Colour Analyser On Line.

Hi guys...Try this link, I was hunting around to try and come up with accurate colours for Canadian stamps.
http://labs.tineye.com/


Here is what I was playing with...

Wert,
TinEye has a good reverse image lookup.

And this tool is good, but it is not identifying the color of your stamps. It is identifying the color that your scanner/video card/PC and associated drivers are generating.

In other words, you could send me the same stamp and it would ID different colors than what you came up with when I scanned it on my scanner and PC. In fact, if you close or open the cover of your scanner you will probably get different colors.

There are calibrated instruments that will identify colors such as pantone. Here is an example http://www.pantone.com/pages/Produc...id=1032&ca=1
There are not cheap, but have controlled light sources which will yield consistent results. So you could scan and identify a Pantone color of a stamp and another person would get the same results if they also owned the same instrument.

But you will never be able to use your scanner/video card/PC to do this unless everyone else has the exact same setup (including exactly the same version operating system, drivers, etc.).
Don

Don...I was hoping my friend that you would chime in...color your world uses the same (almost) portable system as you mentioned...kind of expensive....well you know me...back to the drawing board...thanks Don.
ROBERT

Hi Don, You are 100% right. That is why I never trust or comment on colors or shades of stamps posted online. I look for varieties of printing and errors online. If I am looking for a specific color or shade I want to see the actual stamps. By the way even human eyes see color differently between individuals. That is why I don't worry about minor color differences. I collect the ones I find and catalog them as shades.

Yes, but ... this may not matter.

It is my thought that, no matter how many of your neurons are firing, or which of your neurons are firing, the 'distance' in the color spectrum from pink to carmine (for example) is going to be pretty much the same for everybody.

Therefor, even though the 'absolute' colors differ to different human eyes, we would pretty much all see the same differences.

First one to think "vive la difference" loses.

Cheers,

/s/ ikeyPikey

Yes we would both see a color difference. My point is the name of that color requires a learned thought process. What I might see as red might be orange red to someone else. It is how we perceive the colors that differs. Just as a device like a scanner interprets the data scanned to determine a color. Most scanners also have software that can be used to alter the way the scanned data can be interpreted to determine color and saturation. Our brains essentially do the same thing and relate it to a learned color. If I learned the color called red was white then I would call red white when I saw it. This is a bit extreme but what we learn as a color name is how we see that color or shades there of.

I think you are treating a two-step process as if it were a single step.

Scanners react to photons. Physical variations (in the source, in the mirrors, in the lenses, in the sensor pixels) result in digital variations, the character of which are lost forever. They are entirely deterministic (physical, replicable, etc) but, to the downstream viewer, might as well be random, though probably with a normalish distribution.

Downstream software uses those digital values to 'determine a color'. It may or may not have been calibrated to the scanner; the samples used for the calibration may or may not have been true; and the calibration may or may not have preceded later changes (aging, dust, etc) in the scanner.

But I was not addressing the use of software, but of the Mark One Eyeball.

Yes, we agree that colors are learned. But that does not mean that they are learned differently by different people, and it does not mean that they are used in a vacuum.

If a stamp comes in pink & purple, to pick an example that it easy for me, there are only those two possible colors. Even if my spectrum is different than yours, my pink is still in the middle of my red (with white to lighten), and my purple is still somewhere towards blue (to keep it simple, and avoid less-familiar words like cyan & magenta & all that).

We should still be able to agree which is pink & which is purple, especially if we have good reference samples of the stamp in front of us.

Where we will get, uh, color-blind-sided is when one of us agrees with the software, and the other does not.

It is up to us to decide if we will be Ruled By The Tool.

I would suggest keeping a nice stack of cheap stamps to use as reference samples when color really does make a financial difference. If Scott X (more pricey) comes in pink & purple, and Scott Y (less pricey) comes only in pink and Scott Z (less pricey) comes only in purple, we've got it made.

But let's keep the software out of the loop.

Cheers,

/s/ ikeyPikey

If all we had to separate was pink from purple, the vast majority of us could do it, but most stamps that have different shades are much closer together than that and each of us not only sees them differently (our eyes are not the same and even our own eyes change over time, plus we learn names of similar colors differently), but also the conditions we are viewing them in(light source, what background they are against) will greatly affect our perception.

