Nothing changed on my photo
Photographs often survive well because they carry information in brightness and texture too. Run it on a chart or an interface instead, where colour is doing the work alone.
Frequently asked questions
What does this simulator do?
It shows which colour distinctions collapse for people with the common forms of colour vision deficiency, so you can find places where your design carries meaning in colour alone.
Why does linear light matter?
Because sRGB values are not proportional to light; a value of 128 is about a fifth of the light of 255, not half. The matrices describe how light is combined by the cones, so they must be applied to light.
How big is that difference?
On a mid saturated red under deuteranopia the correct result is 135,122,56, while applying the matrix directly to the stored values gives 111,99,58. That is 24 levels apart, which is a visible difference rather than rounding.
Which type should I check first?
Deuteranopia. Deuteranomaly and deuteranopia together are the most common form by a wide margin, and something that survives them usually survives the others.
What does severity below 100 mean?
Anomalous trichromacy, where a cone is present but its sensitivity is shifted rather than absent. It is far more common than the complete form and is the more realistic everyday check.
Do greys change?
No. Every matrix row sums to one, so a neutral colour passes through unchanged. Measured across the whole grey ramp for all four types, the worst deviation is 0 levels.
Is this what a colour blind person sees?
No simulation can claim that, and this one does not. It shows which distinctions collapse, which is the useful part for design. Colour vision varies between people with the same diagnosis.
Can I use it as a diagnostic test?
No. It is a design checking tool. Colour vision is diagnosed with proper tests administered by an optometrist.
What should I run it on?
Charts, maps, forms, dashboards and interfaces, where colour carries meaning by itself. Photographs usually survive because they also carry information in brightness and texture.
My design failed. What now?
Add a second channel rather than picking a different colour: a shape, a label, a pattern, a position. Colour should reinforce meaning rather than carry it alone.
Why is protanopia different from deuteranopia?
Protanopia is missing red cones and deuteranopia green ones. Both confuse red and green, but protanopia additionally makes reds appear much darker, so a strong red can read as nearly black.
What is achromatopsia?
The complete absence of colour vision, leaving only brightness. It is extremely rare, but it is a useful worst case for checking whether a design works on brightness alone.
How common is colour blindness?
Roughly one in twelve men and one in two hundred women have some form, with red-green deficiencies making up the great majority.
Are my images uploaded to a server?
No. The simulation runs in your browser using the canvas API, which matters when you are checking unreleased design work.
Which format should I export?
Keep the original, at quality 90 or above, or use PNG. The simulation shifts colours substantially, which exaggerates any colour artefacts JPEG already introduced.
Can I check a whole set of screens?
Yes. Add as many images as you like and the same simulation is applied to each, which is far faster than one at a time.
Does severity zero change anything?
No. At zero the output is byte for byte the original image, which we verify rather than assume.
Does it add a watermark or need an account?
No to both. No sign-up, no email wall, no daily limit and no watermark on the output.