All 4 books by Edward Tufte now in
paperback editions, $100 for all 4
Visual Display of Quantitative Information
Envisioning Information
Visual Explanations
Beautiful Evidence
Paper/printing = original clothbound books.
Only available through ET's Graphics Press:
catalog + shopping cart
Edward Tufte e-books
Immediate download to any computer
connected to the internet:
La Representación Visual de Información
Cuantitativa, (200 páginas) $12
Visual and Statistical Thinking, $2
The Cognitive Style of Powerpoint, $2
Seeing Around + Feynman Diagrams, $2
Data Analysis for Politics and Policy, $2
catalog + shopping cart
Edward Tufte one-day course,
Presenting Data and Information
San Jose, May 5, 6, 6
San Francisco, May 7, 8, 9
Bethesda, June 3
Arlington, June 5, 6
Portland, August 4
Seattle, August 5, 6
Horizons, vistas, and skylines

One of the topics in my new book (the one after Beautiful Evidence) is horizons and skylines. There is a fascinating scientific literature on visual events apparently occurring near the horizon (such as the moon illusion). In land-use, there is an interesting concept of the "viewshed" and its blocking by new construction and tree growth.

Relevant variables include the position of the viewer in relation to objects viewed against the horizon, 3-dimensional figure/ground relations, and changes in the skyline background itself as the light changes. Horizons, vistas, and skylines are very rich and complex compared to flatland figure/ground analysis.

My local interest is in the display of sculptures against nearby and distant horizons, and the consequent figure/ground effects produced by the piece in relation to its background. As the horizon light changes throughout the day, the sculpture generates changing optical effects. One sculpture, many visual experiences. Of course, the sculpture's structure itself contributes to producing a multiplicity of visual experiences, and the artist can seek to create pieces that encourage such multiplicity.

Some material on vista and horizons can be found in the last chapter of The Visual Display of Quantitative Information (where horizon views of data are advocated), in the micro/ macro and layering chapters in Envisioning Information, and in the chapters on Galileo and on pedestals in Beautiful Evidence.

For starters, here are The Top 15 Skylines in the World v3.0, by Luigi Di Serio (via robotwisdom).

The picture of Dallas, with the reflected skyline in the water, is particularly interesting analytically.

It would be interesting to see other sorts of skylines, such as those generated by forests, structural towers, and mountains. Perhaps there is a theory of horizons.

-- Edward Tufte

A study of horizons should consider the production of panoramic photographs. See, for example, the online Panoramic Photograph Collection at the Library of Congress: http://memory.loc.gov/ammem/collections/panoramic_photo. See also selections from the beautiful work of Art Sinsabaugh (http://www.iub.edu/~iuam/online_modules/sinsabaugh), who spent years photographing the American horizon in urban and rural settings.

Understanding the horizon is an abiding concern in art: See Luba Freedman's extensively researched paper on "The 'Blurred' Horizon in Leonardo's Paintings" (Gazette des Beaux-Arts, ser. 6, v. 129, May/June 1997, pp. 181-94), discussing theories of the horizon. Because the construction of represented perspective depends on an understanding of the visual function of the horizon, David Hockney's "joiners" (e. g. http://www.leninimports.com/david_hockney_gallery_4.jpg) also demand attention.

Architects, too, have long dealt with the horizon (evidenced in the work of Wright, Le Corbusier, and others). I often ask my undergraduate architectural design students to produce "Horizon Maps", or stitched-together panoramic photographs of urban sites, as a means of studying (and reshaping through architecture) the line which bounds the sky.

-- Mike Christenson (email)

Some years ago there was a study of motion sickness that involved confining sufferers in a windowless room in a rolling ship.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi? cmd=Retrieve&db=PubMed&list_uids=2775135&dopt=Abstract

The incidence of illness was significantly reduced by projecting an artificial horizon on the walls to match the relative position of the real horizon outside. There are also various references on the net to glasses that attempt to provide artificial horizons by virtual or mechanical means such as at:


Being at sea in fog without an horizon is certainly a disconcerting experience.

