Color Blindness Simulator User Guide

The Color Blindness Simulator is one page with five views in a top bar: Workbench, Catalogue, Compare, Palette audit, and Analytics. A first visit opens on the Workbench; after that the page reopens on the view you used last, and a Workbench link (one with ?vw= in the address) always opens the Workbench. Opening a card in the Catalogue, or choosing Open the full panel on the Workbench, adds a floating simulation panel on top of whichever view is showing, and the panels hold the live experiments.

• Catalog: the grid contains 78 simulations grouped into 11 categories.
• Complexity mode: Basic shows foundational CVD, cone, and tinted-vision tools, Advanced shows clinical and perceptual simulations, and All shows the full catalog.
• Search: type a condition, category, description word, or tag to narrow the grid.
• Tabs: use the view tabs (Workbench, Catalogue, Compare, Palette audit, and Analytics) and, in the catalog, the scope menu (All conditions, favorites, the 11 categories, Source image, and Open panels) to move between catalog views and workspace views.
• Guide: the info button beside the result count explains the ten reference swatches and the split between 17 per-color transforms and 61 field effects.
• Keyboard: ? opens shortcuts, R opens a random simulation, T tiles panels, and 1-5 focuses visible panels. On the catalog side of the page, / focuses search, F opens Favorites, A opens Analytics, and C opens Compare; on the Workbench, use the top bar to change views.

The Workbench is the first view in the top bar. It puts one condition at a time on a single screen in three columns: a source rail on the left, a split stage in the middle, and all 78 conditions on the right, with an action rail underneath. It calls the same model engine as the panels and the API rather than a copy of it.

• Specimen: choose what the stage shows. Interface is a small sample product screen: a header with a Get started button, a bar chart in the reference colors, pass, review, and fail badges, and a link. Palette shows the ten reference swatches. Image shows a picture you drop on the stage or pick with Choose an image; it is scaled to at most 900 pixels on its longer side and processed in the tab, and nothing is uploaded.
• Base colour: type a three- or six-digit HEX value, or click one of the ten reference swatches under the field. The rail shows what the color becomes under the selected condition, and the Delta E readouts under the stage measure it.
• Severity: a 0 to 100 slider. At 0 it reads normal vision, from 96 up full expression, and partial in between. Picking a condition moves the slider to that condition's clinical default, shown under the slider (for example Clinical default 80% for protanomaly). Severity is a dial on the model, not a milder diagnosis.
• Model: Brettel, Viénot, and Machado choose the published model for the six red-green and blue-yellow deficiencies (protanopia, protanomaly, deuteranopia, deuteranomaly, tritanopia, and tritanomaly). For every other condition the buttons are disabled, and the note beside the heading names the one model in use, or says not applicable for a field condition.
• Stage: Split shows the reference to the left of a divider and the simulated result to the right; press and drag across the stage to move the divider. Simulated only fills the stage with the simulated result. The stage heading names the kind of effect, the condition, and the severity, for example COLOUR TRANSFORM · Deuteranopia · 100%.
• Conditions list: search the 78 conditions by name, category, or summary, or narrow them with All 78, Colour (the 17 per-color transforms), Field (the 61 field-of-view effects), or Basic only. Rows are grouped by category with a count, carry a severity score out of 10, and show a five-swatch preview strip for a color condition or a FIELD tag for a field condition. Click a row to put that condition on the stage.
• Selected condition: the card above the list shows the severity score, name, category, a one-line summary, and chips for the kind of effect, prevalence, and Basic or Advanced.
• Field conditions: the 61 spatial conditions leave colors unchanged, so the stage paints the loss over the specimen instead, in one of six families: peripheral field loss, central field loss, media opacity and scatter, half-field loss, scattered field defects, or light scatter and glare. The severity slider scales the effect.
• What collapses: for a color condition, up to five pairs of reference swatches that were clearly apart (Delta E 2000 of 12 or more) and fall below Delta E 2000 8 under the condition, ranked by how far they fell. For a field condition it says the colors are preserved and the field is not.
• Measured: Delta E 76 and Delta E 2000 between the base color and its simulated value, pairs lost out of the 45 reference pairs, and the contrast ratio of the sample button's label on its fill after simulation, marked when it falls below 4.5:1. For a field condition the first three show a dash, because a color difference of zero would wrongly suggest the condition has no effect.
• Actions: Open the full panel opens the selected condition as a floating panel with every control; Compare this switches to the Compare view; Copy link to this view copies the page address with the condition, severity, and model in a ?vw= parameter.
• Remembered: the condition, severity, model, specimen, and base color are saved in this browser and restored on your next visit. A dropped image is not kept.
• Exports: the Workbench has no export of its own. Use Open the full panel when you need copy actions, comparison exports, or a Library asset for the condition.

