Accurate Aircraft Paint & Colour Discussions

[x4btr]

Greetings All,

Having read this thread, I was invited by our illustrious host (!) to share a few thoughts on this topic, one very dear to my heart. Or the source of terminal nightmares!... It all runs together.

Those familiar with my work will know that I have laboured mightily to do what this thread suggests, that is to formulate some reliable digital references for describing paint colours. In the course of these attempts I have used untold numbers of monitor calibration software and/or routines, plumbed the depths of video device driver technology and collaborated with large laboratories (e.g. Samsung Electronics, Raytheon, etc) as regards monitor (CRT and LCD) function. An attempt was even made to link the colours to a digitised version of a physical colour system (Munsell). Alas, to be very succinct, all efforts failed.

I know that the remit of this thread was to remain positive. However, I must suggest that, from my own point of view, and having had a bit of experience on the matter, I believe that establishing this kind of 'accurate' digital colour reference is impossible. Or, at least between any two digital devices; it is possible, indeed, to establish accurate colour samples on your own hardware by copious testing and trial-and-error work. But, to transmit this appearance using some type of digital reference to someone else's hardware and achieve the desired appearance reliably? I just do not believe that this can be done.

This is not to say, however, that as artists, technical illustrators and profilers that we should give up on creating accurate digital colour chips for our own work! Obviously this is an important part of what we do. But, so long as digital technology continues as it is, I believe that this will have to remain a rather individual process. Furthermore, as artists we all modify colours according to our technique or preference (saturations, scale effects, surfaces, etc); for some, therefore, out-of-the-tin shades or values might not really be of the best use, anyway.

How then should one transmit their ideas concerning accurate colour information? My own thoughts on this topic have gone somewhat full-circle, I suppose. I now find myself looking for means by which to produce physical paint samples. The problem with such sets of samples is that they tend to be very limited. The Federal Standard system, naturally, was never designed to replicate aviation finishes, and thus many colours which would be required for an accurate description of aviation camouflage simply do not exist. Some colour booklets like Methuen are broader, but still incomplete and difficult to find. I believe what we really need is a modern set of commercial paint chips which are both ubiquitous and reliable. This has turned my attention towards the larger residential paint manufacturers in Europe. Some of these companies have issued very comprehensive colour booklets for their available paint lines.

Jotun, for example, has issued a very thorough paint chip book regarding their domestic paint brands (Lady, etc). These books contain a large number of paint samples, and I have the suspicion that they would provide for most aviation colour matches simply by virtue of the sheer breadth and volume of the chips in the book. Some of the other Jotun colour sample books (such as those for marine paints) are additionally valuable by providing a BS and RAL (and sometimes even a Munsell value) equivalent for each entry where possible. However, these books are rather expensive, and I have no idea if they would be available world-wide? Therein lies a potential problem.

Therefore, my question would be this: is there a some system of colour which exists in our daily lives to which one could make a physical reference of appearance which perhaps we are just not recognising? Does anyone know of a commercial colour booklet (of the type described for Jotun, for example) which is neither prohibitively expensive nor difficult to obtain? I reckon those two conditions must be met for this kind of exercise.

As regards accurate digital colour citations, I believe that all attempts to define such will only result in frustration. I say that without malice, and without attempting to abort this thread in any case! But, that is my honest opinion, for whatever it's worth....

[JMSmith]

well!

i have read this twice now, now then one of us is out of sync somewhere.

i stated at the very begining, this is NOT to produce accurate colours, you just can't do it when it comes to profiles, mind you, they can't do it when it comes to real paint.

the idea here is to produce a set of colours we all agree on to be used as a base colour before all other layers are put over the top.
we managed to do this over on AWF years ago, so why can't we do it now

just out of curiosity who are you really x4btr, i get the feeling i know you

[Clint Mitchell]

Thanks for the info! I like challenges lets go for it. We all know of the problems with regards to different monitors etc already. But as an example RLM70/71 as much as it is the correct colours that are needed it is also just as important to get the two colour shades to work together correctly as seen in period photos. Surely whatever colours users monitors twist these new updated colours into the colour shades will still work together in the desired fashion?

