Geezer

On the way.

I have a CL.IV that I will never finish. I will cheerfully donate it - no strings attached - to anyone who is willing to finish it.

I made the same mistake!

After some unfortunate delays, the narrow-shouldered pilot is progressing well. The face, helmet, and scarf are pretty much done and have stared work on the leather jacket.

Sent you the pilot this morning - did you get it?

Heh, heh - had trouble typing last night. That should have been Halberstadt CL.IV. The improved two-seat fighter design that replaced the CL.II in the last months of the Great War.

I have Windsock data files for the Halberstadt C.V and Albatross C.1. Also have equally good drawings for LVG C.VI. Will send scans if you are interested.

While waiting for my main computer to get fixed, I experimented with the hi-res Fokker textures I am working on by running them in large dogfights on my gaming computer. I ran 512 and 1024 versions for comparison and, while the 512s DID run faster, the 1024 skins did not cause frame-rates to fall below 35fps.

I had a major computer problem, but I'm back up and working on the new pilot. Does he need a dueling scar?

I would be happy to help. I have several figures that need to have recently learned lessons applied, so while doing that I could custom-tailor a figure to fit in the Starstrutter. Send me a PM. :yes:

There are two possible causes for the shadows on your model: 1) You need to apply a smoothing group to your polys - this usually makes shadows go away. 2) The polys change angle too abruptly. See the angles of the red polys shown below. There is a shadow on the side of the engine because the polys change angle quickly. There is no shadow on the wing polys because they change angle slowly.

Hinch asked me how my Gladiator was doing. The answer is that I set it aside to work on other stuff, but I could resume work on it after I finish 'dem Fokkers'.

I'm working on new Fokker D.7 skins, and some adjustments that I had to make to Edward's excellent Flanders terrain caused me to think about producing an article about color systems for the knowledge base. First, let me emphasize that there is no such thing as a "bad" color system. The many different color systems found in TW mods are all "good" - but they're all different. Major differences in colors can result in an unrealistic appearance in mods that are attempting to simulate reality. Some guys notice the visual mismatch that can sometimes occur when using a variety of terrain tiles, ground objects, aircraft, etc and some guys don't. Research has found that anyone with normal color vision does perceive "unrealistic" color, but may not consciously realize it - kind of a "you won't notice it if you don't know about it" situation. To illustrate what I mean, below are shots from testing the Fokker WIP and from a new WW1 sim in development. First, why did I modify the Flanders' colors? All color is relative. This is a fancy way of saying that the appearance of a color is influenced by the other colors around it. That's why brick red looks red next to brown, but looks brown when next to red. The way that a group of colors interact with each other creates a kind of group effect that can be called a system. Put another way, the colors all work together to produce a specific visual effect. Adding colors that do not fit the system creates a kind of visual conflict that happens frequently when guys combine different mods together. The greens in the Flanders color system appeared too bluish on my monitor, and this created a mis-match with the cammo colors of the Fokker that made it hard to develop the new skins. To eliminate the conflict of color, I altered the Flanders .bmp artwork - but I forgot to alter the airfield .bmp. First shot below shows how the airfield tile is too bluish to match the new over-all color system created by the Fokker skins and the altered Flanders' tiles. It is not uncommon for computer game graphics to have unrealistic colors for natural scenery - most natural colors are fairly dull, so developers hype the colors for a more dramatic effect. The next shot is a SPAD from a new WW1 sim in development. While the lighting is a bit more subdued when compared to the Fokker shot, the color systems are quite similar. Shot below shows the bluish green from the airfield pasted over the SPAD. The same visual mis-match can be seen - the bluish-green does not match the color system. The next shot shows the SPAD shot reduced and pasted onto the Fokker shot. The SPAD shot was lightened SLIGHTLY to match the high noon lighting of the Fokker shot - no other adjustments of any kind were made. As you can see, both color systems pretty much match up. What I propose to do is produce some guidelines written for modders - not artists - to assist them in understanding color systems. But...it will be a lot of work. How many guys would be interested? Snippet from first draft: INTRODUCTION: The following guidelines are intended to assist those modders who are interested in making their artwork look more realistic. The guidelines are based on the fundamental principles of realistic painting that were first developed by the "Great Masters" – Da Vinci, Michelangelo, Rembrandt, etc. These basic principles have not been significantly altered by the development of the scientific method, although science can now explain the underlying reasons for these 400-year-old techniques. Think of a cake, and then think of its icing - if the cake ain't right, it does not matter how good the icing is. 90% of creating realistic artwork is getting the colors right - think of color selection as making the cake. All the artsy stuff - shading, blending, etc - is the icing. Most modders either don't have the time or the ability to make the icing, but they CAN learn to make the cake. If they do, they have accomplished 90% of what they want to do. 90% ain't bad. Recognizing that the average modder is not interested in spending many years becoming an accomplished artist, I have greatly simplified the basic principles in an effort to get modders up to speed in a reasonable amount of time. The following guidelines are roughly divided into two parts: background and practical application – without an understanding of the background, it will be next-to-impossible to master the practical application. THEORY: First, some simplified physics. 1) Color is light. Our eyeballs are biological light detectors. 2) The different colors that we see are the result of light striking our eyeballs at different frequencies. When we look at an object, what we see is the result of light bouncing off that object - the reflected light travels through the air and strikes our eyeballs. The real world usually changes that light as it travels from the object to our eyeballs. This is crucial to understanding color – the colors that we see are not always the colors that first bounce off the object. This is why carefully matching a color chip does not always produce game textures that "look real". For you techno-types, this is the same frequency shift that can occur with most any kind of radiated energy. Let's call these color changes "optical illusions". Learning how to make realistic artwork is basically learning how to simulate the optical illusions that occur in the real world. Learn this and you will have learned how to make the cake. TERMINOLOGY: A color has three components: - Chroma. This is a fancy word for the color itself. Is it green, or blue, or red, etc. - Value. This is the lightness/darkness of a color. Think of what shade of gray the color would be in a black and white photograph. A light gray in the photo would be a light value. - Saturation. This is the vividness or intensity of the color. Think of fire engine red, and then think of brick red. Fire engine red is highly saturated (very vivid or intense). Brick red is a low saturation color (not vivid or intense). Tint: Adding a very small amount of one color to SLIGHTLY change another color. Atmospheric perspective: The atmosphere is not 100% transparent – water vapor, dust, etc in the air is the primary cause of color change. Atmospheric perspective will cause colors to appear lighter in value and lower in saturation. Distance between the object and your eyeball is the main variable – the more air the reflected light must pass through, the more the color will change. Scale Effect: The same color applied to two different objects can appear different if there is a significant difference in size of the objects. Generally, the larger the object, the less saturated the color appears to be. The bottom line: Simulating real world changes to a color is all about making the color lighter ( in value) and duller (lower in saturation).

