Does anyone know how to measure the color of a Post-It note?

[Xelikai]

I'm not sure if this is the right place for this, but since I'm new here i'll post it here until someone tells me it should go somewhere else. Anyway, I am currently wondering how to measure the color of some post it notes. It started as a friendly argument as to which post it pad was more red, a hot pink, or a pinkish salmon color. After causing a riot in the AP art class at my high school, we could not determine which was more red. I've tried an online histogram of the post it notes, the problem is that it is giving me the same graph for both sticky note pads. Does anyone have an idea on how to scientifically determine which pad is more red?

 

[Borek]

"More red" is a can of worms, unless you define first what you really mean by "red" and what you mean by "more".

Even then measurements will be quite difficult. It is not something unusual to be not able to reproduce a color in print just because precisely calibrating scanner/monitor/printer is not easy.

 

[.Scott]

First off, color is in the eye of the beholder - and not all human beholders agree on the fine distinctions of hue.

The surface of the post-it note reflects different amounts of light across the whole spectrum. The human eye summarizes that response into a hue - which is not only dependent on the retina, but on the lighting and how close the note is to other contrasting colors.

That said, try this:
Agree on a lighting condition. Take a photo with both pads. Pop it up in MS Paint or Gimp, or whatever. Sample the color from the photo and view that color in terms of hue, saturation, and intensity. Red shows up as a hue of zero (right next to 1 and 255).

 

[Xelikai]

So assuming that the actual color of red is the wavelength of 620-750 nm, the average wavelength of red, or the "purest" red, would be a wavelength of 685 nm. So if I could graph all of the wavelengths on a graph of x=wavelength and y=intensity (i'm not sure how I would measure intensity), then calculate the area under the curve, and compare the average to the "pure" red of a 685 nm wavelength, would that yield a viable answer? If it did, how would I even go about doing that? To give some background, I'm a high school student, so I have nearly no education in topics such as these.

 

[.Scott]

You would be looking for a spectrometer. Such at this:
http://ikancorp.com/productdetail.php?id=1975&model=mk350d

This is not something that your high school is likely to have.
But the right technical college may be willing to demo one for you.
Some State labs might have one - and they are often open to demo their equipment to high schoolers.

 

[russ_watters]

So assuming that the actual color of red is the wavelength of 620-750 nm, the average wavelength of red, or the "purest" red, would be a wavelength of 685 nm. So if I could graph all of the wavelengths on a graph of x=wavelength and y=intensity (i'm not sure how I would measure intensity), then calculate the area under the curve, and compare the average to the "pure" red of a 685 nm wavelength, would that yield a viable answer? If it did, how would I even go about doing that? To give some background, I'm a high school student, so I have nearly no education in topics such as these.

Did you read the replies you got so far? That approach is missing some key parts of the issue. I'll try again:

1. The reflectivity of the paper will be different for different wavelengths irrespective of light source, BUT:
2. The color of incident light is different for different light sources.

So the measured color depends on the color of the light source. Simply put: it will look redder outside in the sun than inside under LED lights.

So your first task is to agree on the light source.

Then you can talk about measuring the reflected light. And there are several methods, some already mentioned (like a camera or scanner). But if you want it on a by-wavelength basis, you need a spectrograph (spectrometer).

[edit] All that said, I don't see how that would be more useful than the Photoshop answer, which would be in the form:
50% Red
30% Green
20% Blue

 

[russ_watters]

You would be looking for a spectrometer. Such at this:
http://ikancorp.com/productdetail.php?id=1975&model=mk350d

This is not something that your high school is likely to have.
But the right technical college may be willing to demo one for you.

Hmm... putting some thought into this; shouldn't it be possible to make one using a diffraction grating or prism and a monochrome CCD?

 

[ZapperZ]

So assuming that the actual color of red is the wavelength of 620-750 nm, the average wavelength of red, or the "purest" red, would be a wavelength of 685 nm. So if I could graph all of the wavelengths on a graph of x=wavelength and y=intensity (i'm not sure how I would measure intensity), then calculate the area under the curve, and compare the average to the "pure" red of a 685 nm wavelength, would that yield a viable answer? If it did, how would I even go about doing that? To give some background, I'm a high school student, so I have nearly no education in topics such as these.