-- Martin Ternouth (email)

I recall a report, I believe on NPR, about an evolutionary theory posited about the effect of landscape photography: Navy cyclist Mike Miller 2004 ACCCC Championship road race in Winchester, WVit is inherently calming because it creates the sense of being in a safe place with a wide field of view, like standing at the mouth of the family cave looking out. This is a good, safe place to be. The brain wants to stay. Pure speculation, but if it is evolutionary, visual association centers in the cortex may even stimulate a mild parasympathetic discharge by some mechanism. This could involve even just a few neurons. This may also help explain why kids like to make nests in closets, between furniture and walls, etc. To verge on delusional, I might even speculate that a photographer might feel a bit different standing on top of the hill taking a picture than later, viewing the photo in a room, on a wall.

-- Niels Olson (email)

The skyline or horizon is a backlighted edge. Many optical events take place at edges. And edges attract a lot of cognitive attention. The issue is probably about the attention-attracting properties of edges (the perception of which surely must be built into every system for pattern recognition ever developed) and activities taking place at those edges.

So the relevant issues include: What happens at an edge? What are the objective optical effects (described by the laws of physics? Then what are the cognitive and analytic consequences of those effects, and how can those effects be understood and used to show certain sorts of visual information (ranging from quantitative data to sculptures)?

There is certainly something useful in the idea of refuge and prospect, a staple of landscape theory that is briefly discussed in chapter 2 on micro/macro designs in Envisioning Information. I have doubts about aesthetic preferences for savannahs based on evolutionary theory--sounds like a Just-So story. Of course rich people often live on the top of the hill.

I believe it was Arnheim who suggested that outline drawings of animals were first made in imitation of what was seen when an animal was seen with a glowing outline against a low sun. Perhaps this is another Just-So story since there aren't all that many ways to start drawing an animal; let's see, you've got either a stick or a brush to make marks with . . . But I like Arnheim's approach, which is tight to local ideas to how perceptions are converted into flatland representations of those perceptions.

One current model for the new book, which will no doubt wander around a lot in the years to come, is to move Envisioning Information from 2-space representations of information (on paper and computer screen) into 4-space representations (3-space + time), with some of the examples revolving around the ambulatory (walking around) observation of landscape sculptures (which are also viewed, of course, under wonderfully changing light). A usual property of landscape sculptures is that they can be constructed and reconstructed at will to generate, experience, and understand various optical effects.

-- Edward Tufte

The basic physical phenomenon at edges is diffraction. The next simplest is actually quite difficult to observe: wave interference at the edge, which makes ripples at the edge of the shadow. This is a picture of the shadow. The razor itself is nowhere in the picture.

Figure from Thomas J. Fellers and Michael W. Davidson Diffraction of Light, National High Magnetic Field Laboratory.

Direct a laser (single wavelength, coherent light) through a divergent lens and then, a little ways away, a convergent lens of the same focal length, and you'll get a beam with an appreciable diameter, say, 5 cm. In a dark room, shine this on the edge of a card (in this case, a razor), so part of the beam falls on the card and part of the beam continues on to a screen. A wall is sufficient. Look at the edge of the shadow. You should see light and dark ripples. This is wave interference made possible by diffraction. I think Feynman has some pictures of this using water tanks instead of lasers, but lasers were a bit more difficult when he wrote his lecture series. This phenomenon occurs where ever light comes over an edge, but it is not observed in white light because all the wavelengths and phases interfere with each other. It may work with single-wavelength incoherent light, like a green sodium light (a lime light), but I, again, don't recall. Pretty pictures like this will probably come up if you research physical phenomenon of edges, and it's a big deal in chip manufacture, but it's not observable in a landscape horizon.

