Posted by Jack Williams on 03-29-2007 02:50 PM:

Seeing red - color, dyes, pigment, and fade


What a beautiful visual walk through the world of “red.”. And if I owned the collage of rugs that you use to illustrate the effect of “red”, I would probably spend a lot more time at home, staring at the walls, or floors, or ceilings.

To supplement your presentation, I will offer a few observations about “reds” used in woven carpets. My apologies if this is elementary. I'll add some additional information later.

“Red” of course is a “spectral color” that is defined scientifically as those wavelengths occurring between 625 and 720 nm (see - ). “Red is an additive primary color of light, complementary to cyan. It was once considered to be a subtractive primary color, and is still sometimes erroneously described as such in non-scientific literature; however, the colors cyan, magenta and yellow are now known to be closer to the true subtractive primary colors detected by the eye, and are used in modern color printing” (quoted from - ).

As John has pointed out, what is a spectral color is not necessarily what our eyes perceive as a color. Our eyes have three types of receptors and the perception of color is dependent on factors other than pure wavelength. In my opinion, the Wikipedia sites are easy to understand and have good organization for the basics of “color”. The sections covering “natural colors” and “pigments” are also very good and even more germane to our hobby in my opinion.

There are actually quite a few natural sources for “red” dye. But for the purpose of carpet coloring, probably only five are likely to be frequently encountered. These are (1) madder; (2) Mexican cochineal (an insect dye originating in Mexico and in use world wide after the 16th C Spanish conquest); (3) lac (a cochineal-like insect native to India); (4) kermes (a cochineal-like insect native to the Mediterranean coast); and (5) Polish cochineal – or “St. John’s blood” (a cochineal-like insect native to Poland, central Europe, and at one time distributed across all of arid Eurasia).

Madder is well known to all of us and of course has two primary color constituents, alizarin and purpurin (though many madder dyes have other minor color influencing impurities, and the resultant color is heavily dependent on dying methods and mordant). Neither of these constituents is inherently color fast but I’ll reserve my thoughts on that “madder” for another post.

Kermes, lac and Polish cochineal were well known dyes that were eclipsed in the 16th century by massive import of cochineal from Mexico (see: ). Kermes was commercially used especially in Europe, Egypt, Morocco, and Spain from earliest times (coloring of Egyptian mummies used Kermes) prior to the importation of cochineal which produced a much more vibrant color.

Lac is an Indian sub-continent sourced cochineal-like insect dye that was widely exported - forever - into central Asia (commercial records from Yarkand note import of indian dyes 1,500 years ago). It is present in older carpets from East Persia, Afghanistan, Turkmenistan, and the carpets of East Turkistan (see and also Bidder, H., Carpets from Eastern Turkestan, Zwemmer, London, 1964).

Polish cochineal is not well known to many in the carpet world, but it was a very important commercial dye especially in Europe prior to the introduction of cochineal from Mexico. In the 16th C. when cochineal became widely available Polish cochineal was virtually driven from the commercial markets, but continued to be used locally.

When Russia completed the takeover of Western Poland, Ukraine, etc. in the mid 18th C., Polish cochineal dye became a major export again, but this time mostly exported to Turkmenistan for use in Turkmen carpets. The insects responsible for Polish cochineal were available in Central Asia, and were familiar to the dyers in the area. But the Russian imported dyes apparently became an important source of the red dyes for Turkmen rugs until the introduction of artificial red dyes (see:, and paragraph immediately after “plate 15.”).

Possibly of interest to ruggies are the following undifferentiated notes:

(1) Madder was not used to dye silk very often, for reasons I have not yet discovered. Cochineals were the preferred dyes for silk. But all (?)cochineal dyes are reputed to tend to “run” (this is probably related to the method of producing cochineals in a powered form). It seems that some will even “run” 100 years later if over-dyed initially. Many references mention this including A. Cecil Edwards.

(2) Purple is not a spectural color, but a combination of red and blue (this impacts fastness in the presence of light). Violet however is a spectural color.

(3) It seems that none of the natural red dyes are very colorfast in the presence of light. About 300 hours exposure might be enough to fade many natural red dyes almost completely (see site below). Some references note that cochineal dyes are the most light-fast of the natural red dyes. Given the experimental data below, I have some doubts about this. In another post I will offer my thoughts on the color-fast nature of natural colors, pigments, and probably dyes (?), especially “red.”.

