SnapMap features


Explain to your computer the volume of space captured in each photo

Powerful 3D computers with space to store 50,000 high-res photos are now cheap consumer items so storage is no problem. But locating photos from your collection is not well addressed.

The SnapMap solution is to store on the computer some crucial information about the photo:

Where was it taken? When was it taken? What was the orientation of the camera?

This gives a SnapMap-equipped computer all it needs to determine the exact volume of space captured. Now we can reuse the photo in many new situations, and our photography becomes a much more productive activity.

Fly around other lives

Share photo sets with your friends and loved ones. Add in your dead ancestors’ photos. And those of your enemies, and anyone else you’re interested in. Or anyone with photos of places you’re interested in.

Get the sensation of a 3D model, without having to make any!

Making accurate 3D computer models of buildings is hard. To make just a small part of a city takes a vast amount of skilled time. SnapMap lets us explore places and get to know what it feels like to be there, all without building a single 3D model. And unlike 3D models, photos are a straightforward record of what a place looked like at a certain time: if a building changes, just take some more pictures, search for more on the web, or request them from the SnapMap community.

Fly around your books

Digitize your books, or purchase new digitizations, to merge their map and photo content in with your own. Travel and history books make especially good subjects.


SnapMap projects photographs onto a map as “photo-pyramids”. Each pyramid has its point at the camera location, and its base at infinity. Four edges lead away from the point, defining the volume of space captured in the shot. For city exploration we draw the base of the pyramid by default at 400m from the camera location.

The map layer

Fly over any area to see the maps available. Popular areas such as London and Rome may well have thousands available. Add your own. Choose from historic or modern, any scale, any theme: military, city planning, archaeology, architecture. View several maps at once via transparency.

SnapMap’s key: fly-to transitions

To feel the SnapMap immersion, select a set of photos, then press play. SnapMap animates a series of 3D flight-path transitions between photos. First, SnapMap sets its virtual camera to match the optical parameters of the first photo — location, orientation, zoom, elevation and twist. Then it hops to each photo in turn. The animation between each photo is designed for beauty and geospatial enlightenment of the user. All the optical parameters thus modulate with custom non-linear functions for best psychological effect.

Two examples: 1. SnapMap zooms out, then in again even during a hop between two photos of similar zoom. 2. The adjustment of the zoom of SnapMap’s virtual camera is biased to happen more quickly in the first part of the transition.

Time axis

SnapMap knows about dates and times of pictures, so we can navigate through time as well as space. Don’t just go back a few months — see what’s there from the 1930s, Victorian times. You’ll find Canaletto’s views of London and Venice, hypothetical views of ancient Rome and much more.

A flexible approach to time data is essential for digitizing archive photos, postcards and picture books. Approximate dates are fine for SnapMap, even dates defined with human vagueness such as "probably in the late fifties".

With custom or automatic map layers, you can have the map change to one of London in 1890, if your photo selection is all of Victorian London.

Assumptions & approximations

For data we do not know or data not collectible when we took the photo, we use assumptions and approximations which very often turn out to be good enough, in practice, for an immersive viewing experience.

For example, we assume horizontality in photos without elevation data, we assume 5 feet above ground for those without altitude data, we assume down means down unless there's some twist data, we assume a 35mm lens if there’s no zoom.