Inks under the colorimeter

[amper]

I'm not sure if this has come up before, because I'm not sure what terms to use to search for it, but I found this really cool site where somebody took Greg Clark's Ink Sampler book and ran the samples through a colorimeter, or spectrophotometer, if you prefer. So if you really want to know *exactly* what color ink you have, check it out.

 

http://www.rmimaging.com/projects/inks/inks.html

 

[Titivillus]

I'm not sure if this has come up before, because I'm not sure what terms to use to search for it, but I found this really cool site where somebody took Greg Clark's Ink Sampler book and ran the samples through a colorimeter, or spectrophotometer, if you prefer. So if you really want to know *exactly* what color ink you have, check it out.

 

Thanks for the link. Now we can see how close the blues are to Penman!

 

Kurt

 

[KCat]

This is exactly what I've been asking about for years now - why can't that information be used to develop a Penman Emerald duplicate?

 

Huh? Huh? Why? Tell Me!!!!

[Rapt]

The big problem I see is if there is metamerism involved in the colours.

 

Colorimeters etc don't necessarily provide "perfect" information. Particularly if the colour has differing responses under different light sources.

 

In other words you may have a match with incandescent light bulbs, but not with fluorescents or the sun, etc...

 

 

 

 

 

[KCat]

The big problem I see is if there is metamerism involved in the colours.

 

Colorimeters etc don't necessarily provide "perfect" information. Particularly if the colour has differing responses under different light sources.

 

In other words you may have a match with incandescent light bulbs, but not with fluorescents or the sun, etc...

 

Yeah - that makes sense. Tulipe Noir looks like *&*& under my LED bulb, but lovely in regular incandescent light.

[tfwall]

The big problem I see is if there is metamerism involved in the colours.

 

Colorimeters etc don't necessarily provide "perfect" information. Particularly if the colour has differing responses under different light sources.

 

Any modern UV-Vis spectrophotometer has a sensitivity of at least 1 nm and quite possibly up to 0.001 nm depending on the instrument. I've never heard of the particular instrument used in this study, so I can't comment on its suitability for this work. In general though, the spectrum is independent of the light source used. The light sources only limitation is the wavelengths it can produce (200-800 nm, &c) and hence the area of the EM spectrum you can study.

 

The difficulty in replicating a ink color without using the original dye is determining which wavelengths are important. Then finding a compound which absorbs at close enough wavelengths and blending it with other coloring agents without them all reacting together to make a new compound. Here is where you might have metamerism issues, since you aren't using the original dye there are very likely some other absorbance bands in the spectrum, making the ink look different under UV light v. fluorescent lights, &c.

 

In short, just because you know what something looks like, doesn't mean you can replicate it if you can't use the original components. It is kind of like trying to build a Mercedes from a picture, but you can only use parts made by BMW to generate your mock-up. It's much easier to use UV-Vis to determine what the dye was and then just use it. Not viable though if the dye is something which is not available/safe today (as Penman ink dyes may be).

Edited April 10, 2008 by Chemyst

[Lloyd]

Lamy Blue Black 55.47 5.52 223.78

Montblanc Blue-Black 39.61 15.92 272.82

I thought these were the same?

[Sholom]

The big problem I see is if there is metamerism involved in the colours.

 

Colorimeters etc don't necessarily provide "perfect" information. Particularly if the color has differing responses under different light sources.

 

Any modern UV-Vis spectrophotometer has a sensitivity of at least 1 nm and quite possibly up to 0.001 nm depending on the instrument. I've never heard of the particular instrument used in this study, so I can't comment on its suitability for this work. In general though, the spectrum is independent of the light source used. The light sources only limitation is the wavelengths it can produce (200-800 nm, &c) and hence the area of the EM spectrum you can study.

 

The difficulty in replicating a ink color without using the original dye is determining which wavelengths are important. Then finding a compound which absorbs at close enough wavelengths and blending it with other coloring agents without them all reacting together to make a new compound. Here is where you might have metamerism issues, since you aren't using the original dye there are very likely some other absorbance bands in the spectrum, making the ink look different under UV light v. fluorescent lights, &c.

 

In short, just because you know what something looks like, doesn't mean you can replicate it if you can't use the original components. It is kind of like trying to build a Mercedes from a picture, but you can only use parts made by BMW to generate your mock-up. It's much easier to use UV-Vis to determine what the dye was and then just use it. Not viable though if the dye is something which is not available/safe today (as Penman ink dyes may be).

 

 

Just to clarify what is going on here, a colorimeter is not the same thing as a spectrometer, and because of the way we see color, the light source does matter.