[Jesters-Ink]

My idea was still the best

[x4btr]

Surely whatever colours users monitors twist these new updated colours into the colour shades will still work together in the desired fashion?

Hmmm... perhaps. Typically, it takes some considerable work to co-ordinate RGB across monitors and printers. However, that is not to say that such is impossible; "merely" difficult.

But, in a broader sense, if the desire exists to have a go at creating some digital colour chips to the best values which may be possible, then fair enough. In that event, may I suggest that we begin with an agreement on RGB settings? This will be of the most help when attempting to get a printer to actually replicate what you've drawn, or another monitor to show something like the desired shade.

In my opinion, I would suggest as the best configuration that everyone will want to employ sRGB A Standard Default Color Space for the Internet - sRGB colour space in their software colour management. This should give the best compatibility results across various hardware and platforms, and certainly will with regards to printers. Is anyone using another colour space than sRGB deliberately? Could you tell us why you have done so, perhaps? It may be that in some cases other solutions are preferred.

CRT monitors should be adjusted to 6500 degs. (D65) colour "temperature". LCD monitors which still have such a listing may do likewise, but if not (most no longer have it) one will have to calibrate manually. However, I believe that it is quite unlikely that any LCD screen will need to, nor should be, set to a native Brightness value in excess of 25%, and a native Colour Balance of no more than 20/20/20 RGB. I suggest to start at lower Brightness + RGB levels to begin your calibration [e.g. on my Samsung CX2333 I have the Brightness set to 20 and the colour balance to 18/15/15].

Regardless of the calibration software employed for your monitor, I recommend to have a look at this page additionally []Monitor calibration and gamma. This outstanding discussion of the topic includes some test patterns, and I very strongly suggest to employ the Gamma black-level diagnostic shown here (it's the best I have seen).

That is my opinion for where to start as regards calibration. Does anyone have any other tips or ideas on this topic? Or, perhaps does anyone disagree with these choices? I reckon that if we start from the same fundamental settings that better results are the more likely.

K-- mate, your ideas are always 'the best', any road! lol

[Clint Mitchell]

Hi X4btr,

I was told by my printer to edit and save my artwork in Adobe RGB (1998)?

[JMSmith]

now listen here jester,

if we did it your way, you would be first dead, you know your as blind as a bat

X4,s monitor calibration looks good, all i need is time to do it, running a Samsung Syncmaster 913v and a couple of programs for setting it up
the black box with the two dk grey bands looks perfect, but the two grey squares have a distict yellowish band runing through them
i wish you guys would remember i am only a truck driver, a useless one at that

are you going to let me in on who you are X4 its killing me thinking i know you

[x4btr]

I was told by my printer to edit and save my artwork in Adobe RGB (1998)?

Adobe RGB-98 is essentially sRGB, but with certain small differences. This is not a bad choice, therefore. However, the sRGB offering in Photoshop and Illustrator since version CS is slightly improved vis a vis "white point" positioning and colour substitution, and I recommend it in preference to the older Adobe scheme. This will impact mostly upon printing, yes, but even so there could be an improvement with respect to your monitor's appearance with the change.

Why not have a go? Open some work in sRGB and examine/print the results (do not save work after colour conversion); you can always revert if you don't fancy them.

are you going to let me in on who you are X4 its killing me thinking i know you

Well, if I am using a using an ambiguous log-in name and not applying my identity to any signature, it would suggest that such a question would be best handled via PM, yes?

In the meantime I will open some of the proposed chips in this forum and see what they look like on my hardware....

[Clint Mitchell]

Thanks X4btr, I will give it a go.

[JMSmith]

hi X4BTR,

na, can't be bothered, learned a long while ago when running AWF that people that hide behind usernames are not worth pushing. a lot of serious sites are now asking for real names to be used, a more open policy.