Casey Posted April 15 Share Posted April 15 (edited) Here is the data I've gathered from British Aviation colours of World War Two book: Keep in mind that it is a book from late 70's, so there is a certain amount of ageing on paint chips. Here are the raw data: I've also calculated and tested recipes for those. The recipes are for Golden Fluid Acrylics and are based on mass parts. Spoiler RAF001 - Light Earth - Flat, Total parts 16, DE00 from target: 0.716895 Raw Umber: 2 Raw Sienna: 9 Titanium White: 5 RAF002 - Dark Earth - Flat, Total parts 15, DE00 from target: 0.688953 Ultramarine Violet: 8 Yellow Ochre: 7 RAF003 - Light Green - Flat, Total parts 23, DE00 from target: 0.803751 Raw Sienna: 18 Anthraquinone Blue: 3 Titanium White: 2 RAF004 - Dark Green - Flat, Total parts 14, DE00 from target: 0.608752 Raw Sienna: 6 Bone Black: 7 Teal: 1 RAF005 - Extra Dark Sea Green - Flat, Total parts 2, DE00 from target: 0.232389 Raw Sienna: 1 Cerulean Blue, Chromium: 1 RAF006 - Grey Green - Flat, Total parts 17, DE00 from target: 0.347447 Raw Sienna: 11 Primary Cyan: 2 Titanium White: 4 RAF007 - Medium Sea Grey - Flat, Total parts 25, DE00 from target: 0.764993 Titanium White: 13 Bone Black: 11 Cobalt Turquoise: 1 RAF008 - Dark Sea Grey - Flat, Total parts 17, DE00 from target: 0.393495 Titanium White: 5 Bone Black: 11 Cobalt Blue: 1 RAF009 - Extra Dark Sea Grey - Flat, Total parts 10, DE00 from target: 0.295876 Titanium White: 2 Bone Black: 7 Cobalt Blue: 1 RAF010 - Ocean Grey - Flat, Total parts 8, DE00 from target: 0.416770 Titan Green Pale: 3 Bone Black: 4 Payne's Gray: 1 RAF011 - Light Slate Grey - Flat, Total parts 17, DE00 from target: 0.477654 Raw Umber: 4 Titanium White: 4 Raw Sienna: 7 Primary Cyan: 2 RAF012 - Dark Slate Grey - Flat, Total parts 16, DE00 from target: 0.651086 Raw Umber: 10 Titanium White: 5 Primary Cyan: 1 RAF013 - Sky Grey - Flat, Total parts 7, DE00 from target: 0.475378 Titanium White: 4 Bone Black: 2 Titan Green Pale: 1 RAF014 - Sky - Flat, Total parts 25, DE00 from target: 0.534561 Titanium White: 5 Titan Green Pale: 17 Chromium Oxide Green: 2 Burnt Umber Light: 1 RAF015 - Deep Sky - Flat, Total parts 15, DE00 from target: 0.720011 Prussian Blue Hue: 6 Anthraquinone Blue: 4 Titan Buff: 5 RAF016 - Sky Blue - Flat, Total parts 22, DE00 from target: 0.221515 Titanium White: 18 Bone Black: 3 Cobalt Turquoise: 1 RAF017 - Azure Blue - Flat, Total parts 18, DE00 from target: 0.343922 Titanium White: 14 Bone Black: 1 Permanent Violet Dark: 1 Prussian Blue Hue: 2 RAF018 - Light Mediteranean Blue - Flat, Total parts 15, DE00 from target: 0.573092 Titanium White: 7 Payne's Gray: 8 RAF019 - Dark Mediterranean Blue - Flat, Total parts 5, DE00 from target: 0.335922 Titanium White: 1 Prussian Blue Hue: 1 Payne's Gray: 2 Ultramarine Blue: 1 RAF020 - P.R.U. Blue - Flat, Total parts 21, DE00 from target: 0.426977 Titanium White: 7 Bone Black: 10 Prussian Blue Hue: 3 Phthalo Green (Yellow Shade): 1 RAF021 - Middle Stone - Flat, Total parts 5, DE00 from target: 0.613003 Yellow Oxide: 3 Bone Black: 1 Titanium White: 1 RAF022 - Night - Flat, Total parts 7, DE00 from target: 0.549596 Bone Black: 6 Raw Umber: 1 RAF023 - Yellow - Flat, Total parts 12, DE00 from target: 1.608428 Raw Sienna: 5 Cadmium Yellow Medium Hue: 4 Titanium White: 2 Diarylide Yellow: 1 RAF024 - Red - Flat, Total parts 7, DE00 from target: 0.427769 Raw Sienna: 4 Burnt Sienna: 2 Cadmium Red Medium Hue: 1 RAF025 - Blue - Flat, Total parts 17, DE00 from target: 0.608347 Ultramarine Violet: 13 Permanent Green Light: 1 Cerulean Blue, Chromium: 3 RAF026 - Aluminium - Gloss, Total parts 10, DE00 from target: 0.954047 Iridescent Pearl (Fine): 9 Bone Black: 1 RAF027 - Matt Red - Flat, Total parts 7, DE00 from target: 0.466359 Cadmium Red Medium Hue: 3 Primary Magenta: 3 Quinacridone Crimson: 1 RAF028 - Matt Blue - Flat, Total parts 15, DE00 from target: 0.403399 Prussian Blue Hue: 5 Ultramarine Violet: 8 Titanium White: 2 RAF029 - Semi Matt Black - Velvet, Total