 

[caution: technical stuff ahead - feel free to skip if it sounds like gibberish]

For the technically minded, the colorimeter we use at my company, a Gretag Spectro-Eye, can function as a spectrometer but with only 10 nm resolution, which is more than is needed to relate visible light absorption to human color perception. Chemyst is correct that any decent UV-visible spectrometer made today is at least 10 times better than that, but this fact turns out to be irrelevant as far as human perception is concerned. The human eye-brain combo actually doesn't pay much attention to where the absorption maximum of a dye is unless it is a very "pure" color, i.e. just one absorption maximum, and a narrow one at that. For typical dyes that have multiple absorption maxima. I find that the shape and position of the long wavelength side of the absorption peak (if there is one dominant one) is a better predictor of what color the dye is.

 

A colorimeter is based on the CIE standard, mentioned in the initially referenced web page. The CIE is an international scientific organization that since at least as far back as the 1930s has been developing methods of specifying colors in terms of physically meaningful parameters. The specific version used at the website under discussion is the L*C*h(ab)* system (let's ignore what the stars mean for this conversation). L* is called luminance, and it is intended to represent a linear scale of perceived brightness independent of hue, sort of like converting a color picture to black and white and then evaluating areas of the picture and relating them back to the original colors of those areas; it is sort of like brightness, but with the effect of hue completely separated out. C* is a measure of how much color is there, i.e. how far is the color from black/gray/white; it is a nominally linear and precise measure of color saturation or intensity. h(ab)* is a measure of hue, essentially where on the artists' color wheel is the color; not surprisingly h(ab)* is given in degrees of angle starting at 0 being red.

 

[note: end of the worst of the technical stuff]

The reason the type of illumination matters is because with inks (or any color that is not itself a source of illumination) we are dealing with reflected light, and of the wavelengths (i.e. colors) of light that the dyes can reflect (and are hence seen), the dyes can only reflect those wavelengths that they receive from the light source to begin with. Categorization of light sources is a very complex topic. The ink colorimetry webpage mentions a D65 filter. That means the measurements of color were made using a standardized synthetic daylight (sun, not shade). The numbers will change somewhat with other illuminants, and illuminants that have discontinuous spectra (e.g. fluorescent lights) create a whole set of additional complexities. As mentioned, some dyes, because of the shapes of their absorption spectra, are more prone to change color under different light sources. This phenomenon is called metamerism, and many fountain pen inks have it, even if variables like pen and paper are kept constant.

 

If you are totally overwhelmed by all this, don't feel bad. Human color perception is a tremendously complex subject that I only understand some of the basics of myself. The textbook mentioned at http://www.rmimaging.com/projects/inks/inks.html may be of some help, and there are other books on the subject as well.

 

Chemyst is correct that a UV-visible spectrometer could conceivably be used to identify the actual dye (if there are only one or two in the mix) in an ink. But as noted, the dye industry is always changing in response to taste changes, technology changes, raw materials changes, and our increasing understanding of the toxicology and ecological impact of some of the classic dye manufacturing processes. Just because we might identify the dye in say a 1950s Waterman ink doesn't mean that dye is still on the market.

 

[tfwall]

Just to clarify what is going on here, a colorimeter is not the same thing as a spectrometer, and because of the way we see color, the light source does matter.

 

[caution: technical stuff ahead - feel free to skip if it sounds like gibberish]

For the technically minded, the colorimeter we use at my company, a Gretag Spectro-Eye, can function as a spectrometer but with only 10 nm resolution, which is more than is needed to relate visible light absorption to human color perception. Chemyst is correct that any decent UV-visible spectrometer made today is at least 10 times better than that, but this fact turns out to be irrelevant as far as human perception is concerned. The human eye-brain combo actually doesn't pay much attention to where the absorption maximum of a dye is unless it is a very "pure" color, i.e. just one absorption maximum, and a narrow one at that. For typical dyes that have multiple absorption maxima. I find that the shape and position of the long wavelength side of the absorption peak (if there is one dominant one) is a better predictor of what color the dye is.

 

A colorimeter is based on the CIE standard, mentioned in the initially referenced web page. The CIE is an international scientific organization that since at least as far back as the 1930s has been developing methods of specifying colors in terms of physically meaningful parameters. The specific version used at the website under discussion is the L*C*h(ab)* system (let's ignore what the stars mean for this conversation). L* is called luminance, and it is intended to represent a linear scale of perceived brightness independent of hue, sort of like converting a color picture to black and white and then evaluating areas of the picture and relating them back to the original colors of those areas; it is sort of like brightness, but with the effect of hue completely separated out. C* is a measure of how much color is there, i.e. how far is the color from black/gray/white; it is a nominally linear and precise measure of color saturation or intensity. h(ab)* is a measure of hue, essentially where on the artists' color wheel is the color; not surprisingly h(ab)* is given in degrees of angle starting at 0 being red.