Scope around the universities near your area and see which one of them has a UV-VIS device and is wiling to perform that measurement on your sticky notes. The reflectivity intensity will be rather low, but who knows, you may get a spectrum and compare which spectrum has the highest intensity near the 685 nm value.

This might make a nice little science project, to compare human perception of color versus what actual measurement says.

Zz.

 

[OmCheeto]

...a can of worms...

Agreed.

IMHO, discussing "color" is the closest you can get to "Politics + Religion" in science.

Color Political Parties:

Humanists: Color is a human brain construct!
Scientists: Color is a frequency of the electromagnetic spectrum!
Moderates: Both are correct!​

Color Religious Affiliations:

Evangelical: If you don't believe as I do, then you are wrong!
Atheists: There is no such thing as color!
Unitarian: Both are correct!​

hmmm... I think most here at PF would fall into the "Moderate Unitarian" camp.

So assuming that the actual color of red is the wavelength of 620

According to this website and my eyes, 620 nm is Orange.

-750 nm,

that looks like "Brick" to me.

the average wavelength of red, or the "purest" red, would be a wavelength of 685 nm.

Not a bad rendition of "Red".

So if I could graph all of the wavelengths on a graph of x=wavelength and y=intensity (i'm not sure how I would measure intensity), then calculate the area under the curve, and compare the average to the "pure" red of a 685 nm wavelength, would that yield a viable answer? If it did, how would I even go about doing that? To give some background, I'm a high school student, so I have nearly no education in topics such as these.

hmmmm... High school was also the the highest level of education I completed.

a friendly argument as to which post it pad was more red, a hot pink, or a pinkish salmon color.

Ugh... I would suggest you send an email to the artist Stuart Semple. He claims to have created the worlds "pinkest pink".

And "pink"? Ehr mehr gerd. According to this guy, pink should be called "not green".



ps. Strangest things I've discovered this morning thinking about your question:

1. The red cones in our eyes are most sensitive to spectral yellow light. It's only the lack of input from the green sensitive cones that makes things look red, as we move to longer wavelengths [ref]
by eyeballing his graph {590 nm}:

2. As far as RGB levels go, 'black' is just a really dimmed down version of 'white'.

name   Hex        RGB             HSV
White  #FFFFFF    (255,255,255)    (0°,0%,100%)
Silver #C0C0C0    (192,192,192)    (0°,0%,75%)
Gray   #808080    (128,128,128)    (0°,0%,50%)
Black  #000000    (0,0,0)          (0°,0%,0%)

[ref]​

 

[anorlunda]

Paint stores do that every day. You can take a paint chip (or a postit note) to the paint store. They scan it and it tells you the amount of pigments to add to a white base paint to match the color.

Obviously, the device is a digital camera. It controls the illumination, and the result is sensitive to the reflectivity, and variations in color throughout the sample. Everything else is a matter of calibration and semantics (the names we associate with colors).

The OP might be able to bring his postit notes to a paint store and if he or she is charming enough, convince them to scan them in their machine and to tell the results.

I think the following from Wikipedia is interesting. It says that the machines use two different illumination colors.

There is no straight-line relationship between colors mixed in pigment, which vary from medium to medium. With a psychophysical color circle, however, the resulting hue of any mixture of two colored light sources can be determined simply by the relative brightness and wavelength of the two lights.[13] A similar calculation cannot be performed with two paints. As such, a painter's color wheel is indicative rather than predictive, being used to compare existing colors rather than calculate exact colors of mixtures. Because of differences relating to the medium, different color wheels can be created according to the type of paint or other medium used, and many artists make their own individual color wheels. These often contain only blocks of color rather than the gradation between tones that is characteristic of the color circle.

it also says

The psychophysical theory behind the color circle dates to the early color triangle of Thomas Young, whose work was later extended by James Clerk Maxwell and Hermann von Helmholtz. Young postulated that the eye contains receptors that respond to three different primary sensations, or spectra of light. As Maxwell showed, all hues, but not all colors, can be created from three primary colors such as red, green, and blue, if they are mixed in the right proportions.

 

[sophiecentaur]

Hmm... putting some thought into this; shouldn't it be possible to make one using a diffraction grating or prism and a monochrome CCD?