parts 1, DE00 from target: 0.218784 Carbon Black: 1 Update 2022-04-17: Here are the recipes using different selection criteria - it gives more precise matches, but calls for more paints. I've also added examples of how they behave further in this post. Golden Fluid Acrylics - based on Mass parts, closer match but more expensive mixes (more paint parts) to make. I've also added calculations for resulting mixture gloss to give you an estimate of how much flattening agent is needed to make a paint. To give you an starting point: Liquitex Ultra Matte Medium is having GU of 0.2 by itself and works as great paint retarder. Here are amounts needed to bring paint from satin to 3.0 GU as most of the samples above are. 20GU -> 1 : 0.7 paint/matt medium -> 3.0GU 25GU -> 1 : 0.8 30GU -> 1 : 0.85 40GU -> 1 : 0.95 50GU -> 1 : 1.0 60GU -> 1 : 1.1 70GU -> 1 : 1.15 90GU -> 1 : 1.25 Generally, 1:1 ratio is a good approximation to make those paints look flat with this medium. Spoiler RAF001 - Light Earth - Flat Suggested using total of 16 parts (DE00: 0.72) Expected: #AA8D69, Simulated: #AA8E6D Raw Sienna: 9 Titanium White: 5 Raw Umber: 2 Mixture gloss: 30.01 RAF002 - Dark Earth - Flat Suggested using total of 25 parts (DE00: 0.38) Expected: #78634E, Simulated: #796450 Titanium White: 3 Raw Sienna: 7 Bone Black: 7 Burnt Sienna: 6 Benzimidazolone Yellow Medium: 2 Mixture gloss: 27.58 RAF003 - Light Green - Flat Suggested using total of 15 parts (DE00: 0.44) Expected: #656958, Simulated: #656858 Raw Sienna: 8 Teal: 2 Bone Black: 5 Mixture gloss: 34.08 RAF004 - Dark Green - Flat Suggested using total of 13 parts (DE00: 0.36) Expected: #55584C, Simulated: #56584E Raw Sienna: 7 Benzimidazolone Yellow Medium: 1 Carbon Black: 1 Titan Buff: 3 Jenkins Green: 1 Mixture gloss: 35.76 RAF005 - Extra Dark Sea Green - Flat Suggested using total of 15 parts (DE00: 0.22) Expected: #526059, Simulated: #53605A Bone Black: 9 Titan Buff: 3 Jenkins Green: 2 Teal: 1 Mixture gloss: 37.72 RAF006 - Grey Green - Flat Suggested using total of 24 parts (DE00: 0.47) Expected: #738971, Simulated: #73886E Chromium Oxide Green: 12 Titan Buff: 11 Red Oxide: 1 Mixture gloss: 43.78 RAF007 - Medium Sea Grey - Flat Suggested using total of 22 parts (DE00: 0.51) Expected: #889098, Simulated: #888F98 Titanium White: 15 Carbon Black: 1 Titan Buff: 5 Payne's Gray: 1 Mixture gloss: 41.70 RAF008 - Dark Sea Grey - Flat Suggested using total of 22 parts (DE00: 0.48) Expected: #6B7079, Simulated: #6A717A Bone Black: 14 Titanium White: 7 Ultramarine Blue: 1 Mixture gloss: 35.16 RAF009 - Extra Dark Sea Grey - Flat Suggested using total of 18 parts (DE00: 0.28) Expected: #5D626A, Simulated: #5C626A Bone Black: 13 Titanium White: 4 Ultramarine Blue: 1 Mixture gloss: 34.12 RAF010 - Ocean Grey - Flat Suggested using total of 16 parts (DE00: 0.22) Expected: #6A7377, Simulated: #6A7277 Bone Black: 8 Titan Buff: 7 Payne's Gray: 1 Mixture gloss: 33.63 RAF011 - Light Slate Grey - Flat Suggested using total of 21 parts (DE00: 0.36) Expected: #677268, Simulated: #687267 Bone Black: 11 Titan Buff: 8 Benzimidazolone Yellow Medium: 1 Jenkins Green: 1 Mixture gloss: 35.55 RAF012 - Dark Slate Grey - Flat Suggested using total of 28 parts (DE00: 0.56) Expected: #676961, Simulated: #676962 Bone Black: 17 Titanium White: 6 Yellow Ochre: 4 Jenkins Green: 1 Mixture gloss: 38.67 RAF013 - Sky Grey - Flat Suggested using total of 30 parts (DE00: 2.54) Expected: #A6AAAA, Simulated: #9FA2A1 Titanium White: 28 Carbon Black: 1 Yellow Oxide: 1 Mixture gloss: 44.93 RAF014 - Sky - Flat Suggested using total of 15 parts (DE00: 0.50) Expected: #B3BBA3, Simulated: #B3BAA4 Titan Green Pale: 14 Bone Black: 1 Mixture gloss: 50.32 RAF015 - Deep Sky - Flat Suggested using total of 11 parts (DE00: 0.23) Expected: #385379, Simulated: #385378 Prussian Blue Hue: 4 Anthraquinone Blue: 2 Primary Cyan: 2 Titan Buff: 3 Mixture gloss: 49.55 RAF016 - Sky Blue - Flat Suggested using total of 27 parts (DE00: 1.17) Expected: #B7C3C8, Simulated: #B4C4C9 Titanium White: 20 Payne's Gray: 2 Titan Green Pale: 5 Mixture gloss: 43.39 RAF017 - Azure Blue - Flat Suggested using total of 29 parts (DE00: 0.56) Expected: #899DC6, Simulated: #879EC6 Titanium White: 25 Phthalo Blue (Red Shade): 1 Cadmium Red Medium Hue: 2 Anthraquinone Blue: 1 Mixture gloss: 47.30 RAF018 - Light Mediteranean Blue - Flat Suggested using total of 15 parts (DE00: 0.57) Expected: #5F7290, Simulated: #5C7291 Titanium White: 7 Payne's Gray: 8 Mixture gloss: 35.10 RAF019 - Dark Mediterranean Blue - Flat Suggested using total of 14 parts (DE00: 0.36) Expected: #425473, Simulated: #405473 Ultramarine Violet: 1 Primary Cyan: 2 Titanium White: 2 Bone Black: 5 Ultramarine Blue: 4 Mixture gloss: 27.37 RAF020 - P.R.U. Blue - Flat Suggested using total of 18 parts (DE00: 0.50) Expected: #536E7A, Simulated: #526E7C Prussian Blue Hue: 5 Raw Umber: 5 Titanium White: 6 Teal: 2 Mixture gloss: 41.34 RAF021 - Middle Stone - Flat Suggested using total of 30 parts (DE00: 1.74) Expected: #A6864C, Simulated: #A0874B Titan Buff: 3 Permanent Green Light: 1 Yellow Ochre: 26 Mixture gloss: 59.44 RAF022 - Night - Flat Suggested using total of 27 parts (DE00: 0.47) Expected: #3B3C3E, Simulated: #3D3E41 Bone Black: 25 Permanent Violet Dark: 1 Titan Buff: 1 Mixture gloss: 34.08 RAF023 - Yellow - Flat Suggested using total of 14 parts (DE00: 2.00) Expected: #E7A435, Simulated: #DFA334 Raw Sienna: 6 Titanium White: 2 Cadmium Yellow Medium Hue: 5 Diarylide Yellow: 1 Mixture gloss: 45.58 RAF024 - Red - Flat Suggested using total of 8 parts (DE00: 0.28) Expected: #8D4D40, Simulated: #8E4E41 Burnt Sienna: 2 Raw Sienna: 5 Primary Magenta: 1 Mixture gloss: 22.64 RAF025 - Blue - Flat Suggested using total of 12 parts (DE00: 0.30) Expected: #3C475B, Simulated: #3A465A Titan Buff: 2 Payne's Gray: 9 Phthalo Blue (Red Shade): 1 Mixture gloss: 33.35 RAF026 - Aluminium - Metallic - Gloss Suggested using total of 29 parts (DE00: 0.94) Expected: #CFD0D2, Simulated: #CED0D3 Iridescent Pearl (Fine): 26 Bone Black: 3 Mixture gloss: 29.17 RAF027 - Matt Red - Flat Suggested using total of 18 parts (DE00: 0.39) Expected: #A6363C, Simulated: #A7373F Cadmium Red Medium Hue: 12 Titanium White: 1 Quinacridone Crimson: 4 Naphthol Red Medium: 1 Mixture gloss: 66.76 RAF028 - Matt Blue - Flat Suggested using total of 15 parts (DE00: 0.37) Expected: #42517B, Simulated: #40507A Ultramarine Violet: 6 Titan Buff: 3 Ultramarine Blue: 2 Prussian Blue Hue: 4 Mixture gloss: 29.71 RAF029 - Semi Matt Black - Velvet Suggested using total of 1 parts (DE00: 0.22) Expected: #3B3C3D, Simulated: #3A3B3C Carbon Black: 1 Mixture gloss: 79.23 For those who prefer Tamiya, here is my attempt to calculate same on Tamiya paints based on volume parts. Spoiler RAF001 - Light Earth - Flat Suggested using total of 48 parts (DE00: 0.29) XF-7 - Flat Red: 6 XF-4 - Yellow Green: 17 XF-2 - Flat White: 24 XF-1 - Flat Black: 1 RAF002 - Dark Earth - Flat Suggested using total of 10 parts (DE00: 0.34) XF-4 - Yellow Green: 4 XF-10 - Flat Brown: 5 XF-12 - J.N. Grey: 1 RAF003 - Light Green - Flat Suggested using total of 17 parts (DE00: 0.35) XF-13 - J.A. Green: 7 XF-2 - Flat White: 5 XF-4 - Yellow Green: 4 XF-1 - Flat Black: 1 RAF004 - Dark Green - Flat Suggested using total of 12 parts (DE00: 0.32) XF-13 - J.A. Green: 5 XF-2 - Flat White: 1 XF-4 - Yellow Green: 4 XF-1 - Flat Black: 2 RAF005 - Extra Dark Sea Green - Flat Suggested using total of 25 parts (DE00: 0.48) XF-2 - Flat White: 11 XF-11 - J.N. Green: 13 XF-8 - Flat Blue: 1 RAF006 - Grey Green - Flat Suggested using total of 60 parts (DE00: 0.16) XF-5 - Flat Green: 13 XF-8 - Flat Blue: 1 XF-2 - Flat White: 40 XF-13 - J.A. Green: 6 RAF007 - Medium Sea Grey - Flat Suggested using total of 56 parts (DE00: 0.49) XF-12 - J.N. Grey: 33 XF-8 - Flat Blue: 3 XF-7 - Flat Red: 4 XF-2 - Flat White: 16 RAF008 - Dark Sea Grey - Flat Suggested using total of 23 parts (DE00: 0.48) XF-1 - Flat Black: 3 XF-2 - Flat White: 18 XF-8 - Flat Blue: 1 