Color Match 1.0

That's a real great idea, and it can really be handy.  I'm impressed by the logic behind it and I've written a lot of software over the years.

The RGB function is difficult to get good results from, but it isn't the fault of your code I don't think.  For example, This Photo of an A-10 I built a couple of years ago.  The main fuselage that is visible in that photo is Testors Acryl Light Ghost Gray.  I picked up several RGB codes from various places on the image and the closest it got was Dark Ghost Gray.  It's most likely just some slight color shift in the photo itself since I didn't take the photos in sunlight.  I used incandescent light and then color corrected the images.  I think it might be a good idea to mention this to people (and perhaps you did in the FAQ or article).  The match that one gets from the RGB code is only going to be as good as the photograph itself, and light makes a HUGE difference in what is perceived.

Another example.   The three photos below are of the same F-15C taken on different days:

Photo 1
Photo 2
Photo 3

In fact the second and third photos were taken on the same day, and the first one was 2 days prior.  It is easy to see how different the colors of the fuselage are. 

A great bit of work though.  Very well done.

Unforunately the RGB input has a number of limiatations that can cause issues - though I'll double check the Light Ghost Gray color tomarrow (should be able to pick up atleast two additional samples here in town).  That said...

RGB color gamuts come in a number of different flavors, and you also have the issue of jpeg files shifting slightly darker than the actual color that they are.  You can see an example of this going on here:  Article regarding scRGB Colors (scroll down to the side by side comparison of a JPEG and HD image).  Assuming that the image is encoded properly (unless you are specifically trying to, it is unlikely that it is - not any fault on your part, these issues are not generally covered in scanner and camera manuals), depending on what RGB gamut the eyedropper tool is using in order to sample the colors - you can have issues there.  Since the most readily available and free tools tend to be designed for HTML sampling, that is what I have based the RGB conversion formulas off from.  More advanced software like Photoshop, PSP and even the GIMP allow you to configure that yourself...but most freebies like ColorCop are much more limited.

After all those issues are dealt with - you also need to get an eye for the shape and lighting of the subject.  In your example picture, the top of the fuselage is right out...most of it is a hot spot so to speak.  Along the same lines so is the bottom of the fuselage.  The best place for sampling seems to be the area behind the "Rescue" decal on the lower half of that plate.  When I sampled there, I got two matches - Lichtgrau and Light Ghost Grey.

Anywho, thanks for the feedback - and I hope that bit of explanation goes aways to explain some of the limiatations I am fighting against.  I'll write up a better explanation on the RGB issues later tonight and send it off to be checked (I can tend to get a bit technical in explanations causing more questions than answers ).

An remember, it will get better with time.  If you find colors which are obviously off - let me know.  I'll take new color samples and double check them against the database.  I've fixed several already, but with 1500+ paints I am sure there are still a few corrections left. 

Wow, great tool, Joe! Thanks for writing it. Scott is perfectly correct about images taken at different days, or even time of day, so research research research is the name of the game.

Bookmarked your site and will be giving it a try soon. 

Tonight I did an update to the program, split the finishes from the type of paint so you will be able to search for Flat Acrylics, Glossy Acrylics, Flat enamels and other combinations.

There were also a few backend updates that will enable future upgrades with less effort on my part (and a slight increase in performance).

One of those upgrades plays into the issue of the photos.  I am sure that most of you are familiar with the concept of scale lighting and colors.  This comes into play quite often when taking pictures of aircraft in flight (even though you are using a zoom lens, the colors still tend to be darker than if you take the same picture relatively sized but at a closer distance).  But - the next two features that I am adding are a tool for handling scale color and color correction due to distance.  The other part is along a similiar lin, but it will allow you to calculate shade and highlight colors for a given source color (more useful when doing dioramas and shadow boxes where you are simulating lighting as well).

Anywho, thanks for the comments and feedback from everyone.