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
Portland, August 4
Seattle, August 5, 6, 7
Denver, August 11
San Jose, December 15
San Francisco, December 18, 19
What color is your salmon, flamingo, leaf, soil, golden retriever, yolk, beer, diesel fuel? Measuring color in the field

In a curious sort of information design, salmon farmers apparently choose the desired orangness-pinkness of their product from a color chart (SalmoFan) provided by Hoffmann-LaRoche which supplies the food dyes to color farmed salmon that would otherwise be gray, khaki, pale yellow, or pale pink. The SalmoFan is in tomorrow's New York Times, Farmed Salmon Looking Less Rosy.

[Update: the article is available here.]

Note the cute fins on each SalmoFan color chip.

This story in the Times takes on immediate personal interest because tonight we made fresh salmon homebrew sushi for dinner. A leftover piece, held up to the screen, falls about 6 color chips in from the right on the SalmoFan color wheel (assuming accurate monitor color, currently calibrated in the faint WYSIWYG hope of corresponding to CMYK printing on paper).

Or maybe it was wild salmon tonight. Which poses its own environmental concerns.

-- Edward Tufte

I'll trade all 3 of my books (autographed if you like) for a SalmoFan, which looks like a future classic of design.

-- ET

Go to this site and scroll halfway down...


I was amazed to learn that, "In one study, consumers from the East and West coasts of the United States were shown salmon fillets matching the hues on the Hoffman-La Roche SalmoFan??. Colour 33 was preferred by a two-to-one (2:1) margin. This market research suggests that consumers believe darker salmon colours are indicators of higher quality and better taste."

Doesn't look too appetizing on the screen. Two paragraphs later there is an 800 number where you can order a SalmoFan.

I already have all three books, autographed. How about a copy of "The PowerPoint Cognitive Style"? Or how about, since I live in Cheshire too, some leftovers? (Provided, of course, they are close enough to color 33.)

-- Danny Arsenault (email)

Hoffmann-LaRoche Canada has promised to send me out a Salmon colour wheel today! The moral is this: Avoid the heavy traffic of US business, deal with Canadians. Of course, I don't have it until it arrives. I expect to know more early next week.

Yes, I want autographed books! Melissa Spore

-- Melissa Spore (email)

O Melissa!

-- Edward Tufte

Upon receipt, ET, could you post some picture/presentation of this marvelous contraption?

-- Kent Karnofski (email)

SalmoFan photographic studies will be conducted, subject to our usual experimental controls, including varying lighting conditions and reference/calibration pieces of fresh salmon, farmed and wild.

-- Edward Tufte

Melissa Spore's SalmoFan™ package arrived this morning, including the amazing SalmoFan™ Lineal.

Below are various scientific measurement tools to assist in the assessment of SalmoFan™ Lineal.

Careful research protocols are now being prepared for submission to the Salmon Subjects Review Committee at the Graphics Press World Headquarters Research Laboratories. Upon obtaining the necessary approvals, we will proceed with control-color studies and field testing of the SalmoFan™ and, of course, the SalmoFan™ Lineal.

-- Edward Tufte

Above a comparison of Pantone colors, the SalmoFan, and homemade gravlax (from Atlantic salmon)

-- Edward Tufte

The SalmoFan Lineal in action at our test laboratory:

Note that several relevant variables in this image are out of control. The angle of light falling on the SalmoFan Lineal differs substantially from the angle of light falling on the Atlantic salmon. The measurement tool and the specimen should, of course, be on the same plane or very close to the same plane. In addition, the film is not adjusted for the warmth of the halogen lighting in the test laboratory and so, for example, the white cutting board appears distinctly orange. (And why assume that the film miscalibration has a uniform effect on the array of relevant colors in the experiment?) Despite these flaws in experimental design, the resulting sashimi was excellent.

-- Edward Tufte

Tonight our World Headquarters Test Laboratory was visited by Dr. Jorn Scharlemann of the Conservation Biology Group, University of Cambridge. Promptly demonstrating the virtues of international scientific collaboration, he pointed out the following, as mentioned at the San Francisco Zoo website: "In the wild, flamingos eat algae, crustaceans, brine shrimp, diatoms and aquatic plants. At the Zoo, a special 'flamingo fare' is served, a commercially prepared diet. To preserve their rosy color at the Zoo, they are fed a diet high in caratenoids. Initially, zoos fed carrots, red peppers and dried shrimp, but today it has been found that if synthetic canthaxanthin is added to this, nesting and breeding is more successful."

The scientific implications are astonishing; this cross-order, cross-species analysis finds enormous communality between salmon farmers and zoo-keepers--busy with doses of canthaxanthin (or similar) for fish and flamingos!