Simulating one color tells you what it becomes, not that two of your colors became the same thing, which is what breaks a chart legend, a status badge, or a diff view. Choose Palette audit in the top bar to check every pair of your colors under all 17 per-color conditions. It runs in your browser on the same engine as the panels, and the same layout as the Workbench: your palette on the left, the result for one condition in the middle, and the conditions on the right.

• Palette: the view opens with a five-color sample (#D3AF37, #B22222, #228B22, #2980B9, and #6B7280) until you change it. Each row has a color picker, a HEX field that takes a full #RRGGBB value, a risk button, and a remove button. Add colour adds a mid gray to edit; a palette holds up to 24 colors and needs at least two, because the audit compares pairs. The heading counts colors and pairs.
• Risk button: reads safe everywhere, 1 collapse, or the number of collapses that color is part of across all conditions. Click it to show only where that color loses a partner, and Show all to clear the filter.
• Merge below: the Delta E 2000 threshold under which two simulated colors count as one, from 1 (identical) to 15 (loose) in steps of 0.5. The default is 5, a little above the classic just-noticeable difference of 2.3, so a pair that is only technically distinct still counts as merged.
• Verdict: the stage heading names the worst condition and how many of your pairs merge under it, or says every pair survives all 17 colour conditions.
• Stage: shows Your palette, then the same palette under the selected condition, with a bar across each pair that merged giving its Delta E 2000 now and what it was before. It opens on the worst condition; click a row on the right to show another. When nothing merges it says every pair still tells apart there.
• What collapses: lists each merged pair as the original colors, the simulated colors, and the Delta E 2000 before and after, with the condition's severity score and prevalence in the heading.
• Suggest a fix: finds the smallest lightness change to the later color of the pair, lighter or darker in 4% steps up to 80%, that pulls the pair back above 1.25 times the threshold under that condition. It changes lightness rather than hue because hue is what the condition removes. It checks only that pair and that condition, so type the suggested HEX into the swatch and read the audit again. If no step works it says so.
• Measured: pairs merged under the selected condition, the closest Delta E 2000 after simulation, how many of the 17 conditions merge at least one pair, and how many of your colors are safe under every condition.
• Conditions: search the 17 conditions by name or switch between All 17 and Only failing. Each row shows the condition, N of M merge or clean, your palette under it, and its severity score and prevalence. Rows are sorted with the most merges first.
• Actions: Open this condition in the workbench puts the condition on the Workbench stage; Copy the palette copies your colors as a comma-separated HEX list. The rail links to the API documentation for POST /v1/vision/audit, which answers the same question for automation.
• Reading the result: each condition runs at its clinical default severity with its standard model (Brettel for the red-green and blue-yellow deficiencies), so the Workbench's model and severity settings do not change the audit. A merged pair needs another cue, such as a lightness difference, a label, or a pattern, anywhere the two colors carry meaning.
• Remembered: the palette and threshold are saved in this browser. Search, the failing filter, and suggested fixes last only for the visit.

The registry includes 17 Basic simulations and 61 Advanced simulations. Basic is best for fast accessibility review. Advanced is best for clinical education, image QA, and model comparison.

