Casey

Are you aiming for paint resiliency? Vallejo are made to withstand a beating (helmets, RC cars, resistance to fuel residue, heat), especially with urethane varnish. The color selection suggests more "shiny car" and primaries for mixing, not "exact historical colors replica".

If something works counter intuitive, it is awesome thing because it allows new understanding. Here is my quick theory: DecalFix may be containing some form of paint stripper (like Tamiya one). It may react to the top surface, and not cause the discoloration itself but make surface delicate and combined with whatever you did to it modified the surface to effectively became more matte, which gives 'white patch' effect. For now this is theory but maybe it would be worth setting up an experiment to confirm or invalidate it - that kind of troubleshooting guide would be very useful for others. First step is to make the process repeatable., but I do not have revel coat so I can't test it out quickly. Any volunteers?

What amuses me most are the names associated with this math. Bernhard Riemann, Hermann von Helmholtz and Erwin Schrödinger are pretty well known so it is quite a 'wow' moment for me. Even if the effect applies to perception of large color differences, and has not much meaning for precise color matching, it is quite applicable for other topics, like computer vision, and color communication as the paper suggested (quote "improve color maps for data visualization, to make them easier to understand and interpret").

SDS is a Safety Data Sheet detailing the composition of the product. It is intended for safety purposes and should list all the ingredients and their safety precautions. Some companies go quite open about their products, for example I found Tamiya SDSes that actually give the % ratios of pigment composition in their paints. Humbrol is much more secretive for some unknown reasons. Maybe we should look for the https://en.m.wikipedia.org/wiki/Safety_data_sheet Here are examples of one I found: http://www.hegner.co.uk/downloads/msds/SDS0161.pdf for Polycement produced by Colart Fine Art & Graphics Limited for Humbrol, Humbrol Matt/Satin/Gloss spray - https://safety365.sevron.co.uk/substances/accessSDS/SDS-6041-577288528b6b11.58665871 I have ones for enamel paints and so on but I cant find Decalfix SDS that easily for some weird reason. Probably same reason why it is easier to find a link to humbrol mythical 3D printer on their page than the safety sheet of their really available probably not safe to drink liquid (https://support.humbrol.com/hc/en-gb/articles/5615501361682-Humbrol-Creator-3D-Mini-Printer)

A lot of people have successes using simply diluted floor polish instead of decal fix to eliminate the silvering - Combine it with some form of decal softener and thats most probably what DecalFix is, but until I see the SDS I can only speculate by looking of how it behaves in practice. It does leave permanent glossy residue on matte paint.

I would love to see SDS for decalfix to get that mystery solved.

If you are talking about decalfix from Hunbrol, this liquid is an diluted gloss varnish by itself, if you apply it on a matt surface you will see it. You did not mention what paint you did use. Some paints react badly to some of the decal liquids, for example Mission Models + Microscale Microsol and Microset caused first discolorations on a paint for me and then worked as paint stripper...

Sadly I do not have those, so I'm not having much except the recipes. To do an exact match I need a spectral measurements, so I'll make a Tamiya mix myself.

FYI - Original BS381C from 1996 and the recipe Left, the BS381C 223, right: recipe. Transparent Yellow Iron Oxide: 5 Bone Black: 1 Chromium Oxide Green: 1 I always thought the color was green because of actually containing chromium oxide, and the lack of this pigment caused the color to be phased off in later years. And this is a quote from @Mike Starmer - https://www.tapatalk.com/groups/missinglynx/tamiya-mix-nobels-khaki-green-no-3-g3-dark-green-t155571-s12493901240.html But the samples you show are much more brown.

If I had a spectrophotometer data for this color I could do a match... otherwise it is bit hard. Your last result may be because of the finish. Golden fluid has no fillers, you get the paint finish exactly as the pigments are, bone black for example is very matte, and quinacridone magenta is extra glossy. The cheaper paints usually have a lot of filler, and the matte medium is one of them. Also, I wonder what is the validity of equivalence of "BS 381C 223 Middle bronze green" to that color.

Another very good resource is our own @Jamie @ Sovereign Hobbies Shows a really good practical example and an simple explanation why the widely-used base color probably was not having much, if any expensive pigments. It may also help you a lot with mixing the G3... With modern pigments, I would try: 17 parts raw sienna,6 parts bone black and maybe 2 parts of phthalo green yellow shade. You can check it on Golden mixer. Or replace phthalo green with green chromium oxide which is probably more historically accurate but it is much weaker pigment so you will need different proportions.

There are many good publications, for varying level of technical needs: Easy: Books: 1500 Color Mixing Recipes - https://www.amazon.com/Color-Mixing-Recipes-Acrylic-Watercolor/dp/1600582834 Or more technical information, which can be bit pricey but they have wealth of knowledge: "Billmeyer and Saltzman's Principles of Color Technology" - https://www.amazon.com/Billmeyer-Saltzmans-Principles-Color-Technology-dp-1119367220/dp/1119367220

I will happily go into the rabbit hole of colors! Bring it on!

I only have descendant this color in form of original British Standard 381C chips from this equivalence table: https://www.cybermodeler.com/color/bs381c_table.shtml No clue how close it is considering the time period difference but I can give you a match for that one, just PM me if you need any of BS 381C colors actually.

I love the one in Navy colors. I am so going to paint one in glossy sea blue!... for fun.