 

[note: end of the worst of the technical stuff]

The reason the type of illumination matters is because with inks (or any color that is not itself a source of illumination) we are dealing with reflected light, and of the wavelengths (i.e. colors) of light that the dyes can reflect (and are hence seen), the dyes can only reflect those wavelengths that they receive from the light source to begin with. Categorization of light sources is a very complex topic. The ink colorimetry webpage mentions a D65 filter. That means the measurements of color were made using a standardized synthetic daylight (sun, not shade). The numbers will change somewhat with other illuminants, and illuminants that have discontinuous spectra (e.g. fluorescent lights) create a whole set of additional complexities. As mentioned, some dyes, because of the shapes of their absorption spectra, are more prone to change color under different light sources. This phenomenon is called metamerism, and many fountain pen inks have it, even if variables like pen and paper are kept constant.

 

If you are totally overwhelmed by all this, don't feel bad. Human color perception is a tremendously complex subject that I only understand some of the basics of myself. The textbook mentioned at http://www.rmimaging.com/projects/inks/inks.html may be of some help, and there are other books on the subject as well.

 

Chemyst is correct that a UV-visible spectrometer could conceivably be used to identify the actual dye (if there are only one or two in the mix) in an ink. But as noted, the dye industry is always changing in response to taste changes, technology changes, raw materials changes, and our increasing understanding of the toxicology and ecological impact of some of the classic dye manufacturing processes. Just because we might identify the dye in say a 1950s Waterman ink doesn't mean that dye is still on the market.

 

Ah, excellent explanation. The only time I ever hear "colorimeter" used is by biology types, who really mean their ancient spectrophotometer that they can insert test tubes into instead of quartz cuvettes. I didn't realize their was an actual dedicated machine for the purpose you describe.

[Deirdre]

Lamy Blue Black 55.47 5.52 223.78

Montblanc Blue-Black 39.61 15.92 272.82

I thought these were the same?

Montblanc has two blue-blacks: the iron gall (bottle) and the regular (cartridge).

[Chris]

Fascinating - so there are blacker blacks than Aurora (such as Penman Ebony).

Chris

[mercy]

With this tool we can now tell what is the truest blue, greenest green etc all the way around the color wheel. I see that Blurple is indeed almost the same, but not exactly, as Tanzanite just like Richard has said. I have been interested in a peacock blue and with this tool I can see which one on the list would be the best candidate. Unfortunately the list doesnt include Diamine.

[dcwaites]

The big problem I see is if there is metamerism involved in the colours.

 

Colorimeters etc don't necessarily provide "perfect" information. Particularly if the colour has differing responses under different light sources.

 

In other words you may have a match with incandescent light bulbs, but not with fluorescents or the sun, etc...

 

Yeah - that makes sense. Tulipe Noir looks like *&*& under my LED bulb, but lovely in regular incandescent light.

Incandescent bulbs have a wide range of light wavelengths/frequencies, with the overall balance (warm red, white, through to actinic blue) depending on the temperature the filament is run at. Halogen bulbs run at such a high temperature that they have a UV component, meaning that if you have a halogen desk lamp, you run the risk of sunburn on your hands.

 

LED lights have very spikey frequency range, typically red, green or blue. A white LED light will be made up of combining red, green and blue emitters, so you will have three spikes, one each at red, green and blue, looking to the eye like white. These will interact with the dyes in the ink differently from the way a continuous range of frequencies from an incandescent bulb will.

 

And then, you have what fluorescent lights will do, with the UV that they put out causing some of the dyes to actually fluoresce, looking brighter than they normally would, and changing the whole appearance of the ink.

 

 

[OnPoint]

The chroma chart at rmimaging.com seems to have been removed.

[Ondina]

QUOTE(Lloyd @ Apr 10 2008, 04:32 PM) <{POST_SNAPBACK}>

QUOTE

Lamy Blue Black 55.47 5.52 223.78

Montblanc Blue-Black 39.61 15.92 272.82

 

I thought these were the same?

 

Montblanc has two blue-blacks: the iron gall (bottle) and the regular (cartridge).

 

So, are they the same, or not? I'm awfully curious, right now.

Edited July 24, 2008 by Ondina

[Gawain]

That website was just updated and "projects" no longer loads. I informed the owner.

[OnPoint]

The chroma chart owner at rmimaging.com says that the file has been restored (there were website changes that removed several things on the site).

[Robert Hughes]

Lamy Blue Black 55.47 5.52 223.78

Montblanc Blue-Black 39.61 15.92 272.82

I thought these were the same?

And I love how they're carried out to 3 and 4 significant digits. That's scientific!

 

Inaccurate, yes, but...

[Gawain]

No Noodler's on that rmimaging list of inks!

I wonder why?

[CelesteCAT]

No Noodler's on that rmimaging list of inks!

I wonder why?

 

It looks like they used data from The Ink Sampler 2001; so it wouldn't have newer info. :-(