XF-7 - Flat Red: 1 RAF009 - Extra Dark Sea Grey - Flat Suggested using total of 23 parts (DE00: 0.45) XF-2 - Flat White: 16 XF-1 - Flat Black: 4 XF-8 - Flat Blue: 2 XF-7 - Flat Red: 1 RAF010 - Ocean Grey - Flat Suggested using total of 18 parts (DE00: 0.35) XF-1 - Flat Black: 2 XF-8 - Flat Blue: 1 XF-12 - J.N. Grey: 15 RAF011 - Light Slate Grey - Flat Suggested using total of 7 parts (DE00: 0.47) XF-2 - Flat White: 4 XF-11 - J.N. Green: 1 XF-13 - J.A. Green: 2 RAF012 - Dark Slate Grey - Flat Suggested using total of 5 parts (DE00: 0.41) XF-2 - Flat White: 3 XF-1 - Flat Black: 1 XF-4 - Yellow Green: 1 RAF013 - Sky Grey - Flat Suggested using total of 29 parts (DE00: 0.38) XF-2 - Flat White: 25 XF-1 - Flat Black: 1 XF-12 - J.N. Grey: 2 XF-14 - J.A. Grey: 1 RAF014 - Sky - Flat Suggested using total of 50 parts (DE00: 0.44) XF-2 - Flat White: 28 XF-14 - J.A. Grey: 21 XF-5 - Flat Green: 1 RAF015 - Deep Sky - Flat Suggested using total of 27 parts (DE00: 1.96) XF-8 - Flat Blue: 18 XF-2 - Flat White: 8 XF-5 - Flat Green: 1 RAF016 - Sky Blue - Flat Suggested using total of 60 parts (DE00: 1.41) XF-2 - Flat White: 55 XF-8 - Flat Blue: 1 XF-12 - J.N. Grey: 3 XF-10 - Flat Brown: 1 RAF017 - Azure Blue - Flat Suggested using total of 10 parts (DE00: 2.21) XF-2 - Flat White: 9 XF-8 - Flat Blue: 1 RAF018 - Light Mediteranean Blue - Flat Suggested using total of 16 parts (DE00: 2.22) XF-8 - Flat Blue: 3 XF-7 - Flat Red: 2 XF-2 - Flat White: 11 RAF019 - Dark Mediterranean Blue - Flat Suggested using total of 27 parts (DE00: 0.49) XF-8 - Flat Blue: 14 XF-7 - Flat Red: 3 XF-2 - Flat White: 9 XF-5 - Flat Green: 1 RAF020 - P.R.U. Blue - Flat Suggested using total of 10 parts (DE00: 0.39) XF-8 - Flat Blue: 3 XF-3 - Flat Yellow: 1 XF-14 - J.A. Grey: 6 RAF021 - Middle Stone - Flat Suggested using total of 19 parts (DE00: 1.93) XF-4 - Yellow Green: 13 XF-7 - Flat Red: 3 XF-5 - Flat Green: 1 XF-3 - Flat Yellow: 2 RAF022 - Night - Flat Suggested using total of 5 parts (DE00: 0.38) XF-1 - Flat Black: 4 XF-12 - J.N. Grey: 1 RAF023 - Yellow - Flat Suggested using total of 34 parts (DE00: 1.19) XF-3 - Flat Yellow: 31 XF-2 - Flat White: 2 XF-9 - Hull Red: 1 RAF024 - Red - Flat Suggested using total of 13 parts (DE00: 0.48) XF-7 - Flat Red: 9 XF-4 - Yellow Green: 3 XF-5 - Flat Green: 1 RAF025 - Blue - Flat Suggested using total of 4 parts (DE00: 0.41) XF-8 - Flat Blue: 3 XF-10 - Flat Brown: 1 RAF026 - Aluminium - Metallic - Gloss Suggested using total of 26 parts (DE00: 1.98) XF-16 - Flat Aluminum: 25 XF-12 - J.N. Grey: 1 RAF027 - Matt Red - Flat Suggested using total of 1 parts (DE00: 2.11) XF-7 - Flat Red: 1 RAF028 - Matt Blue - Flat Suggested using total of 12 parts (DE00: 0.71) XF-8 - Flat Blue: 7 XF-2 - Flat White: 4 XF-7 - Flat Red: 1 RAF029 - Semi Matt Black - Velvet Suggested using total of 5 parts (DE00: 0.39) XF-1 - Flat Black: 4 XF-14 - J.A. Grey: 1 To give you an overview of what you can realistically expect from doing the mixes, I've prepared them all (well, except pure black), mixed and scanned them with source materials. I tried to create an realistic 'worst case' example - so I've mixed the smallest possible volume of paint to maximize the measurements errors. In each case, the whole batch of mixed paint ended up on a drawdown card, so you can visually see how little of each was mixed (and explain why some drawdowns look more like a blots) Mixing larger batches will reduce the errors and results should be better. Additionally, my scanner does show the difference in gloss ratio, the original paint samples are very matt, the mixed paint is usually satin. This is especially visible on Night color which is almost a perfect match on a spectrophotometer readout but without adding a matt medium or varnish it looks very different. I'm also providing the spectrophotometer readout differences for each sample. I've also cut part of the scanned picture and moved it next to a paint sample to give you chance to closer inspect the samples. RAF001: Measured