All 4 books by Edward Tufte now in
paperback editions, $100 for all 4
Visual Display of Quantitative Information
Beautiful EvidencePaper/printing = original clothbound books.
Only available through ET's Graphics Press:
catalog + shopping cart
All 4 clothbound books, autographed by the author $180
catalog + shopping cart
Edward Tufte e-books
Immediate download to any computer:
Visual and Statistical Thinking $5
The Cognitive Style of Powerpoint $5
Seeing Around + Feynman Diagrams $5
Data Analysis for Politics and Policy $9catalog + shopping cart
Edward Tufte one-day course,
Presenting Data and Information
Chicago, September 23
Chicago IL, September 24
Minneapolis MN, September 26
Brooklyn, October 31
Brooklyn, November 1, 4, 5
Is it better when making a presentation with a Proxima(tm) type projector to use light/white text with a dark background or dark text with a light/white background?
-- Bob Nash (email)
The usual metaphor for screens (projection and computer) these days seems to be black type on a white background, that is, a paper metaphor. This sometimes results in video glare, with lots of rays coming from the background. Sometimes the old fashioned computer screen seems less tiring, showing lit-up text on a dead backround.
So you should try to reduce video glare, perhaps by reducing the figure/ground contrast. For example, our screens on this website usually have a light tint on the ground to reduce the electric blue-white video glare. Of course on the computer screen, one can just turn down the brightness control when working at night or in a darkened room. Television control rooms and airplane cockpits generally have darkened backgrounds upon which to display lighted data.
On paper, white type on a black background can result in filled-in serifs because of the printing process. It is also harder to read white type on a black ground (there surely must be some evidence about this). But we should never be working at the edge of legibility in any situation, so legibility tests might not be relevant.
For projection presentations, obviously the figure/ground contrast must be sufficient to overcome the ambient light. Check your presentation room out in advance, test your projectors, find out how to control the room lighting and the window curtains if necessary. You should simply look at the various design solutions under real conditions to see what is going on, and not depend on verbal discussions such as this to decide!
-- Edward Tufte
I frequently use white text on a black background, because it removes the "frame" around the image or text being shown. The frame conveys no information.
Further, the frame limits where things can be placed on the slide. Because we are used to margins in printed text, we expect some space between images / text and the projected frame (the paper metaphor E.T. referred to). By using a black background to remove the frame, an image can be made larger and thus more visible. This is particularly helpful if the image does not have the same proportions as standard slides (i.e., 2:3 for film slides; 3:4 for computer screens, PowerPoint).
-- Zen Faulkes (email)
-- Edward Tufte
Other advantages to projecting white letters on a black background rather than vice versa:
-- Zen Faulkes (email)
Another good reason to have dark backgrounds and light lettering on PowerPoint slides is because of an optical illusion. Try this. Compare two shapes, say circles, of the same size, one white with a black background and one black against white. White against black always appears larger than black against white.
Hence, across America we see interstate highway signs with a dark green field and white lettering that "glows" when illunimated at night. In Beijing, China, where I teach, the street signs are dark blue with white lettering.
-- Michael Buschmohle (email)
I recall an undergrad Psychology course from about 1971 in which the text authors claimed research showed black text on yellow paper was best for readability (comprehension?). The text used exactly that scheme: black text on yellow paper. Anyone else aware of this research? I have no idea who the authors were.
-- Jim (email)
-- Keith Babuszczak (email)
Are there any different suggestions for longer term exposure to projected presentations? For example, every day for three weeks if you are on a training course. Should we change the colour every day to keep them interested? Should we change it for the period just after lunch when they become lethargic? Should it change to reflect the content or how they learn?
I am involved with the production of training courseware for military end users. As these courses are entirely computer based, the students may spend all day in front of the projector. The environment is controlled (lighting and such like) and standardised. We have followed the paper paradigm with black text on a white background for a number of reasons.
Primarily, because it is visually uncluttered (title, graphic, text). You have total continuity between screen and paper (we also have to produce the manuals to go alongside the presentations) which this also keeps printing costs down. It's quite a spartan approach (no watermarks, no background).
-- Adam Poole (email)
Can I make a plea for people to remember that not everybody has perfect eyesight. In my work I get involved in public talks to groups of all ages (literally from babes to 100 year olds) the fancy background so loved by many of the presenters at these events are not suitable for older members of the audience. And that's before one considers issues such as colour blindness.
In addition to the links given earlier for colour studies I use Vischeck to show me what pages will look like to those with colour blindness problems. There is a option there to test your own web page(s). Worth the effort. Not everybody has perfect eyesight like you.
-- Trevor Jenkins (email)
I think that inclusive design has started to become more mainstream thinking recently, in many areas of design. The huge (in the UK) DIY company B+Q have a range of powertools designed to inlcude those with limited hand mobility (i.e arthritis). These are well designed regardless of the inclusivity brief and stand apart aesthetically and functionally (i.e. an electric screwdriver that you squeeze to operate). Inclusivity can lead to new ways of thinking and solving problems.