-- Edward Tufte

A new use for salmon coloring: to test if wild salmon is actually wild. As Marian Burros devastatingly reports in the New York Times, 6 of 8 samples of salmon recently sold in New York at up to $28.99 per pound as "wild" were in fact farmed salmon, as revealed by testing for additive carotenoids. This is why so much "wild salmon" was available off- season, from November to March.

How about blind taste-testing that compares farmed and wild salmon to see if renowned chefs can detect any difference? Blind wine-tasting indicates that sometimes it is hard to tell the difference between red and white wine (as reported in witty essay by Calvin Trillin).

This news story uses a good diversity of evidence in doing detective work.


-- Edward Tufte

Calvin Trillin's essay. Apparently wine-tasting is still in case studies. At this rate double-blind placebo-controlled trials are a ways off.

Two of my fellow students in a pharmacology class do investigative work for the FDA's fisheries section. Apparently the salmon dyeing is nothing compared to concerns over shrimp because foreign farms use some pretty nasty pesticides to keep the shrimp lice at bay.

-- Niels Olson (email)

Hi! Someone sent me your site after reading a post on my blog about farmed salmon. This is a great thread. The flamingo story is fascinating. And I love all your pictures!

The post on my blog was about farmed salmon, the SalmoFan, and how salmon farmers are using it to ensure a rosy color in their harvest - info I became aware of through the Smith and Lowney site that you also referenced. Curiously, I received an unsolicited email from an employee of the US National Oceanic and Atmospheric Administration (NOAA) saying that "The Salmofan can only be used to grade flesh color after a fish has been harvested. The Salmofan contains no instructions on how to achieve a desired color prior to harvest." (His letter is here: http://fanaticcook.blogspot.com/2005/03/equal-time-in-salmon-aisle.html)

Now, I don't know what to believe!

By the way, I also saw that story on wild salmon in the NYTs yesterday. I agree with you. It's devastating.

-- fanatic (email)

Just wanted to say thank you for a very interesting thread. One might even say its been salmon-chanted evening.

-- bridget hannigan (email)

The Salmon color wheel reminds me of a reference tool I used while working as an archeologist: The Munsell Color Chart. The Munsell chart is used to identify the color of soil (and the backs of sunburned archeologists). The Munsell book features color swatches -- like the Salmon wheel -- but also a small hole next to each swatch. The archeologist scoops up a trowel full of soil, and then holds it under the page of swatches. In this way the soil can be seen through the holes and immediately adjacent to the swatches (without staining the swatches themselves).

Here is a sample page from the Munsell book: http://store1.yimg.com/I/ascscientific_1881_1354636

-- Dug North (email)

Here's a picture from a dog breeder in England that, according to the web site, was "used in The Breed Standard Illustrated, to demonstrate the wide range of acceptable colour in the Golden Retriever".

Here's another picture from the same site, just because I like retrievers. Maybe it would fit better in the Dog camouflage: Whose hermeneutic shall prevail? thread but most of these dogs look all captioned out.

-- Simon Shutter (email)

Here's the soil color measuring device, with instructions for its use below.

-- Edward Tufte

A recent TV advertisement in the UK has stirred up the salmon farming industry. The advertisement shows a man in a lab describing how farmed salmon is coloured with the chemical Astaxanthin. The speaker then holds up on of the now famous "Salmon Colour Charts"

SCOTTISH SALMON PRODUCERS' PRESS RELEASE http://www.scottishsalmon.co.uk/mediacentre/releases/2006/280406.asp


"The Scottish salmon farming industry is very disappointed that Birds Eye has chosen to take this rather melodramatic and cynical approach..."


?? Salmon cannot synthesize carotenoid pigments and so must find them in their diet, to assist the reproductive process and protect the eggs. Colour charts are used to ensure that the salmon are fed sufficient quantities of carotenoid pigments to meet their and the consumers' needs.

-- Tchad (email)

There is another color reference from Roche, the Yolk Colour Fan: http://www.dsm.com/en_US/downloads/dnp/51559_poultry.pdf

-- Alessandro Gentilini (email)

Aha, what color is your yolk?

This topic on field color charts is a gift that keeps on giving.

-- Edward Tufte

a leaf color chart


-- Gregg Drube (email)

What Color is Your Urine?

I recently saw an episode of Channel 4 - The Worst Jobs in History about medieval jobs, including the Barber-Surgeon . On the show they showed the Barber-Surgeon using a urine color chart for diagnosis. Though not identical to the one showed on the show, I found 2 images of these charts:

-- Adi Shavit (email)

Hello Tufte crew! I hope this is not an incivility. This thread is the reason the Internet was invented.