Of course the rugs we are interested in almost never have spectural colors, and rarely have colors that approximate the original colors used by the weavers. I suspect our hobby is a little short of scientific terminology and accurate measurement applied to colors, pigments, and dyes when compared to...say...the world of painting and water colors.

And in our art form/hobby, in my experience little is usually made of color change in rugs with time, unless an obviously nasty color or "fade" is present. Then, it seems to me that most comments are made from heresay rules-of-thumb, often merely someone's opinion, rather than from science.

For color change in pigments caused by light, I highly recommend the following web site and all of its sub-links. This site is the best I have found and is quite a resource for understanding colors, pigments, etc. The author seems to be something of a real character and he writes in an engaging way... (see - .)

Regards, Jack Williams

Posted by R. John Howe on 03-29-2007 05:54 PM:

Hi Jack -

Thanks for the additional indications and links on red.

It is true that there are a lot of natural sources for red. I just ticked off a list that Chenciner provides, that doesn't pretend to be exhaustive at all.

As I was talking with Susan McCauley, who presented a rug morning at the TM recently on contemporary Southeast Asian textiles, I learned not only that lac is still produced and used there (Sheridan Collins had previously reminded me that it is also used in Bhutan) but that there is another local source of a natural red used there.


The list of natural sources for red is likely more extensive than the listings we have managed suggest.

And it has also been noticed by some scientific folks in the rug world like Ned Long that we do not have accurate ways of referring to specific colors. He wrote once in ORR, arguing that we should adopt some existing standards based on light wave frequency (if that latter is the correct term) and give up our "Madison Avenue" descriptions of color.


R. John Howe

Posted by Steve Price on 03-29-2007 07:50 PM:

Hi John

You can express the spectral position of a color as either the frequency or the wavelength, since they are linearly related to the reciprocals of each other. In terms of wavelength, the visible spectrum extends from about 400 nanometers (blue to ultraviolet) to about 650 nanometers (red to infrared). The various tinges that "colors" have result from mixtures of wavelengths that they transmit or reflect.

The light sensitivity of dyes results from the fact that they absorb light at the wavelengths that they don't transmit or reflect. Since light is a form of energy, it can cause reactions in molecules that absorb it. The energy in a unit of light varies with wavelength, being greatest at short wavelengths and smallest at long wavelengths. Thus, if all other things were equal (which they never are) red dyes would be most light-labile because they absorb light of short wavelengths (hence, of high energy).


Steve Price

Posted by Lloyd Kannenberg on 03-30-2007 02:40 PM:

Hello John, Jack, Steve and all,

"Color" is a vexed issue because there is a mismatch between "spectral color", defined by its wavelength or frequency, and "physiological color", which is what we SEE. This is nowhere more problematic than among the reds. Some of the most striking reds , from dark "red delicious apple red" to those tending over toward reddish violet, aer not found in the spectrum at all (look carefully at the rainbow the next time you see it - the first red you see is actually a rather washed-out orangy red). Our eyes synthesize these wonderful hues from combinations of spectral colors, a phenomenon known, at least in part, to Isaac Newton. Because we "process" the color images in our brains, what we see depends, among other things, on the kind of illumination we have at hand (all ruggies are familiar with this!). International color standards claim that the best light source is daylight (overcast sky, not blue, and not direct sunlight), or in default of that a xenon lamp, not a filament or mercury - or, heaven help us, a fluorescent lamp! There are standard references on color (I use a Hickethier Color Atlas), but the truth is that such "scientific" measures are really of much less use in studying and enjoying rugs and other textiles than a sensitive eye and extensive experience in looking at the objects themselves. That's why John's presentation is so uniquely valuable!

Best to all,

Lloyd Kannenberg

Posted by Steve Price on 03-30-2007 09:33 PM:

Hi Lloyd

I think you're in danger of confounding the physical entities that cause us to perceive color (light at various wavelengths) with the perceptions and aesthetic judgements we make about the colors of various objects. They are related, of course, but I think it's important to maintain the distinction between them.

First, nearly every color that we encounter in the real world is a mixture of wavelengths, which is why you can't locate many of them on a spectrum.

Second, light sources consist of mixtures of wavelengths, and the wavelength mixture reflected from an object depends on the light source for that reason. Incandescent lamps have a heavy yellow component, fluorescent lamps are weighted toward the blue end. Daylight (and halogen lamps) are more nearly balanced across the spectrum.

Third, the receptors through which color interacts with our nervous systems (the cones) are of three types, each with a different spectral sensitivity. Their differing spectral sensitivities result from the fact that the molecules in them that absorb light differ.