A friend of mine looks at stars that way. The luminance values (after some calibration) tell him the temperature, significant gases etc. etc. when he plugs in some freebee software.
The paint store idea would be good. I would bet that the CMY / RGB values for all the manufacturers' ranges are available. Many of them are standard colours.

 

[Merlin3189]

If you're not likely to have a spectrometer and the time to do that sort of measurement, maybe you can find an old photographer who still has a simple light meter. (Or just make one.) If you then measured the brightness of each post it under identical illumination from some red light sources, you might see which reflected most light. But I expect the difference would not be great enough to be noticed on the log scale of a photographic meter.

The eye. is not absolutely calibrated nor fixed in sensitivity, so you have do comparisons side by side and simultaneously. Have a look at how people used to compare brightness with wax spots and the like.
To avoid the question of which is redder, you can ask people to make them look the same. If the two samples are lit by two sources whose colour can be controlled, you should be able to make them appear the same. Then you record the settings of the illumination on each. Use plenty of trials for each subject with the pads being swapped sides pseudorandomly. If there is any systematic difference, you should be able to detect it.
A big difficulty is training the observers. You can practice yourself and then help the observers. So long as you yourself are also blind as to which way round the pads are (have them set by someone else) I don't think that would introduce bias.

 

[CWatters]

In industry the Pantone colour charts are commonly used to specify and identify colours but their sample books are expensive. You could try asking the manufacturer of the post-it if they can tell you the Pantone colour reference then look up the hex value for that colour which I think is in the format RRGGBB. That would tell you how much red Pantone think is in that colour.

 

[Borek]

Pantone colour reference

I can be wrong, but I don't think every Pantone color can be converted to RGB, some printable (CMYK) colors are outside of what can be displayed in RGB (and vice versa).

 

[sophiecentaur]

I can be wrong, but I don't think every Pantone color can be converted to RGB, some printable (CMYK) colors are outside of what can be displayed in RGB (and vice versa).

But, to be fair, @CWatters was referring to a Post it note, which is well within the gamut of any three primaries that you will find in any real imaging system.
The only problem with doing the Pantone things is that a set of Pantone Reference Colours will cost you an arm and a leg (I saw £500 to £1500 on the Pantone site). If you want to use your computer monitor with the Pantone on-line system, you would still need to calibrate your monitor; more expense.
This thread is getting to be typical of PF threads. The OP doesn't specify an accuracy for the measurement so everyone is posting opinions based on their own assumption about requirements. Good fun but we have reached the end of the line until we can agree on how accurate we want to do this measurement.

Personally, I reckon going to the local paint supplier and getting a handful of colour charts would do the trick. Especially if you manage to get a repeat match outdoors on sunny and cloudy days and indoors with tungsten, CFL and LED lighting. (I believe in Father Christmas and the Tooth Fairy.) The same thing would apply to Pantone, too.

 

[Xelikai]

So based on the responses, I've come to the conclusion that I will be doing two-three sets of tests on these notes. I found a teacher at the print shop of our high school who has the Pantone color charts, and has agreed to let me use them. Second, this weekend I will be looking for local Paint shops to see if any of them can scan the notes and would be willing to scan them for me. Finally, I will be asking them if they would be kind enough to send me an email of the images they take, where I can upload them into MS Paint or a similar program, and measure the RGB/CMYK point values of each note and determine which has a higher red composition.

 

[sophiecentaur]

I found a teacher at the print shop of our high school who has the Pantone color charts, and has agreed to let me use them.

Brilliant. Exploit any resource you can access. That's the secret of progress in Science.

 

[russ_watters]

Second, this weekend I will be looking for local Paint shops to see if any of them can scan the notes and would be willing to scan them for me.

"Paint shop"? I don't understand why a "Paint shop" would have a scanner, but there are tons of places to find a scanner - every office will have one or several. Heck, I have 3! Regardles...

But I need to emphasize a major pitfall you still seem to be ignoring: YOU NEED TO TELL THE SCANNER WHAT COLOR LIGHT TO USE WHEN MAKING THE SCAN*.

Sorry for yelling, but it still doesn't seem like you recognize that the color of an object depends on the color of the light shined on it.