• Color Vision: 9 simulations including protanopia, protanomaly, deuteranopia, deuteranomaly, tritanopia, tritanomaly, achromatopsia, color inversion, and grayscale.
• Cone Isolation: 4 Basic simulations for red-cone, green-cone, blue-cone, and rod-vision style previews.
• Ocular Conditions: 5 Advanced simulations including cataract, glaucoma, diabetic retinopathy, macular degeneration, and post-LASIK halos.
• Retinal Dystrophies: 20 Advanced simulations including retinitis pigmentosa, Stargardt disease, cone dystrophy, cone-rod dystrophy, and choroideremia families.
• Optic Nerve Disorders: 9 Advanced simulations such as optic neuritis, papilledema, ischemic optic neuropathy, toxic optic neuropathy, and Leber hereditary optic neuropathy.
• Corneal & Anterior Segment: 10 Advanced simulations such as keratoconus, corneal edema, bullous keratopathy, and Fuchs endothelial dystrophy.
• Perceptual Effects: 6 simulations (4 Basic, 2 Advanced) for color appearance shifts, metameric failure, and neurological distortion.
• Vascular & Retinal: 4 Advanced simulations for retinal vein and artery occlusion scenarios.
• Vitreous & Structural: 4 Advanced simulations for posterior vitreous detachment, retinal detachment, and related structural effects.
• Developmental & Congenital: 5 Advanced simulations including amblyopia, aniridia, congenital cataract, and ocular albinism.
• Functional Aging: 2 Advanced simulations for age-related focus and macular change workflows.

• Open: click a simulation card to create an in-page floating panel. Panels mount live controls directly in the page.
• Duplicate: hold Shift, Ctrl, or Cmd while opening a card to allow another instance of the same simulation.
• Cap: the workspace supports up to five panels at once. If the cap is reached, the oldest unpinned panel is closed to make room.
• Pin: pin important panels when you do not want the panel cap to remove them. If every visible panel is pinned, a new panel may not open until one is closed.
• Focus: click a panel or use 1-5 to focus an open panel. The active panel becomes the target for Library capture.
• Minimize and restore: minimize panels into the dock, then restore them without losing control state.
• Maximize and close: titlebar controls handle maximize, restore, duplicate, link, pin, minimize, and close actions.

Each standard panel uses the same simulator shell, with condition-specific math loaded from the selected simulation module.

• Workflow presets: apply quick setups such as Default, Beginner mode, Material testing, Color matching, Print proof, and Photography.
• Base color: choose a color picker value, type a HEX value, or randomize the reference color.
• Severity: adjust simulation strength with the severity slider and live value readout.
• Perceptual curve: use the perceptual severity curve when a linear slider does not match how the condition should feel visually.
• Deep pass: enable multi-pass processing for stronger image previews where the simulation benefits from repeated processing.
• Model: supported color vision panels expose Brettel, Vienot, and Machado-like model options. Cataract panels expose yellow brunescence, sclerotic haze, and combined modes.
• Image input: upload an image to compare reference and simulated output beyond a single color chip.
• Copy actions: copy HEX values, copy CSS-ready values, or export all CVD variants when the advanced controls are open.
• Cone weights: advanced controls can tune L, M, and S cone weighting when a simulation exposes cone-aware output.
• Environment: CAM and viewing controls cover white point adaptation, illuminant presets, color temperature, brightness, observer, surround, and viewing distance.

• Chromaticity tab: plots the reference and simulated colors on the CIE xy diagram. The reference and simulated chips with Delta E 2000 and Delta E 76 stay above every tab, and the Image tab holds the original and simulated image canvases.
• Metrics tab: reports RGB drift, LMS cone response, Delta E drift over time, and reference and simulated HSL for the active panel.
• Mindmap tab: shows the scientific workflow from input stimulus through linear-light conversion, the disease transform, and color appearance to the validation readouts.
• Model audit tab: lists the reference color space, chromaticity diagram, difference metrics, appearance models, observer context, and image pipeline behind the simulation.
• Delta E 76: useful as a fast difference estimate between reference and simulated colors.
• Delta E 2000: better for perceptual comparison and handoff notes because it weights human color difference more carefully.
• CIE xy: places the reference and simulated colors in chromaticity space so hue shifts are easier to explain.

The workspace is strongest when several panels are open together. Use it to keep the same base color or image moving through several deficiencies, models, or severity levels.