DE*00=2.37 RAF002: Measured DE*00=3.30 RAF003: Measured DE*00=4.74 RAF004: Measured DE*00=2.34. As an example - AK Interactive AK2011 has a DE*00 of 2.56 in this case. RAF005: Measured DE*00=2.77 RAF006: Measured DE*00=1.31 RAF007: Measured DE*00=1.69 RAF008: Measured DE*00=1.85 RAF009: Measured DE*00=3.15 RAF010: Measured DE*00=1.49 RAF011: Measured DE*00=2.12 RAF012: Measured DE*00=1.80 RAF013: Measured DE*00=1.33 RAF014: Measured DE*00=1.36 RAF015: Measured DE*00=0.49 RAF016: Measured DE*00=1.57 RAF017: Measured DE*00=2.56 RAF018: Measured DE*00=1.95 RAF019: Measured DE*00=2.21 RAF020: Measured DE*00=2.41 RAF021: Measured DE*00=3.34 RAF022: Measured DE*00=1.03 - Note the difference on scan vs the spectrophotometer readouts - my device ignores difference in matt/gloss levels there. And yes I am aware that this paint originally should have Ultramarine Blue added to Black, but this is what I got from my math. I suspect the yellowing of the paint chips may play a role. RAF023: Measured DE*00=1.96 RAF024: Measured DE*00=1.22 RAF025: Measured DE*00=1.88 RAF026: Measured DE*00=1.03 - This is metallic color. Shade of grey is same, but the other properties of those are much more complex to do math on RAF027: Measured DE*00=1.11 RAF028: Measured DE*00=1.51 And last but not least, example of Tamiya paint mixes. I've used precise volume measuring device for those (micropipettor) and the Tamiya paints are flat by design so the results may look like a closer match. They also call for larger amount of parts in total - that allows my color matching algorithm to find better fit. That was a lot of watching-paint-dry fun Update 2022-04-17: Added more expensive but more precise mixes option, and here is an example RAF001: Measured DE*00=1.04 RAF002: Measured DE*00=2.49 RAF003: Measured DE*00=1.06 And there is another comparison when you add Matt medium into the Golden recipes above. Notice how the 'bit too dark' Dark Earth now looks more like the original sample. Edited May 13 by Casey Fixed XF1 and XF2 swap in Tamiya recipes 7 12 Link to comment Share on other sites More sharing options...
Ngantek Posted April 15 Share Posted April 15 This is a wonderful resource, thank you so much for posting! Link to comment Share on other sites More sharing options...
stevehnz Posted April 16 Share Posted April 16 Again, I'm hugely impressed by your efforts here Casey, real world methods to achieve useable results. Steve. Link to comment Share on other sites More sharing options...
Space Ranger Posted April 17 Share Posted April 17 On 4/15/2022 at 5:07 PM, Ngantek said: This is a wonderful resource, thank you so much for posting! Yes, indeed. This thread and its images should be saved somehow so it doesn't go away. Link to comment Share on other sites More sharing options...
Steben Posted April 21 Share Posted April 21 Wooooooow. Good job. Link to comment Share on other sites More sharing options...
Ngantek Posted April 23 Share Posted April 23 Just interested, roughly what kind of algorithm do you use to get these mixes? Presumably with, say, tamiya 1 to 14, there's a wide range of colour overlap. How would it differ just using, say, just red, blue, yellow, black and white? I have no knowledge whatsoever on methodology or colour classification. Link to comment Share on other sites More sharing options...
Steben Posted April 23 Share Posted April 23 (edited) 1 hour ago, Ngantek said: Just interested, roughly what kind of algorithm do you use to get these mixes? Presumably with, say, tamiya 1 to 14, there's a wide range of colour overlap. How would it differ just using, say, just red, blue, yellow, black and white? I have no knowledge whatsoever on methodology or colour classification. Not speaking in name of the topic starter, but with the right basic colours, it is possible. But mostly you get faster to your goal when using colours more suitable to "get there". Edited April 23 by Steben Link to comment Share on other sites More sharing options...
