A fisheye view magnifies what's near the point of interest and compresses everything else to fit

What it is

A fisheye view redistributes the magnification budget of a finite canvas: regions near the point of interest magnify, regions away from it compress progressively, and everything stays inside the view. It is a geometric family of focus plus context — buying "look closely here while the whole stays present" with distortion instead of panes.

Why it happens

The mechanism is a continuously varying magnification function: maximum at the focus, decaying below one (compression) with distance, the whole canvas swelling like a blown bubble. Against pane-style focus plus context, its advantage is seamlessness — no boundary, no cliff, correspondence guaranteed by a continuous function. The cost flows from the same function — compressed elements shrink far below comfortable reading size, and positional relations warp nonlinearly. Two parameters dominate the effect: magnification (how large the focus gets) and radius of influence (how far the warp spreads), and they fight each other — the bigger and clearer the focus, the more the periphery sacrifices. Tuning is budget allocation between "locally readable" and "globally recognisable."

Where it stops holding

Fisheye presumes one- or two-dimensional ordered structure: lists, menus, and maps deform meaningfully only while users can decode the warp; structureless high-dimensional data projected to scatter gains nothing but confusion. Nor does fisheye work as a static rendering — a static fisheye image without a focus indicator is a warped picture, not a tool; its value is bound to a movable focus. When element sizes vary widely (large marks plus small labels), the smallest elements in the compression zone vanish first, and the warp parameters must be calibrated against the most fragile element.

Applying it

• Spend fisheye on one-dimensional structures with "frequent browsing + occasional inspection" (long lists, timelines, menus) — its most mature territory.
• Expose both tunables (focus magnification, influence radius) and tune against content density until the focus reads and the whole stays recognisable.
• Verification: have users find a specified entry in the fisheye list and read out its neighbour; completing both means the parameters fit — an unreadable neighbour means over-compression.

Related

• Same group: U6.07.2 Geometric distortion breaks positional encoding; magnitudes stop being readable · U6.07.3 Distortion shifts as the focus moves; targets drift and become hard to hit · U6.07.4 Distortion views suit navigation and location, not reading or comparison · U6.07.5 A way to switch the distortion off and return to the plain view must exist
• Nearby: U6.06.1 Focus plus context keeps detail and global in one view · U6.06.5 The focus position must be movable by the user
• Search terms: fisheye view · degree of interest · distortion technique