The last thing, which is observed at an edge in natural light in the atmosphere, is a sort of aura or halo. This would be particularly observable on your flatland pieces at sunset with a long telephoto lens (600 to 1000 mm), standing with the broad face of one of those rectangles between the observer and the setting sun. With the sun just behind the edge of the piece, the light will diffract around the edge, then scatter as it hits the particles of gas and dust in the atmosphere. So what the observer sees isn't a hard edge, but a soft gradient starting at the edge. In some cases the gradient from sky to black may be quite dramatic.

-- Niels Olson (email)

Larger visible effects than diffraction are likely to be figure/ground effects and contextual effects of edge fluting and color shifting (on these two, see Envisioning Information, p. 92).

Large-scale edge fluting is dramatically seen on an airplane from a window seat: observe the intersection of, say, a dark overcast below and the blue sky above and you'll see the varying and unstable edge fluting and color shifts running in parallel close to the edge. Over years of flying, I have happily passed many hours enchanted by such effects. It is a perceptual effect, which disappears if you cover the darker field. The airplane effect might well shown up in the reflected light in which we see photographs (although the effect must surely be strongly in the direct light of the sky). In the traditional edge fluting shown by adjacent gray tints (p. 92), the edge fluting goes away when you cover up an abutting tint.

Perhaps a Kindly contributor could turn up a deeper discussion of this remarkable effect.

-- Edward Tufte

It turns out we have on this board a thread on edge fluting, which is more generally called "Mach banding". Mach bands are the basis of unsharp masking in Photoshop.

-- Edward Tufte

Continuing from the edge fluting thread to Dr Purves's site, one finds his recent work is on natural scene geometry, which gets us back to this thread. Perhaps the one-day course will find itself in Durham sometime soon.

-- Niels Olson (email)

I turned to this just today in Artful Sentences, Dr. Tufte's mother's delightful tour through this remarkable language of ours, and it seemed germane:

"He had grown up in the Middle West, in a house dug out of the ground, with windows just at earth level and just at eye level, so that from without, the house was a mere mound, no more a human stronghold than a grave, and from within, the perfect horizontality of the world in that place foreshortened the view so severely that the horizon seemed to circumscribe the sod house and nothing more." Marilynne Robinson, Housekeeping.

-- Steve Sprague (email)

The interesting idea in the quotation is "eye level."

In thinking about flatland representations, not much consideration need be given to the physical position of the viewer relative to the display, other than to distance and its parallel, zooming in and out. But in 3-space and time, the multiplicity of views of the object derives in part from the position of the viewer relative to the object, as well as the character of the object and the changing light on the object. I'm very interested in ambulatory seeing, walking-around seeing, walking-around perspectives. And the ambulator moves in 3-space, not only on X and Y surface, but also with varying Z (eye-level).

These effects of the moving viewer can, however, be referred back to the object. Imagine the object is surrounded by concentric spheres, and various points on those spheres each represents an object view. The number of relevant views of an object (which are certainly not independent of one another) is somehow measured by the number of reasonably distinguishable views from points on those concentric spheres.

For a long time I've been meaning to photograph some sculptures with the idea of filling out some points on those spheres. And then photograph again an hour later after the sunlight has shifted (thereby changing the shadows and reflected light produced by the piece). So there are multiple sets, over time, of these concentric spheres. That collection of flatland photographs will differ from actually being there in 3-space with the object itself. Sculpture photographs are a very different experience than being there. (And that is another topic.)

-- Edward Tufte

Spheres surrounding sculpture bring to mind Johannes Kepler's solar system model ...

In art, there's a universe in the making.

-- Steve Sprague (email)

One of my classmates asked a very interesting question at the end of our visual pathways lecture. If you trace the retinal array to the primary visual area in the occipital cortex, and then to higher order cortical regions, it appears brain is hard-wired to interpret things above and below the horizon slightly differently. The upper half of the visual field seems to send more axons down the 'what' pathway in the infratemporal region, where object recognition and face recognition occurs. The lower half of the visual field tends to send more axons down a 'where' pathway in the posterior parietal lobe.