The RNIB (Royal National Institute for the Blind) launched guidelines for inclusive design for print. I am not sure how widely this has been used but it would be interesting to find out how this could lead current layout styles (small text vs big text).
Being partially colour blind myself, I tend to stick to black, white, red, green, blue, cyan, magenta and yellow. I also try and memorise colour codes of successful colours as looking at a colour gamut chart in photoshop is a no-go!
-- Adam Poole (email)
The recommendation for limiting eye strain on VDUs is to make the screen background a similar shade to the surroundings. For a projected image, then, a black background is preferable because it is exactly the same shade as the surroundings.
I use Trebuchet font, coloured 'ivory' (not white) on a plain black background. White text is good for subtle highlighting. The text goes right to the edge of the slide, but since the background is black, who misses the margin?
-- Ewen McLaughlin (email)
Everyone here seems to prefer light type on a dark background. There are reasons to disagree. Most typefaces including screen fonts are designed for black-on-white. Second, many talks are given under conditions where the presenter will have no control over lighting. White backgrounds ensure that there will be at least some light in the room. Third, in my own experience at scientific meetings, where one sees a weeklong succession of 15-20 minute talks, each prepared by a different person, the light type-on dark slides are *not* easier to read, and screen glare is a hypothetical, not actual problem.
One presenter points out that the background for a VDT should be close to the luminance of the VDT. At least for scientific talks, the object of interest very often ends up being a black-on-white data graphic embedded on the slide. A black slide background often contrasts brutally with the embedded data panels, and makes such figures much harder to resolve. A great deal depends on the specific slide layout and typography, but the hypothesized advantage suggested by many here does not, in my experience, exist.
Finally, the darker the room is, the easier it is to sleep, and the harder it is to make notes.
-- Alex Merz (email)
Regarding the matter of sleepiness:
From Tortora & Gabowski's Principles of Anatomy and Physiology, eighth edition:
One horomone secreted by the pineal gland is melatonin, a biogenic amine. Less melatonin is liberated in strong sunlight, according to the following sequence (Fig 18.25):(1) Light enters the eyes, strikes the retina, and stimulates photoreceptors.
(2) Retinal neurons activated by photreceptors transmit impulses to the suprachiasmatic nucleus of the hypothalamus.
(3) From here nerve impulses are transmitted to the superior cervical ganglion.
(4) Sympathetic postganglionic fibers from the superior cervical ganglion extend to the pineal gland and form synaptic contacts with cells of the pineal gland.
(5) In sunlight, norepinephrine released by the sympathetic fibers inhibits secretion of melatonin by cells of the pineal gland. In darkness, retinal neurons transmit fewer impulses to the pineal gland by way fo the suprachismatic nucleus and superior cervical ganglion.
(6) In sunlight, inhibition of matonin secretion results in lack of sleepiness. In darkness, lack of norepinephrine stimulates secretion of melatonin by cells of the pineal gland and the result is sleepiness. Thus the release of melatonin is governed by the diurnal (daily) dark-light cycle.
Here's a good article with pictures.
Light = wakefulness is a double-edged sword. I meet many struggling college students in my job and the most common problem is sleep. I suspect spending too much time in front of a bright VDT late at night has something to do with this.
Use white backgrounds and keep the lights up in presentations. Send your kids to school with laptops, not desktops: at least they can close the screens.
-- Niels Olson (email)
Niels's post raises another reason why white backgrounds are preferable in presentations. Several optical aberrations are more severe in large-aperture systems compared to small- aperture systems of the same focal length. This is why most camera lenses have less contrast and resolution when used at full aperture (e.g., f/2) than when the aperture is closed (e.g., f/8). The same is true of the eye, which is why it is often possible for a person with myopia or astigmatism to read small print by squinting.
In dimmer light, the eye's aperture - the pupil - opens, and optical defects increase in severity. Hence as our vision deteriorates with age, we need brighter light to read by. The lesson for presentations seems obvious: dark backgrounds will result in dilated pupils and poorer vision. This effect should be worse for people with worse eyesight. At least that's the prediction. Quantitative tests of this notion would be straightforward and useful...
-- Alex Merz (email)
I give many presentations, lectures and trainings in a variety of settings, often to audiences diverse in age, ability, language comprehension, etc. I agree that the best choice is what works best in each individual location. However, it is a luxury to have that much time, information and control over the situation.
Using dark text on light backgrounds has almost always been the best choice for me. This is especially true when photocopies of the presentation are required and when other visual props and devices are also used.
-- Ana-Marie Jones (email)
The aberrations of the wide-pupil eye are the reason I switched from using white on black screens to black on white, even back in the days of DOS text screens. In addition, the white background "blooms" into the black letters, giving an impression of finer type and on graphs, finer lines. White on black letters seem bloated, by contrast.
However, nowadays I find that a pure white background is a bit harsh. I have set my Windows pallettes to emulate this website -- Tufte's ivory white.