Wouldn't it be cool if we all saw colours according to our unique brain make-up?

Eg. Red to me is actually Blue to Bob and Green to Sue in our brains;

we all call Red the same name because when we look at an object, our brain sees the same colour we are used to calling Red (though if Bob could see what I see, it would be Blue to him and for Sue it would be Green).

Did I mess you up at all?

WOW 3Dadeo...You think like me..I have always told people that the green I am looking at and if I could see through their eyes it might look pink, of blue..Guess we will never know..



its just a tool ikeyPikey...It will never replace a steady hand, a good eye and a magnifying glass...just a tool....

You're right wert, I guess we think the same. To me it is a possibility only because I ask myself "Why not?" God could have made us that way just because.

No human being understands the brain to that degree.

I like the quote (sorry I forget the source), "If our brain was so simple we could understand it, we would be too simple to do so."

ikeyPikey...I agree with what you are saying.

All our scanners are different...All our monitors are different...All our colour perception is different...That is why I (THINK) that I may have a solution that takes these electronic devices out of the loop when it comes to colour classification of stamps.

Listen to what I am going to say and tell me if you disagree with my idea...??

I have a paint store I always go to and they have a machine that they put a paint sample under and assign a colour number to it..(example #32899)...Now two things can happen.

1- I can get them to show me their "paint sample" strips that you can take home...Then I can hold it up to all the stamps that I am classifying.

2- I could make a sheet with all (or part) of the sample painted on say white cardboard.

Now YOU could go to the same paint distribution company and get the EXACT colour associated with the number I supply...Everyone who is given the numbers could build their own classification sheet and NO scanners...NO monitors involved in the loop.

Am I making sense..??

wert, I understand your idea and agree.

Richard Morris used paint chips in his Admiral and Small Queen colout guides for this reason. These are the most stable samples of colour; they do not fade or change over time.

He used multiple samples of stamps to match the accepted names to the chips based on previous expert identifications of these stamps.

That is about as close as one can get when it comes to matching a colour with the catalog description.

Of course, there will always be some room for doubt.

Wert,
We've covered this before. The paint stores can match colors because they are using a device which control the light source. As long as the device is in calibration (which they do not have to do) the THEORY is that it should be an exact match with another paint shop using the same model device in another location.

But if the device is not calibrated and inspected on a schedule, then the ability of the device to match an identical color becomes compromised. For example the light source (bulbs) grow weaker over time.

And the paint stores are probably not going to standardize on a color naming convention (such as Pantone) but rather some paint line marketing names. And I am unaware of any standardized naming color convention used across all stamp catalogs; what Scott calls 'Rose' may or may not be the exact same shade as what Gibbons calls 'Rose'.

I have supplied you with links to device which can match colors. They are expensive for a reason. They required regular inspections, calibration, and re-certification to consistently match colors properly. There is an entire industry surrounding color matching and color standardization and I see no indications that philately is in sync with it. There have been some great stamp books and efforts which attempted this but none succeeded in becoming universally adopted.

There is only one way to make his happen. Get philately to standardize on a common color chart (like Pantone), have everyone agree upon a color matching device that everyone buys, and then put into place a way to keep the devices all inspected and calibrated.
Don


Edit; ALL paint chips, even those made of plastics, change color over time. We normally buy new color charts every 2-3 years due to this. Out-gassing, light, and other environmental conditions can, and do, change paint chips.

Well with all the negative remarks to me at least trying to help stamps collectors with colour matching, I guess it is time for me to give up...

Onward and forward to me next project to be down played....haha

Kind sir, I have never made a negative remark towards you in any way. I have indeed pointed out the challenges of trying to standardize on colors found in the stamp world. These are real challenges but meant to help you understand the issue(s) better. First, you wanted to figure out a way to do this with each stamp collector having different equipment. Then you suggested that paint store equipment be used. In both cases the need for a standardized color chart remained undefined.

To solve the issue of a color standard in philately you will need to address these specific issues. So please don't shoot the messenger here, if there had been an easy solution to this issue we would all already have had it.

Everything moves forward by bringing ideas to the table and allowing others to mark them up. No one should be insulting each other but that doesn't mean others will not point out things that don't work or make sense.
Don