*If equipped; if not equipped, you will need to find out the scanner's lamp/scan color and do your own calibration. See here for discussion of the issue:
https://forums.macrumors.com/threads/need-to-capture-accurate-colors-from-scanner.1903569/

 

[CWatters]

Many DIY paint shops will mix paint to match a sample. They use a scanner for that but I don't know if it will tell you anything.

 

[sophiecentaur]

Many DIY paint shops will mix paint to match a sample. They use a scanner for that but I don't know if it will tell you anything.

A bit of snake oil, perhaps as it would depend as much on the lighting that's used in your particular room. However, if the paint sample is from the same manufacturer (same basic primaries), then a simple RGB / CMY match would tell them which the original shade was. Near enough for a match where you dumped a bottle of wine on your parents' wall at the party you had when they were away. The carpet is another problem.
Paint colours are so consistent these days that there's hardly ever any need to mix two tins of paint together to avoid a visible change from one wall to another from a different can.

 

[Merlin3189]

It started as a friendly argument as to which post it pad was more red, a hot pink, or a pinkish salmon color. After causing a riot in the AP art class at my high school,

Even if there had been a clear "scientific" answer, would that have ended this sort of argument? I see what I see and you see what you see.
Probably the most scientifically valid way of deciding is to do what they already did - ask people to look and decide. There may be Psychological-scientific ways of improving their procedures, but what people see as red defines what red is.

Perhaps the most useful answer they can get here is that, color is a much more complicated business than they may have thought.

 

[sophiecentaur]

There may be Psychological-scientific ways of improving their procedures, but what people see as red defines what red is.

You are taking all the fun out of this.
Fashion companies spend millions on making a PVC handbag the 'same' colour as the Cotton Dress and the Polyester belt. Colour matching can't just be dismissed as smoke and mirrors. It really works for such a big proportion of the population that it has established itself as a valid technology.

 

[Merlin3189]

. Colour matching ... really works for such a big proportion of the population that it has established itself as a valid technology.

Isn't that what I said? Red is not defined by physics nor physicists, but by what ordinary people say is red. Yes, some people are abnormal, so we take the general consensus.
The important thing is not to ask the physicists! Some are ok, but too many reverse the, "light of 640 - 700 nm looks red", to define "red is light of 640 - 700 nm. "

As for fun, the OP and friends could have a lot. I'd love to have a class to do practical work on this But arguing about it here I don't find to be much fun. I enjoy explaining things to people, but I get very frustrated if I think the experts here are actually wrong.
PF rules that everything is supposed to be expert, referenced and accepted knowledge. That makes it difficult to talk about topics like this, which are perhaps not so simple and clear cut. How can there be any argument once an expert has pronounced?

 

[sophiecentaur]

Red is not defined by physics nor physicists, but by what ordinary people say is red.

Absolutely. However, my point is that it is very much a matter of Physics (Science, in general) to predict what combination of one sort of primary will match what combination of different primaries for the majority of people. The name that's actually used is up to the 'creatives' in the company that provides the pigments but the matching has to be systematic. The swatch of 'reds' can contain hundreds of colours and it's not feasible to expect a human to carry one of those patches in the memory and identify it the next day, in a different situation. A 'name' for the colour won't help. That job has to be done by instrumentation and that side of things needs the same rigour as any other bit of Science - hence the need for referenced knowledge, As for "clear cut", there is a vast amount of frontier Science that is far from clear cut but there is also a lot of stuff, well inside our present model that is as clear cut as the fact that the Sun will rise at a given time on any day next year.
Personally, I am stoical about the 'experts' being right when I don't want them to be. The only weapon against that is to prove them wrong with a good reference or some experimental proof (beyond my finances in most cases.)

too many reverse the, "light of 640 - 700 nm looks red", to define "red is light of 640 - 700 nm. "

They are just ill informed (the politest terms I can think of).

 

[Hendrik Boom]

A bit of snake oil, perhaps as it would depend as much on the lighting that's used in your particular room. However, if the paint sample is from the same manufacturer (same basic primaries), then a simple RGB / CMY match would tell them which the original shade was. Near enough for a match where you dumped a bottle of wine on your parents' wall at the party you had when they were away. The carpet is another problem.
Paint colours are so consistent these days that there's hardly ever any need to mix two tins of paint together to avoid a visible change from one wall to another from a different can.