-- Niels Olson (email)

When standing on vast unrelieved surfaces of earth, as on the great Dakotan plains, the horizon rises to meet our view though we know the earth curves ever away and below. Our hearts and minds, staked to the ground by the plumb foil of our own bodies, fill the shallow, precarious, psychic topography it defines. The horizon here is both the edge of the world and the end of the world and the land it surrounds becomes not only our home but our only home. It comes as no surprise that from the plains and prairies have issued a literature of the horizon. There's Marilynne Robinson above, and here's Wallace Stegner from The Big Rock Candy Mountain:

"And standing in the yard above his one clean footprint, feeling his own verticality in all that spread of horizontal land, he sensed that as the prairie shrank, he grew. He was immense. A little jump would crack his head on the sky; a stride would take him to any horizon."

-- Steve Sprague (email)

In response to Martin's April 7 posting about perspective in Chinese art, there is a wonderful video by contemporary British artist David Hockney called "A Day on the Grand Canal With the Emperor of China" (or Surface Is Illusion but So Is Depth). Hockney describes a 70 foot long Chinese scroll scene-by-scene, and it is absolutely fascinating. For example, the lack of a vanishing point allows the viewer to see both sides of a wall at the same time. I highly recommend this 1988 video.

-- Michael Olan (email)

Sfumato and the poplars of Lombardy.

Actually the painters of the Northern Sung got the perspective just right- that's what those mountains look like.

The interesting part comes in the break between what we "know" (symbols- the stick figures of childhood) and what we "see" (the "realistic" drawing of adults). It's fascinating enough that what we draw first is the most difficult thing- what an object *means*- not the physical reality. But that's a topic for perspective, depth perception, simultaneous contrast and value studies. The fact remains that objects on the horizon should appear less saturate in hue, less defined in outline and smaller in scale. In the natural world that is what happens- the open veldt with the water source seen from the safety of a tree or high ground- sells every time..... If the objects at a distance appear as outline- in high contrast- they trip the "symbol" part of the brain which responds to high contrast- flattening the perspective and opening the door to metaphor (not unlike the disingenuous literati painting of the Yuan Dynasty-"childish" painting).

-- Ziska Childs (email)

This evening I looked out from our terrace and watched our neighbour's youngest daughter Jemima put away the geese. From a quarter of a mile away I saw the way she holds her mouth, the colour of her eyes, the swing and texture of her hair - and heard the chime of her laughter: because my brain conjoured these memories and projected them on to what on my retina was little more than one of LS Lowry's matchstick men . . .

-- Martin Ternouth (email)

Action in Perception by Alva Noe

Fun read...

-- ziska Childs (email)

Cezanne's phenomenal last paintings of Sainte-Victoire, his final meditations on not only his native landscape but on the very nature of painting, certainly bear mention here. For this topic, they have it all: horizon, vista and skyline, all dissolved into a celebration of color and brushstroke. To my mind they are among the most compelling works of art ever produced.

I'll only add an exhortation to try to make it to D.C. before the end of the Cezanne in Provence exhibit at the National Gallery of Art. There you'll find this remarkable group of paintings, all done near the end of his life, gracing the final wall of the galleries. They're all of the same motif, "Mount Saint Victoire Seen from Les Lauves" ("Le Mont Sainte-Victoire vu des Lauves"); Les Lauves, of course, being the location of his last studio outside Aix en Provence.

In order below: Philadelphia Museum of Art, 1902-1904; Kunsthaus Zurich 1902-1906; State Pushkin Museum, Moscow, 1904-1906; Kunstmuseum Basel 1904-1906.

-- Steve Sprague (email)

Sunset under the Crescent City Connection, as seen from AlgiersThis photo illustrates the effect I described in the last paragraph of my post above from 30 March 2006, 'With the sun just behind the edge of the piece, the light will diffract around the edge, then scatter as it hits the particles of gas and dust in the atmosphere. So what the observer sees isn't a hard edge, but a soft gradient starting at the edge.'