Casey Posted April 23 Author Share Posted April 23 (edited) 7 hours ago, Ngantek said: Just interested, roughly what kind of algorithm do you use to get these mixes? Presumably with, say, tamiya 1 to 14, there's a wide range of colour overlap. How would it differ just using, say, just red, blue, yellow, black and white? I have no knowledge whatsoever on methodology or colour classification. Be very, very careful when you are asking me about math because you may just get an answer I am using Kubelka Munk theory (https://en.wikipedia.org/wiki/Kubelka-Munk_theory) to predict how the pigment and their mixtures react to the light. I found this video to have a good explanation of the underlying theory and how it can be used in practice to solve the (so far mostly unsolved) problem of why the digital painting does not work like real one. Edited April 23 by Casey 2 1 Link to comment Share on other sites More sharing options...
Casey Posted April 23 Author Share Posted April 23 (edited) 6 hours ago, Steben said: Not speaking in name of the topic starter, but with the right basic colours, it is possible. But mostly you get faster to your goal when using colours more suitable to "get there". It is possible if I only used 3 colors and white. Black can be mixed too. In theory, all that is needed is to pick 4 pigments that cover the largest possible color space, lets say Hansa Yellow, Phthalo Blue, Quinacridone Magenta and Titanium White. The problem is that recipes that use those colors will depend on very precise measurements of pigments and even small error will be very noticeable, especially in the colors in the middle of the color space - which is exactly where most of the aviation modeller color range sits. It is more practical to use multiple pigments to have easy and more error resilient recipes. Fun fact: Your home printer uses very precise measurements of Hansa Yellow, Phthalo Blue and Quinacridone Magenta (further called: Yellow, Cyan and Magenta). Printer paper is quite often covered in Titanium White. And black is either mixed or supplied additionally to battle the difficulties of practical color mixing. I wonder why Edited April 23 by Casey 2 1 Link to comment Share on other sites More sharing options...
Steben Posted April 23 Share Posted April 23 Correct Link to comment Share on other sites More sharing options...
Troy Smith Posted April 23 Share Posted April 23 9 hours ago, Ngantek said: How would it differ just using, say, just red, blue, yellow, black and white? as the other answers suggest, theoretically yes. Basic thing taught in primary school. And I remember being a very frustrated 8 year old trying to make 'black' with humbrol red, yellow and dark blue... BUT, you just don't really get pure pigments, and even printer struggle, which is why the better ones use a hex colour system. A better idea can be got from looking at an artist paint mix guide , watercolour use the white of the paper, so here's oil paint basics Titanium White [note this is a blue white] Cadmium Lemon or Cadmium Yellow Light (cool yellow) Cadmium Red Light (warm red) Alizarin Crimson (cool red) Ultramarine Blue (violet blue) Dioxazine Purple (violet) Cerulean Blue or Phthalo Blue (green blue, dark) Viridian Green (cool green, middle value) Black Burnt Umber (warm dark brown) from https://www.virtualartacademy.com/oil-paints-for-beginners/ The Tamiya basic Red, Yellow and Blue are pigments mixes. The sub tints caused by these make things muddy.... I remember one trying to explain the the daughters mother who like wacky hair colours, in this case green, that the reason she was getting a browny green was that the blue was a purple blue and the yellow was a lemony (ie greeny) and the colour mix was correct, the green and red aspect making brown.... though she kept telling me that blue and yellow make green so why didn't it work .... I think after the third explanation of the same basic point I was ended up getting a bit snappy... Anyway, it's also of note that the basic colours referred to are the ones used in 4 colour printing, (which is what most colur printing uses, know as CMYK, Cyan, yellow Magenta and Black, the K stand for key BTW as it's oten used as the key colour) and the rise of the digital screen transmitted as opposed to reflected light) is another factors. As regards camouflage paint and the real world of painting lots and lots of large bits of machinery, it required a VAST amount of paint, @Mike Starmer says that the British army in WW2 needed 8,000 tons of paint a year! The amount needed for the aircraft industries was probably a lot more. I keep quoting this as it was the most useful bit of information I have read about actual military paint, courtesy of @Jamie @ Sovereign Hobbies as modellers we are used to tiny little pots, not dealing with having to make thousands of gallons, where pigments cost and availability become VERY important I'm going to point out some facts about real-life paint manufacture and either the reader will understand and "get it" or will not understand and are in no position to contradict me. 1) Usually camouflage colours are fairly low saturation colours because these blend in better with nature. They're seldom bright and bold. Low saturation colours are normally manufactured by adding coloured pigments to a base made from inexpensive white or white and black pigments. 2) Colour pigments are expensive. The expense varies depending on the specific pigment, but they're expensive. 3) The only way to over-saturate a colour so much is to substantially over-dose your base with the expensive colour pigments. I'm not talking about a few percent more or less - that causes minor differences which you only confirm the presence of with one swatch adjacent to another - I'm talking more in the order of a double dose to get something you obviously look at and think "woah". 4) In the case of colours like dark olive, these are mostly white, black and ochre (which is relatively inexpensive for a colour pigment) sometimes further tinted with a bit of red or green (which are often very expensive). 