-- Beth Parker (email)

50 people experience the seasons

This is a compilation of 200 photographs, 50 each tagged with "winter" "spring" "summer" and "fall" (left to right, top to bottom) in flickr, selected and combined algorithmically. One of 12 images in brevity's 50 people see ... set. I love the impressionistic shading and the subtle diffences in hue between seasons.

-- Rob Simmon (email)

I just finished working in the Middle East and they actually had urine charts at all of the urinals (with some people adding helpful blue and black colours meaning "You are almost dead", "You are dead"). It actually worked to enourage us to drink more and we joked that "I am at #3 today" and so on.


-- Bill Paton (email)

Many things are mixtures of similar colors, not a color

All these new contributions are superb, greatly extending the range of field color research.

Here's another problem in field color, the problem of mixtures of colors. An architect friend of mine had some clients who wanted their kitchen "toast color." Well, toast is many different colors. Similarly, I once wanted our house painted the color of straw; upon realizing that straw is many different colors, I wondered if we could glue a thin layer of straw to the sides of the house.

-- Edward Tufte

E.T.'s point about the real thing not having "a colour" reminds me of a feature of some of Professor Jack Cohen's biology lectures, where he asks the audience to call out the quality that owls have. After a few cries of "wisdom!" he advances the slide carousel to a picture of three of the goofiest-looking baby owls you ever saw. He then points out the obvious fact that no real owls of any sort actually possess wisdom, that's just a bit of human cultural shorthand (and that of one particular human culture: the Inuit associations of foxes and owls, for example, are with different qualities from ours. The fox is brave, not sly)

When people say "toast", "straw", "salmon-coloured" and "river", they usually expect their listeners to infer the stereotypical colour, and they're usually not so wrong to expect that. Growing up by the Thames, I expect real rivers to be brown, but I am much less likely to fail to interpret a blue meandering line on paper correctly, as the representation of a river, than I am a brown meandering line.

(The Harry Beck Underground map bears me out - blue Thames)

-- Derek Cotter (email)

Here are maple syrup colors from the Vermont Maple Sugar Makers website.

-- Mark Kasinskas (email)

Beer color scales, with fine color gradients, are available. There are at least 2 scales: Lovibond units and the Standard Reference Method of the American Society of Brewing Chemists.

And, consider the Lovibond Beer Color Database, which records the CMY color (cyan-magenta-yellow) tintometer readings for about 150 beers, ranging from the darkest, Wild Goose Oatmeal Stout (2.0%, 24.0%, 73.0%) to the lightest, Corona (0%, 0.4%, 0.9%).

It would be a lot of work, but a noble Kindly Contribution, if the colors in the list could be constructed into color chips tied to the name of each beer to make up a supertable of nouns, numbers, and colors. This is probably best done in Adobe Ilustrator; I'd be happy to tweak the table after the data and colors are laid out.

Perhaps also Kindly Contributors can search internationally for beer color scales and report back. For example, is there an image of the color scale used by the European Brewers' Convention?

Finally, can Kindly Contributor recover a working link to the NYTimes story that opened this thread?

-- Edward Tufte

Lovibond Beer Color Database

Only seven of the beers in the Lovibond Beer Color Database registered Cyan. The remaining 91 beers seem to fall into a pattern:

-- Dave Nash (email)

That's what statisticians call "multicollinearity". Time for some dimensionality reduction? Excellent analysis by Kindly Contributor Dave Nash, who may want to contribute to the Proceedings of the National Academy of Brewers.

-- Edward Tufte

Response to What color is your salmon, flamingo, leaf, soil, golden retriever, egg yolk, beer, diesel fuel? Measuring color in the field

Aaron Swartz's link generator only goes back to June 2003, so something else will have to be done to recover a functional link.

Here's a Pure Vermont Maple Syrup Grading Sampler I got in Texas. It took some experimenting to capture a reasonably reproducible shot. The Navy uses a very similar procedure for daily lubricating oil analysis. The Pure Vermont Maple Syrup Grading Sampler, regretfully, did not come with a fifth empty bottle in which to pour your own sample for analysis.
On a child's chair in front of the dishwasher On a lightbox On the lightbox with a flash
On a child's chair in front of the dishwasher On a lightbox On the lightbox with a flash
On the lightbox, outside in the shade Seconds later: a weird redness from the lightbox
On the lightbox, outside in the shade Seconds later: a weird redness* from the lightbox
In front of the lightbox, outside. This closely approximates the Navy's clear-and-bright lube oil test. Seconds earlier-the weird redness  demoting Vermont Fancy to Grade A Medium Amber, and even to Grade A Dark Amber when compared to the next picture up.
In front of the lightbox, outside. This closely approximates the Navy's clear-and-bright lube oil test. Seconds earlier: the weird redness* demoting Vermont Fancy to Grade A Medium Amber, and even to Grade A Dark Amber when compared to the next picture up.