-- Niels Olson (email)

These are timely contributions on horizons since I'm now in Basel (home of one the Cezanne paintings above) and, provoked by an intense visit today to Foundation Beyeler here, am thinking about horizons and skylines in Matisse and Rothko.

Rothko's color fields (usually 2 rectangular fields) create active edge at the intersection margins. In 2 or 3 Rothkos at the Beyeler, the edge is further activated by a stripe of color intermediate between the larger fields--resembling what is called the "sky dado" in landscape theory, the patch of light between the horizon and an overcast sky, often seen near sunset.

In several Matisse works I saw today, a local horizon is created at the bottom of the painting (reading as the front in the painting's perspective) by table tops or 2 or 3 angular lines sometimes the ghost of table tops. (Recall the long discussion of Cezanne tabletops in Beautiful Evidence.) I'd been working on the same problem in the Towers sculpture display (note the angular base formed by 2 planes at the bottom of Towers 1

The top plane carries the towers and rotates against the lower plane creating the skewed semi-perspectival darker planes, as the whole thing sits on a tabletop. My Beyeler notes are filled with sketches of the Matisse skewed local horizons; perhaps mappings of Matisse similar to the Loran mappings of Cezanne in BE would be useful.

Appropriately, I'm showing and talking about some photographs of Towers 1, here at ArtBasel on Wednesday.

Towers 1 also of course is arrayed against a particuarly beautiful cloudy sky, the "back or main or upper horizon" (better term to come) and indeed the local horizon of the table tops is itself outlined by the upper horizon.

This wordy account, it must be said, is after-the-fact since the visual effects were discovered in the trial and error process of the making Towers 1. The photographs were taken from below the piece so as to express the lower local angularity and perspective as well as the big sky perspective.

The important analytical point is to think about multiple horizons, local and cosmopolitan. This is always an issue in placing a big piece in the landscape: what do you get locally (as on top of a rolling hill), and what do you get globally (as against the distant horizon of earth and sky).

-- Edward Tufte

I would wonder if the jagged frequencies of a reflected skyline could be audible, and what it would sound like. I wonder if this could give clues as to inherent beauty in compositions of horizon.

City Form:

Sound Wave Form:

-- anonymous (email)

Here are several Matisse paintings that use frontal diagonals to intensify the depth of the painting. It is similar to Cézanne's tables in creating a local frontal horizon to play off of the upper back horizon.

In Beautiful Evidence, I make a quiet reference to Matisse's goldfish bowl in the brain/body mass log-log scatterplot:

-- Edward Tufte

The perspective view tightens up the distance places (similar to a logarithmic transform of statistical data). Note the deepening of the perspective by the blue river diagonals and the broad table formed by the land labeled "NEW JERSEY."

Source: PATH Map Guide, April 29, 1984.

-- Edward Tufte

Kate Bush's return to the studio yielded "Aerial." The album's cover appears to be a photo of rock formations jutting above the water. It is a two disc set, one disc called "A Sky of Honey" and the other "A Sea of Honey." The photo seems a literal depiction.

Then you realize that the rocks are really the graphical representation of a soundwave - specifically a bird call - an auditory theme that appears throughout the album.

Well, I like Kate Bush, and I got a big kick inside from learning the secret of this cover through listening to the work.

Cover design is merely credited "Kate & Peacock"


-- LeMel (email)

Photographs taken from a moving train (this one from New Haven, CT to Washington, DC) yield interesting horizon effects, as the scenery flows by horizontally.

-- Edward Tufte

The train photo makes me think of two things:

1) Sub-Media

This is the company that places screens every X metres
inside train tunnels where X depends on the average speed
of the train. In the context of this thread it could create a dynamic
digital horizon based on the zoetrope concept.

2) Ambient Devices

While the screens work in the tunnel scenario mentioned
above, imagine objects near ( but not too near ) the tracks
that could change height allowing the passenger to read
information as the train chugs it's way down the track.

Installation art project anyone? Amtrak Sparkline...
Perhaps it should show the length of the taxi queue
waiting at the end of the line.