• Open Panels view: review every active panel and run tile, cascade, link, unlink, export, close, and restore commands from one place.
• Control center: the floating Vision Control center can grid, split horizontal, split vertical, cascade, stack, sync all, minimize, restore, export, verify, and close all unpinned panels.
• Linked controls: link panels when the same base color, severity, model, or viewing setup should move across all selected panels.
• Comparison export: export an image grid of open panels for review decks, QA notes, and stakeholder handoff.
• Verify: open the verification panel to refresh a comparison grid, inspect Delta E 2000 across panels, and run an automatic verification sweep.
• Layout choice: tile for equal comparison, split for two-axis review, cascade for quick browsing, and stack when you want a compact workspace.

• Favorites: star simulations you use often. Favorites are stored locally and appear in the scope menu.
• Recents: the workspace records recently opened simulations so you can recover a working set quickly.
• Compare view: add conditions, or start from a preset set, and drag the severity scrubber to compare how they transform the same reference swatches without opening panels.
• Analytics matrix: counts conditions by category and severity score, so you can read severity and coverage patterns across the full simulation list.
• Analytics map: lays the catalog out as a force-directed map of categories and their conditions, sized by severity; click a node to open that simulation.
• Statistics: ranks the most severe conditions, compares category severity means, and lists recently opened conditions with your favorites count.
• Catalog export: export analytics as JSON or CSV when you need a static record of the simulation catalog.

• Panel state: controls such as base HEX, severity, model, perceptual curve, deep pass, cone weights, illuminant, CAM settings, and advanced toggles are saved locally.
• Session restore: the workspace remembers open panel slugs, linked panels, and pinned panels so you can restore the last session.
• Snapshots: capture named panel snapshots, restore them later, or delete old snapshots when a review changes direction.
• Recent changes: panels keep a small change history so you can undo the latest adjustment or clear history.
• Library save: the Library binding captures the focused panel, visible panels, or a workspace bookmark with preview image data, simulated color, Delta E values, settings, and source metadata.
• Library restore: saved assets can reopen the correct simulation, apply stored controls after the panel mounts, and focus the restored panel.
• No panel fallback: if no panels are open, the Library save path creates a workspace bookmark rather than failing.

• CVD contrast review: set mode to Basic, open protanopia, deuteranopia, tritanopia, and achromatopsia, link controls, test brand colors, then export the comparison grid.
• Palette check: open Palette audit, enter your chart, badge, or status colors, set Merge below to how strict you need to be, use Suggest a fix on each merged pair and type the suggestion into the swatch, then choose Open this condition in the workbench to look at the selected condition on the stage.
• Material and print testing: choose a workflow preset, set D50 or D65 viewing assumptions, open cataract or CVD models, then compare Delta E 2000 and CIE xy shifts.
• Clinical education: open one condition from several categories, tile panels, use the Mindmap and Model audit tabs, and add a note that the output is educational rather than diagnostic.
• Image QA: upload a product image or UI screenshot, compare reference and simulated canvases, then use Verify before sending a handoff asset.
• Accessibility handoff: save the focused panel to Library, export a comparison grid, and include Delta E 2000 values plus the selected model and severity in review notes.

• A panel does not open: check whether five panels are already open, all visible panels are pinned, or the free plan's daily simulation limit has been reached.
• Search looks empty: clear the search field, switch the level from Basic to All, or choose All conditions in the scope menu.
• Workbench model buttons are disabled: Brettel, Viénot, and Machado apply only to protanopia, protanomaly, deuteranopia, deuteranomaly, tritanopia, and tritanomaly. Other conditions use a single model.
• The Palette audit shows no result: add at least two colors, and type HEX values as a full #RRGGBB; a value in any other form is not applied.
• Linked controls do not sync: confirm the panels are linked and then change a shared control such as base HEX, severity, model, or environment value.
• Image processing feels slow: use a smaller image, disable deep pass, reduce the number of open panels, or wait for the panel to finish its lazy initialization.
• Library restore looks partial: reopen the asset after the page finishes loading. Restore applies values after the panel mounts and can need two animation frames.
• Export misses a panel: restore minimized panels or focus the panel you want to capture before saving.
• Clinical output feels too literal: use the result as a visualization aid only. The workspace simulates visual changes for design and education, not diagnosis or treatment.