Fourth, and probably the most complicated, is the aesthetic values that we put on colors, which depend on all of the first three plus the results of whatever we've learned and experienced as well as the situation of the moment.


Steve Price

Posted by Jack Williams on 03-31-2007 12:48 AM:

Steve, all you have posted is absolutely true. But so far as I have seen, our hobby has barely even begun to apply science to the descriptions of rugs. It is frustrating that we don't even have an agreed upon way to describe rug colors.

Fortunately, a great deal of work in the area of defining red colors and pigments has been done for 200 the world of artistic painting. A lot of what we are concerned about including a definition of colors, has already been confronted in that world because of significant color problems in the past. Whistler's paints for instance are apparently faded into some nasty synthetic colors because of his failure to anticipate color change.

Here is a link into the site-map of the last site I posted that shows some red colors, their fastness, scientific color designation, names, etc....

...and here is a link into another seciton devoted to natural pigments...

One reason I posted this link on water colors is that site has the most fantastic detail on color, pigments, the creation, testing, history, preparation, source color definition, light fastness, etc., of artist water color paints. The site has an almost inexhaustable set of data covering everything from the mechanics and science of color to production of pigment.

Because artistic paints developed from pigments that were almost always derived from substances that were intended for some other use, often dyes, a lot of what is already cataloged and defined in the painting world could probably be adopted almost directly to our hobby.

The world of artistic painting was probably the source for a lot of studyof colors and cataloging of pigments. and for the study of colors and color theory, beginning 200 years ago. For instance, here is a quote from the site:

"...Vision and color are at the heart of painting. Here is the most comprehensive discussion for artists of color perception, color psychology, "color theory" and color mixing available online, and one of the most comprehensive available anywhere in any format."... go to...

I believe that the treatment of pigments in this stufy applies almost directly to dyes. As yet, I just cannot figure out a way to reduce and post any meaningfull amount of the infomation. But for example, please go on that site and take a look at the "blue wool" test for light fastness of pigments, including madder, cochineal, etc., I promise, it will open a lot of paths for research.

Apparently because of a 200 year history of research and cataloguing of paint pigment colors, there is a ready made set of catagories that might accurately describe a specific color in our rugs...if we would adopt the system.

I am fascinated by what is linked here. I hope those with a scientific bent in our hobby will find the depth of scientific characterization adopted in painting as an inspiration and as a resource. Regards, Jack

Posted by R. John Howe on 03-31-2007 04:40 AM:

Dear folks -

Well, we can see now that there's a great deal more to "color theory" than initially "meets the eye," so to speak.

So what is the advice of the more technically fluent amongst us?

What aspects of "color theory" seem accessible and reliably useful to folks like ourselves, who already balancing a great many variables and qualities as we collect?

The sort of "snippet" treatment I resorted to seems such a gloss (and a distorted one at that) that I wonder, now, whether it was a useful thing to raise. Is this an instance in which a "little knowledge" is, in fact, a dysfunctionally distorting thing?


R. John Howe

Posted by Steve Price on 03-31-2007 06:56 AM:

Hi Jack

That is an interesting and meaty article.

The instability of dyes resulting from exposure to light is a fundamental photochemical phenomenon, the basic understanding of which doesn't require expertise, but does require a little knowledge about the nature of light and its interactions with chemicals. That's why I said a few words about that in my first reply.

It doesn't require any background in the perception or aesthetics of color, neither of which have any bearing on it. My impression from your opening post was that you intended to focus this thread on the light-induced fading of dyes, and I think it would be wise to try to keep that focus from being derailed by the other interesting, but very wide ranging issues that were introduced. I think trying to cover everything we care about that relates to color and its perception in one thread can only lead to confusion. The technology of dying and coloring are related to the main focus, being specific applications of the basic science. In my opinion, it's best to separate their discussion from the other topics because it's so specialized.


Steve Price

Posted by Patrick Weiler on 04-01-2007 05:47 PM:

Run, red, run


I hope this comment does not put the thread on a tangent, but a couple of things in Jack's post caught my eye:

"But all (?)cochineal dyes are reputed to tend to “run” (this is probably related to the method of producing cochineals in a powered form). It seems that some will even “run” 100 years later if over-dyed initially. Many references mention this including A. Cecil Edwards."