5) There can certainly be variances in a manufactured paint, but these tend to be greatly overstated, i.e. used as a ready made excuse for all sorts of mistakes. Ultimately, the only way a manufactured paint can end up so oversaturated is to have dumped in a vast amount of the expensive pigments, if not adding in new additional pigments in large quantities not expected in the recipe. Frankly, it's difficult to see how any manufactured paint could end up so drastically off target, particularly in the over-saturated sense, by any business that wasn't actively trying to bankrupt itself by roasting through obscene quantities of pigments like chrome green which were already expensive at the start of the war and in particularly short supply during. 6) I'd venture that most of the "there was a war on, you know" type apologists for such spectacular errors probably don't have any actual experience of what is and isn't possible when mixing different proportions of 2,3 or 4 pigments when 2 of those are usually black and white just to make your base to tint. You simply cannot end up with a Humbrol 30-esque bluish green using only the ingredients to make olive - i.e. you'd actually have to sabotage it by introducing if not blue then an obviously bluish green. Same goes for that bright green Spitfire above - you can't achieve that with black, white, ochre and a touch of red - you'd need to fire in a lot of bright green pigment in to get that saturated on an overly-light base. It would be more tan-like just using the basic olive green ingredients which only turns obviously olive when tinted enough with black. Put another way, with a fixed number of pigments in various ratios you WILL end up somewhere within a certain envelope, and usually when colours like this bright green are discussed it's because it's well outside that envelope. The point of all the above? In essence it's harder to make a credible explanation for how such a colour might have been arrived at in a real-life paint manufacturing environment than it is to demonstrate that someone would have had to go to a lot of trouble to get it so far wrong. That is harder to rationalise than just getting it closer to correct. from https://www.britmodeller.com/forums/index.php?/topic/235078859-accuracy-of-ammo-by-mig-jiménez-raf-wwii-colours/page/3/#elControls_4045174_menu which is an interesting thread in itself. Note the shortage of green pigments in Britain in early WW2 forced the army and navy to stop using green pigments, these being reserved for aircraft, and example, the army used Khaki Green G3 until 1942 when they had to switch to SCC2 brown which was made using an aluminium by product pigment. 3 hours ago, Casey said: Be very, very careful when you are asking me about math because you may just get an answer I am using Kubelka Munk theory (https://en.wikipedia.org/wiki/Kubelka-Munk_theory) to predict how the pigment and their mixtures react to the light. very interesting. I don't think anyone would mind the math, but might just be baffled... though be surprised as we have some very knowledgeable members.. 1 Link to comment Share on other sites More sharing options...
Casey Posted April 23 Author Share Posted April 23 (edited) 1 hour ago, Troy Smith said: Usually camouflage colours are fairly low saturation colours because these blend in better with nature. They're seldom bright and bold. Low saturation colours are normally manufactured by adding coloured pigments to a base made from inexpensive white or white and black pigments. I am actually making making use of that fact - I've prioritized pigments based on their cost so my math first tries to achieve the result with an inexpensive black, white and ochre-class pigments, then proceeds with more expensive ones. This is done only for Golden Fluid calculations since I could assign the 'cost' range based on known pigment composition, Tamiya does not give me any clue about what is in the pot... 1 hour ago, Troy Smith said: I don't think anyone would mind the math, but might just be baffled... though be surprised as we have some very knowledgeable members.. I'll keep it in mind. I'll probably make a separate thread for that. One thing at a time. My current hobby work pile is: Characterization of paints to use for mixing: Golden Fluid Acrylics - done Tamiya - 16 paints done Color reproduction: British Aviation Colours of World War Two - this thread RLM Merrick and Hitchcock - I upated it with improved mixes using the assumption above Humbrol colors reproduction - I foolishly placed measurements and mixes on separate threads. I will probably need to make one thread out of those once I finish with all the Humbrol paints I could find. I am very close to that goal. Official United States Aircraft Colors 1908-1993 Camouflage & Markings of Imperial Japanese Navy Bombers in WWII FS595 - I've (partially as a joke) shared tamiya recipes for those on one thread, but I am working on full set RAL BS 381C:1996 Maybe I should make a vote somewhere? Day has only 24hours for me.... that stubborn rounded Erf spinning all the time... Edited April 23 by Casey 4 Link to comment Share on other sites More sharing options...
Troy Smith Posted April 24 Share Posted April 24 51 minutes ago, Casey said: Tamiya does not give me any clue about what is in the pot... I have seen mention of what in the pot, possibly taken from safety data sheets? If I remember more on this I'll add it in. maybe something done by Nick Millman, he no longer posts here, but you can find him on the aviation of Japan blog, and I think you have a lot of common ground. Cheers T Link to comment Share on other sites More sharing options...
Ngantek Posted May 11 Share Posted May 11 Hi, I'm bumping this with a request that it be pinned please! Thanks! Link to comment Share on other sites More sharing options...
Troy Smith Posted May 11 Share Posted May 11 On 15/04/2022 at 19:14, Casey said: RAF002 - Dark Earth - Flat Suggested using total of 10 parts (DE00: 0.34) XF-4 - Yellow Green: 4 XF-10 - Flat Brown: 5 XF-12 - J.N. Grey: 1 I finally got some of these paint colours. Dark Earth has really bugged me, so this is one I tried first. The above mix gives a red hued brown, even under artificial light this is nothing like Dark Earth. The closest on the chart is Red, it's a lot browner, but still very red, the mix reminds me of the red-brown used on Japanese propellers. Now, I did mention that I found Tamiya to be inconsistent between batches in colours I have, so that may play a part. FWIW, XF-4:2/ XF-10:1/XF-62(Olive Drab):1 looks promising avenue, but will have to check in daylight. Link to comment Share on other sites More sharing options...