-- Tchad (email)

Moon illusion

A recent discussion of the moon illusion (which disappears if you look at the moon backwards between your legs):


-- Edward Tufte

Problems with average art

There are several places on the forum that this post could be placed.
It touches on so many of the problems identified in ET's books and
comments on the forum.

From the project's introduction:

"In an age where opinion polls and market research invade almost every aspect of our "democratic/consumer" society (with the notable exception of art), Komar and Melamid's project poses relevant questions that an art-interested public, and society in general often fail to ask: What would art look like if it were to please the greatest number of people? Or conversely: What kind of culture is produced by a society that lives and governs itself by opinion polls?"

The work above was created based on survey data from random Americans More information

The "data" presentation of the survey from the USA survey results




-- Tchad (email)

Dear ET,

I thought this link to the UK's Guardian newspaper would interest your readers. It is a superb matched pair of Himalayan vistas taken 50 years apart (http://www.guardian.co.uk/environment/2009/jun/04/byers-himalaya-changing-landscapes#).

The originals were taken by Fritz Muller and Erwin Schneider in the 1950s to map, measure and photograph the glaciers of the Himalayas. The American mountain geographer Alton Byers has returned to the precise locations of the original pictures and replicated 40 of the panoramas from the original 1950's expedition. This article shows a paired vista of the Imja glacier - as the paper says "the juxtaposed images are not only visually stunning but also of significant scientific value".

In particular this pair of images shows the transformation of the Imja glacier in the 1950's to the present day Imja lake.

Matt R

-- Matt R (email)

An interesting map project is 'Here and There' from the design consultancy Schulze and Webb, in which a conventional perspective view of the New York skyline merges seamlessly into a conventional map: http://schulzeandwebb.com/hat/

The ringworld-like effect of the distorted perspective means the horizon disappears. Jack Schulze discusses the influences on the project at http://schulzeandwebb.com/blog/2009/05/04/here-there-influences/, which include the Hockney discussion of the Chinese painting mentioned above - there is a link to a youtube clip of the TV programme.

-- Jakob Whitfield (email)

Blind Photographers

Dear ET,

I frankly do not know where to include this posting - but I thought you would find it interesting. It is an exhibition called "Sight Unseen: International Photography by Blind Artists," that was held at UC Riverside's California Museum of Photography. The exhibition assembled 87 works by 11 artists and one collective. The website was down when I posted this but the LA times has a collection of images from the exhibition here (http://tinyurl.com/noc5zo).

One of the artists is Michael Richard, who died in 2006, his obituary is here (http://tinyurl.com/y9noaqq).

The image below is a photo by Michael Richard called "Double Take".

Best wishes


-- Matt R (email)

Here is a paper (http://vis.berkeley.edu/papers/horizon/) on a form of time series plot called a horizon.

Sizing the Horizon: The Effects of Chart Size and Layering on the Graphical Perception of Time Series Visualizations

Jeffrey Heer, Nicholas Kong, Maneesh Agrawala


We investigate techniques for visualizing time series data and evaluate their effect in value comparison tasks. We compare line charts with horizon graphs -- a space-efficient time series visualization technique -- across a range of chart sizes, measuring the speed and accuracy of subjects' estimates of value differences between charts. We identify transition points at which reducing the chart height results in significantly differing drops in estimation accuracy across the compared chart types, and we find optimal positions in the speed-accuracy tradeoff curve at which viewers performed quickly without attendant drops in accuracy. Based on these results, we propose approaches for increasing data density that optimize graphical perception.

ACM Human Factors in Computing Systems (CHI), 2009. pp. 1303 - 1312.

Awarded a Best Paper Award.

Below are a few images from the paper.

An example of this technique in ProtoVis is also available (http://vis.stanford.edu/protovis/ex/horizon.html)

-- Matt R (email)

Threads relevant to nature studies:
Theoretical speculations on leaving the flatland of paper and computer screens and working now in real-land and space-land.