Yes, lighting. Back in the 60's I was at a party where different parts of the room were illuminated with different colours of lights. My girlfriend had red eyes under a green light and green eyes under a red light. Complicated, indeed!

 

[sophiecentaur]

Yes, lighting. Back in the 60's I was at a party where different parts of the room were illuminated with different colours of lights. My girlfriend had red eyes under a green light and green eyes under a red light. Complicated, indeed!

The secret of good paint matching (fabrics and paper is to get a match under widely varying light conditions. This is only possible if the spectrum of reflected light is wide band (not narrow band peaks) so that the eye's analysis gets an input from all wavelengths and then it can some with more varied illuminants. A peak in one narrow range of wavelengths can miss some of the wavelength bands that you can get with CFLs and that will give lunatic results.
Those CFLs were a temporary blip in the history of colourimetry and we are well rid of them, imo.

 

[James Demers]

The RGB values assigned to a color are highly relevant to your question, because they're designed to be relevant to RGB monitors, which in turn are intended to serve as input devices for the human eye. I would photograph the two pads side-by-side, under any reasonable lighting conditions - the flash from a digital camera or cellphone ought to do nicely. (The light source affects appearances, as notedd by others, but any reasonably white light won't much affect the relative "redness" of two similar colors.) Open the image on a Mac using Preview, and use the color picker as shown above to get the RGB values. (On a PC, Photoshop or the like will do the trick.) Let's call the results r, g and b.
If you can agree that the "redness" of a color can be measured by its distance from pure red (the lower left corner of the above color cube), you have a re-statement of your question that can be answered quantitatively. The distance from pure red (255, 0, 0) is the square root of [(255-r)² + g² + b²].
In the example above, all points on the front plane of the cube are fully red (r = 255), but the selected point is removed from the pure red corner by g=129 and b=139. Per the formula, the distance from pure red is the square root of [0² + 129² + 139²], or about 189.6. (Your colors are likely to have values of R that are other than 255, meaning that they're somewhere within the volume of the color cube, but only the surfaces are visible in the image of the cube I'm using as an example.)

 

[sophiecentaur]

View attachment 220738
The RGB values assigned to a color are very relevant to your question, . . . . . . . . .

This is ok as far as it goes but you are concentrating on Colour Synthesis. The really important process is the Analysis that your eye does and, without getting onto Primaries at all, there are many different spectral distributions that will give any chosen 'colour', as appreciated by the brain. Matching to a given paint sample, using RGB values from an analyser will not tell you the spectrum that's being reflected off that paint. Whilst the Colour TV approach is very successful and definitely near enough for many purposes, it is an approximation. A spectral analysis of the light from a surface plus the spectral analysis of the illuminant will tell you much more and it will allow you to produce a 'perfect' match under any other illuminant.
I just mean that it ain't that simple.

 

[James Demers]

A spectral analysis of the light from a surface plus the spectral analysis of the illuminant will tell you much more and it will allow you to produce a 'perfect' match under any other illuminant.
I just mean that it ain't that simple.

Agree 100%, but the question of "which is redder" is a question of perception, at least to my way of thinking. To the extent that one agrees on the use of distance in RGB color space as the measure of "redness", I think my solution answers the OP's question.

For example, when asking which of two colors is "more yellow", it doesn't matter whether a perceived "yellow" is yellow light, or the equally-yellow result of red and green light. Spectral analysis can tell you this, but it doesn't help answer the question.

 

[sophiecentaur]

Agree 100%, but the question of "which is redder" is a question of perception, at least to my way of thinking. To the extent that one agrees on the use of distance in RGB color space as the measure of "redness", I think my solution answers the OP's question.

For example, when asking which of two colors is "more yellow", it doesn't matter whether a perceived "yellow" is yellow light, or the equally-yellow result of red and green light. Spectral analysis can tell you this, but it doesn't help answer the question.

No problem with that except that your RGB space relates to particular primaries, I think. The tristimulus values X,Y (and possibly Z) would not be 'primary specific'.