Casey Posted May 11 Author Share Posted May 11 35 minutes ago, Troy Smith said: I finally got some of these paint colours. Dark Earth has really bugged me, so this is one I tried first. The above mix gives a red hued brown, even under artificial light this is nothing like Dark Earth. The closest on the chart is Red, it's a lot browner, but still very red, the mix reminds me of the red-brown used on Japanese propellers. Now, I did mention that I found Tamiya to be inconsistent between batches in colours I have, so that may play a part. FWIW, XF-4:2/ XF-10:1/XF-62(Olive Drab):1 looks promising avenue, but will have to check in daylight. Ok, let me mix my batch and let's see what it goes into. If Tamiya is inconsistent I may need to skip doing the Tamiya math... Link to comment Share on other sites More sharing options...
Casey Posted May 12 Author Share Posted May 12 I did whisked the color, and it looks tiny bit different but not that much: BUT the result paint pair is very metameric (https://en.wikipedia.org/wiki/Metamerism_(color)) - you can see it on the spectrum (read: it is a match in certain light conditions only. In my case, D65 - full day). Normally I avoid it by using single pigments to match the curve, but Tamiya has mixes by itself, and does not leave me with much choices there... 2 Link to comment Share on other sites More sharing options...
Troy Smith Posted May 12 Share Posted May 12 14 hours ago, Casey said: and it looks tiny bit different but not that much: Left RAF museum chip, right Tamiya mix XF-4 - Yellow Green: 4, XF-10 - Flat Brown: 5, XF-12 - J.N. Grey: 1. Direct sun. 50620161 by losethekibble, on Flickr The mix is very red under any light, direct sun, indirect daylight, artificial light which are LED bulb. I will try some more of the mixes, If you want I can send brush out of the Tamiya base paint used, to see if this is down to poor Tamiya batch matching. Link to comment Share on other sites More sharing options...
Steben Posted May 12 Share Posted May 12 Still in progress...... These are Revell 86, 88 and 84 Link to comment Share on other sites More sharing options...
Casey Posted May 12 Author Share Posted May 12 This indeed looks too red. And @Steben, I cant see any picture there... Link to comment Share on other sites More sharing options...
Steben Posted May 12 Share Posted May 12 13 minutes ago, Casey said: This indeed looks too red. And @Steben, I cant see any picture there... https://www.facebook.com/photo.php?fbid=276316014284271&set=pb.100057176242792.-2207520000..&type=3 Link to comment Share on other sites More sharing options...
Casey Posted May 12 Author Share Posted May 12 (edited) Now, some math This is the (probably?) the best possible match using first 16 Tamiya paints without going to crazy with paint ratios. Here is what my math returned for such Tamiya mix (you can compare it to the real spectrophotometer results a few posts above) This mixture match is metameric match. I would conclude it is probably impractical to match Dark Earth using only the first 16 Tamiya paints. To compare, I tried same target crazy ratios. It is bit closer but it is still metameric. Real Dark Earth seems to be best matched with mixture of black, sienna and titanium dioxide. Considering they were the cheapest pigments available. Here is an example mix using only those with relatively small ratios. My Golden mix also adds yellow to further fit the book paint sample curve better, but I am not so sure if it was really used in practice. Edited May 13 by Casey 1 Link to comment Share on other sites More sharing options...
Ngantek Posted May 12 Share Posted May 12 Would it be really annoying (or even worthwhile) to try some of the more difficult mixes by maybe starting with an extra colour most adjacent to the final match. So xf52 for dark earth say. So each one would the 'closest match' colour with the algorithm matching using 1 to 14? Link to comment Share on other sites More sharing options...
Troy Smith Posted May 12 Share Posted May 12 2 minutes ago, Ngantek said: So xf52 for dark earth say. I have tried numerous (and I mean 50 + mixes) for Dark Earth. Some have got close, XF-52 is no use, it's got far too much white in it. This is a common problem with Tamiya, they are already reasonably complex pigment mixes. In the case of XF-52, you can't get rid of the white, meaning any mix is too washed out. Some of the best results were with a mix XF-49 khaki, XF-62 olive drab and XF-64 red brown, but, again, the khaki has too much white I'm trying now using XF-4 yellow green to replace XF-49 khaki. I got some interesting results using black, yellow red with a drop of white. I suspect the real pigments are the real paint is black, white and yellow ochre, (the black and ochre gives the greeness) and a little red brown (sienna) What would really be good is a set of tamiya paints using single pigments.... @Casey The Dark Green Tamiya mix looks good though. Link to comment Share on other sites More sharing options...
Casey Posted May 12 Author Share Posted May 12 8 minutes ago, Troy Smith said: What would really be good is a set of tamiya paints using single pigments... So far I found two. Carbon Black and Titanium White... 8 minutes ago, Troy Smith said: I suspect the real pigments are the real paint is black, white and yellow ochre, (the black and ochre gives the greeness) and a little red brown (sienna) I can confirm it from my math simulations above. Burnt sienna and raw sienna are similar family to yellow ochre, here is comparison. 1 Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now