"When Russia completed the takeover of Western Poland, Ukraine, etc. in the mid 18th C., Polish cochineal dye became a major export again, but this time mostly exported to Turkmenistan for use in Turkmen carpets. The insects responsible for Polish cochineal were available in Central Asia, and were familiar to the dyers in the area. But the Russian imported dyes apparently became an important source of the red dyes for Turkmen rugs until the introduction of artificial red dyes"

Runny reds in Turkmen carpets supposedly from prior to the advent of synthetic dyes have been noted and my initial suspicion was that it was madder that had run, but these statements perhaps point to cochineal being the suspect color. Others felt that runny reds only indicated a later weaving.
We can all now go back to collecting runny red Turkmens again......

Patrick Weiler

Posted by Jack Williams on 04-01-2007 10:59 PM:

running dog imperial red

Patrick, actually, you have identified something important involoving dyes, pigments, etc. I too have experienced some runny colors, especially reds but also others. On one especially nice turkmen rug, the run was strictly limited to reds on silks.

This is what initially attracted my attention...and lo and behold, I found quite a few sources that discuss how different silk dyeing is than wool, and that mention dying silk only with cochineals. I 've looked for an opportunity to discuss the use of cochineals in Turkmen carpets...but i see that Filiberto also had the information on Polish Cochineal, and connected it back to Turkmen carpets using the same article that originally caught my eye.

From the oddities involved in dying silk, of perhaps more interest was the frequent references to cochineal dyes and "running." From that, I got interested in common color changes in fabrics caused by one of the five major reactions. This led to questioning the rules of thumb that were widely believed, such as....

Running colors are synthetic (nope, running has little or nothing to do with dye type)

Fast colors are natural (nope, you don't even know if the color is "fast" because you don't know the original color, how long it has been exposed to sunlight, etc...)

tip fading is an indicator of synthetic dyes (how do you know? Fading is a result of adding energy to the dyed material and the amount of fade is related to the dye, method of dying, color and type and amount of energy...and because tips receive energy first just about any dye should begin fading at the tips).

Bright orange or bright yellow is a synthetic (Orange by definition is "bright." It can be produced in many different ways naturally, many both natural and synthetics react differently through time. Yellows were usually natural local dyes and are often an especially naturally unstable dye. Some of the brightest yellows were natural. Orange produced by combo of yellow and red can change color dramatically with time as the yellow preferentially fades...same with green produced by blue and yellow combo) we don' know.

The more we look, the more we discover that we cannot positive tell natural or synthetic dyes by look. And scholorship indicates that red might be the most common synthetic dye used in Turkmen carpets.

Finding a turkmen carpet with some red run might actually indicated a natural dye (cochineal base). But of course madder can easily be made to run for years by faulty dying, faulty use of mordant, etc.

I think that it would do the hobby a great service if we could just adopt a standard color designation such as used in painting. And then perhaps a color check front and back could be compared to "blue wool" strip tests for fading giving some indication of dye source. But heck, maybe it is just me...this desire to be able to catalog the hobby might be evidence of compulsive engineering mentality.

In any case, I'll add some fade data next week. And maybe the source of the dye is not as important as we often conclude. A bad runny color is...a bad runny color.

Posted by Filiberto Boncompagni on 04-02-2007 02:02 AM:

Hi Jack,

Yes, you are tight, we should revise our more common opinions about dyes. Albeit there are dyes that are obviously synthetic, if a dye looks good, that doesn’t prove it’s natural.

Uh, and the notion of running cochineal red isn’t new on these threads. Wendel Swan mentioned it in a discussion of salon 53:
Elena (Tsareva) did say that she has been told that excessive cochineal in wool can bleed since cochineal does not bind in wool as firmly it does in other fibers. However, she has no direct evidence of that fact since most Turkmen wool is dyed with madder - an especially fast dye that does not run.


Posted by R. John Howe on 04-02-2007 04:03 AM:

Filiberto et al -

I don't think that Jurg Rageth has dealt so far with "bleeding" reds in older Turkmen pieces.

Most usually, folks working with natural dyes (on wool at least) claim that no runs occur with them. They say that if a dyeing with a natural dye on wool is somehow defective (say a mordant that fails to operate properly), the result will be that the natural dye will not color the wool at all.

The odd thing to me is that Rageth's testing of an admittedly small number of Turkmen pieces (he has included most of the usual formats) indicates that the cochineal encountered is invariably of the Mexican sort. One would expect that, at least sometimes, the cochineal in Turkmen weavings would be from more local sources.


R. John Howe