*I suspect the redness in the shots is due to fluorescent flicker.

In trying to find a picture of the Navy's clear-and-bright test for oil, I ran across this picture grading copper corrosion, courtesy of Alcontrol Laboratories:

-- Niels Olson (email)

The maple syrup scale photographs are excellent and provide a new form of analysis: color scaling of color scales. A further meta-analysis would involve comparing Niels Olson's various color renderings of the maple syrup scale on different computer monitors with different gammas.

Reproduction of field-color scales introduces color variation into the desired color scale. Perhaps a further topic for this thread then is on synchronizing the reproduction of field color. And then there's the field light which may interact differently with the field object and the field color scale.

-- Edward Tufte

Response to What color is your salmon, flamingo, leaf, soil, golden retriever, egg yolk, beer, diesel fuel? Measuring color in the field

-- Edward Tufte

Response to What color is your salmon, flamingo, leaf, soil, golden retriever, egg yolk, beer, diesel fuel? Measuring color in the field

As a further, belated contribution: MyCuppa mugs, 'to help you mix your favourite brew to just-how-you-like-it by matching the colo[u]r guide on the inside'.

-- Emanuele Pucciarelli (email)

Response to What color is your salmon, flamingo, leaf, soil, golden retriever, egg yolk, beer, diesel fuel? Measuring color in the field

Dear Readers,

I had been chuckling to myself about this thread for a while, but Niels Olson's picture of the ASTM copper corrosion standards finally did it for me! Many years ago I began my laboratory life as a technician in Shell Research in the hydraulic oil lab and had the task of using precisely this standard. The pieces of copper act as catalysts for oil oxidation and had to be carefully polished to a precise protocol prior to placing in oil. In actual fact in my experience they were a good field guide as they are not photographic reproductions but a collection of real copper strips.

Field measurement of colour is a notoriously difficult thing to do and even making comparisons of different "whites" is very difficult with either humans or image analysis equipment (ask the paper industry). There are effects due to culture (some countries prefer their "white" shirts to have much more blue in them than others!), human perception, lighting control, what colour the test swatch or strip is placed next too etc.

Great thread though.


-- Matt R (email)

Response to What color is your salmon, flamingo, leaf, soil, golden retriever, egg yolk, beer, diesel fuel? Measuring color in the field

I just found tomato colors today:

-- Andrea Cesari (email)

Response to What color is your salmon, flamingo, leaf, soil, golden retriever, egg yolk, beer, diesel fuel? Measuring color in the field

I don't believe this example has been posted yet ... http://infosthetics.com/archives/2007/08/mycuppa_pantone_color_coffee_cup.html - an interesting example of both color matching *and* instructions at the point of use.

-- Cathy Oliveri (email)

Response to What color is your salmon, flamingo, leaf, soil, golden retriever, egg yolk, beer, diesel fuel? Measuring color in the field

Just happened to be looking for blood smears of anemias, and came across this on Flickr. It is well known that anemia can be seen in the eye lids, but, at least in humans, the distinction is typically "pale" or "pink". Pale bad, pink good. I had no idea that someone had devised such as thing as a FAMACHA score.

-- Niels Olson (email)

What a fantastic thread!

The SalmoFan instantly made me think of the CoralWatch Coral Health Chart - a colour chart which divers take underwater to monitor coral bleaching. The colour swatches on the chart represent different stages of coral bleaching or recovery, calibrated from laboratory experiments.

What is coral bleaching?
Request a free chart from CoralWatch.

-- Rebecca Weeks (email)

As a former bio-optical oceanographer, I would be remiss if I did not contribute one of my favorite classic oceanographic tools to this excellent discussion. The Forel-Ule (FU) scale dates back to the 1890s and is used to gauge the transparency and productivity of marine and fresh waters. While digital sensors are more commonly used for the task of measuring ocean color and turbidity these days, the FU scale remains an excellent teaching tool; it demonstrates ocean color concepts in the field in a way that digital data can not. An excellent review of the FU scale is here.

A Forel-Ule scale contains clear vials of variously colored liquids that are held up to the water body for visual comparison. Here is a photo of one version of the scale:

(image credit: Heather Haberman, NOAA Teacher at Sea).

-- Amanda Whitmire (email)

Threads relevant to color: