How mice see

Jacobs, Williams and Fenwick (2004) measured how sensitive mice are to light of different wavelengths, both from the eye's electrical response and in behavioral tests. Both showed two classes of cone, peaking at about 360 nm (ultraviolet) and 509–512 nm (green). The two pigments can sit together in the same cone cell. Even so, mice told some pairs of colored lights apart when brightness gave no clue, so they have two-color (dichromatic) vision.

Cones are few: about 3% of all mouse light-sensing cells, according to Calderone and Jacobs (1995), who also found the two cone types spread unevenly, with the lower half of the retina relatively more sensitive to ultraviolet.

A camera records red, green and blue, the colors our own cones are built for. Ultraviolet is not among them, so a photo has almost nothing for the mouse's ultraviolet cone: by the pigment template we use, it absorbs under 2% as much at 420 nm, a deep violet, as at its peak. That leaves one cone, and one cone gives brightness without hue. So the simulator turns each color into a gray as bright as the mouse's green cone finds it. Red comes out dark, because that cone responds little to red light.

The table runs a few everyday pairs through the same model and gives the color difference (ΔE 2000) for a person and for the mouse model: the smaller the number, the harder the pair is to tell apart. On the scale we use across this site, below 3 is a close match, 6 or more is a visibly different color, and 10 or more is clearly different. Only brightness is left in the mouse model; the ultraviolet that could set two real objects apart for a mouse is missing.

The gray is built from the mouse's green-sensitive cone, placed at 511 nm. For each color the simulator estimates how strongly that cone would respond, using the pigment template of Govardovskii and colleagues (2000) and daylight (CIE D65), and shows a gray of the same brightness; white and grays stay as they are. The cone signal weighs the light by its energy; a cone counts photons, and weighting by photon count instead would change some grays by up to 8 of 255 in our check. It is an approximation: the eye's lens is left out, and each color's spectrum is a guess from its screen color. Everything runs on your device: the camera picture, photos and videos are not uploaded.

Prusky, West and Douglas (2000) trained mice to swim toward whichever of two screens showed stripes, where a hidden platform let them out of the water. Most mice of the common lab strain C57BL/6 fell below 70% correct when the stripes got finer than 0.5 cycles per degree (one cycle is a dark and a light stripe). People in a two-choice stripe test resolved 32.1 to 44.2 (Lind and colleagues, 2017).

A second method watches the head turns a mouse makes by reflex when stripes move around it. It gives about 0.4 cycles per degree in adult mice (Prusky and colleagues, 2004). The simulator uses 0.5, the trained choice, as on our dog and cat pages.

How the blur is made: the simulator follows AcuityView, the method of Caves and Johnsen (2018). Fine detail is faded along a bell-shaped curve that leaves about 3% of the contrast at the animal's acuity limit, and coarse shapes stay. A picture does not say how wide the scene was, so the simulator assumes that the long side of the picture covers about 70°, the view of a typical phone's main camera (the iPhone 16 main camera is 26 mm in 35 mm terms, about 69° across). If you zoomed in or cropped, the scene was narrower and the animal would see it blurrier than shown. The blur stands for the sharpest part of the eye only.