-- Phil Lawless (email)
A retired heart surgeon with many years of medical experience approached me with some wisdom about the color of projected backgrounds. He said that the light receptors (rod and cone cells) in the eyes of young and old alike are strained and become tired when forced to look at a brilliant white projected image in a very dark room environment. His advice was to do the following:
1. In venues where the room environment is dark, a stark white projected background creates a stronger more straining and single point of focus for our eyes. Therefore less strain on our eyes would be caused if the background were darker and the text lighter. I found this to be true.
2. On a computer screen in an environment where the general lighting conditions around your monitor are illuminated, the background color best serves the viewer with less strain on the eye by being a lighter color or value with the text being darker.
My observations over the last twenty years of designing and producing presentation visuals for projection, print, and computer are centered in three principles:
1. Design for the audience and the environment.
2. Make it as easy as possible for the audience to get the message.
3. Use a pallete of comfortable color contrasts that work to decrease eye strain.
-- Larry Nielsen (email)
For a long time I used a text editing program on the mac (intended mainly for programmers I think) that had a button to toggle between "day mode" and "night mode". You could customize the colors for each, but the idea was that "day mode" was dark-on-light and "night mode" was light-on-dark. I didn't always use this feature but it was nice to be able to quickly switch if I didn't like one or the other.
-- Reed Hedges (email)
As a frequent presenter I've resolved this question for myself as, "It depends," finding that dark text / light background serves best in well-lit rooms, and light text / dark background serves best in darker rooms. For particularly important presentations when I do not have control of the lighting or can't predict what I'll find when I arrive, I will prepare two versions of a presentation, one light / dark and one dark / light, and use whichever best serves the context.
I also use a non-standard screen size -- a golden rectangle -- for my presentations. I do so to make certain projected images and slide layouts more visually compelling, but it only works with light backgrounds or with full-bleed images when the frame is obvious. Of course, with full-bleed images the question of background color ceases to be a question. With text rich slides (which I do not as a rule use) I would heed the counsel of other posters and go dark background to do away with the margin.
Finally, I might respectfully suggest that if we are too concerned with the readability of our text on-screen, or that eye strain may become an issue, this is a signal that the information is better served by a printed handout than an on-screen display. If there's one thing I took from ET's Cognitive Style of PowerPoint essay, it was (my paraphrase): "Pictures on screens, words on paper, unless the words serve as pictures."
-- Alan Nelson (email)
Much of my work involves PowerPoint, and I've done a lot of work for companies presenting to government agencies. Usually, these meetings are held in hotel ballrooms, where the ambient light desaturates even the most colorful slides. After being asked to "brighten the slides", my attempts to satisfy the client by using the most saturated dark blue (R0, G0, B255) and by attempting to "brighten" the other colors still failed. Attempts to generate constrast by means of color were also unsatisfactory.
I attempted to solve this problem by using a strip of dark blue for the title area (R0, G0, B204) and by using plain white for the content area, but plain white introduces other limitations. For example, the apparent contrast of cyan (R0, G255, B255) on the dark blue mentioned above is quite different on plain white, when projected.
Consideration must be given to the relative luminance of the color choices. A green line (R0, G255, B0) and a red line (R255, G0, B0), when laid on a white field and projected in a venue with high ambient light, will appear surprisingly similar. When laid on a dark background, the differences are more apparent.
Digital projectors technically can project 16 million colors, however, in practical terms, the useful color range is much more limited. In conditions with high ambient light, background colors that use RGB values below a certain level are indistinguishable from black, which, in my experience, was roughly equivalent to a 60% black in print. Various "green" colors may project as indistinguishable from yellow; colors with any amount of red in the RGB mix tend to project much darker than they appear on the computer screen. We usually test projectors with a test slide that contains various widths of rules and all the colors we're using in the presentation.
-- jdominic (email)
As a compromise between avoiding eye strain due to light glare caused by pure white on one hand and, on the other hand, the need for showing data plots and to make slide printouts (both of which look terrible on dark background), I second Tufte's use of an ivory background or also Michael's use of gray shades, with a little twist: try different shades of white, e.g., Cosmic latte, Cream, Ivory, Magnolia, Old lace, Seashell, Beige, Lemon Chiffon (see Wikipedia). Make sure you test these with a projector, not just your computer screen!
-- Felipe G. Nievinski (email)
An interesting discussion!
I'd like to introduce another parameter: the light output of the image source. We have a facility that is in effect a very large, very bright rear-projection system with a really well-anchored black field in its dark state. The brightness of the system was set assuming that the ambient light in the room would be on, allowing the speaker to be clearly seen (based on the bias that the presenter should be the main focus of a presentation, generally) and allowing the other participants to see their notes, other participants, etc. Might the custom of turning off lights for projection be an accommodation for older generation projectors, from magic lanterns to Ektagraphic Carousels to recent small portable projectors?
Regardless of how we got here, people presenting white background slides in this facility with the lights subdued find that their audience is often vocally uncomfortable with how bright the slides were. Turning the ambient lights back on relieves this discomfort. Dark backgrounds with contrasting bright text are easier to look at as well.
What effect does the emergence of large, bright, self-lit displays (such as large flat-screen) have on the answer to the question posed?
-- Bob Smith (email)