Two shots below demonstrate how the use of extremely dark grays instead of black will produce a more realistic looking result. First shot shows "black" and "white" cross on upper wing of Fokker. They are actually extremely dark or light grays - note true black line drawn through center of cross in second shot. True black should be used very sparingly to simulate interior spaces that are not struck by sunlight - compare to black cockpit interior. ">

I Googled Superbug Navy Hornet, but couldn't find any shots that show unusual colors. If you are having problems, could you point me towards an example? When lifting colors from digital images, I often blur a small patch of the color in question, and then use the eyedropper tool to get a sample. This does not always work well, but it will produce a kind of "average" color sample. I'm still learning the eccentricities of the TW game engine which will (like any other game system) alter colors in-game. Also, digital imagery can sometimes be misleading for a whole bunch of technical reasons. Example below shows Fokker lozenge sample DL-ed from the internet - the flesh-colored lozenges don't "match" the rest of the colors. I altered them for two reasons: - Common sense. Cammo colors are supposed to hide an aircraft, not make it stand out. - Real world observation. I've examined several Fokkers in US and German museums, and the over-all impression was of dull, muted colors without any bright, vivid "spots". If you examine the other four colors in the sample, you can see how they were slightly lightened (value) and dulled (saturation) as described in an earlier part of this thread. Sometimes, a kind of visual mis-match can occur when using a combination of dull cammo colors and vivid squadron markings. Is that what is happening with your Superbug skins?

Edited and expanded draft: INTRODUCTION: The following guidelines are intended to assist those modders who are interested in making their artwork look more realistic. The guidelines are based on the fundamental principles of realistic painting that were first developed by the "Great Masters" – Da Vinci, Michelangelo, Rembrandt, etc. These basic principles have not been significantly altered by the development of the scientific method, although science can now explain the underlying reasons for these 400-year-old techniques. Think of a cake, and then think of its icing - if the cake ain't right, it does not matter how good the icing is. 90% of creating realistic artwork is getting the colors right - think of color selection as making the cake. All the artsy stuff - shading, blending, etc - is the icing. Most modders either don't have the time or the ability to make the icing, but they CAN learn to make the cake. If they do, they have accomplished 90% of what they want. 90% ain't bad. Recognizing that the average modder is not interested in spending many years becoming an accomplished artist, I have greatly simplified the basic principles in an effort to get modders up to speed in a reasonable amount of time. The following guidelines are roughly divided into two parts: background and practical application – without an understanding of the background, it will be next-to-impossible to master the practical application. THEORY: First, some simplified physics. 1) Color is light. Our eyeballs are biological light detectors. 2) The different colors that we see are the result of light striking our eyeballs at different frequencies. When we look at an object, what we see is the result of light bouncing off that object - the reflected light travels through the air and strikes our eyeballs. The real world usually changes that light as it travels from the object to our eyeballs. This is crucial to understanding color – the colors that we see are not always the colors that first bounce off the object. This is why carefully matching a color chip does not always produce game textures that "look real". For you techno-types, this is the same frequency shift that can occur with most any kind of radiated energy. Let's call these color changes "optical illusions". Learning how to make realistic artwork is basically learning how to simulate the optical illusions that occur in the real world. Learn this and you will have learned how to make the cake. TERMINOLOGY: A color has three components: - Hue. This is a fancy word for the color itself. Is it green, or blue, or red, etc. - Value. This is the lightness/darkness of a color. Think of what shade of gray the color would be in a black and white photograph. A light gray in the photo would be a light value. - Saturation. This is the vividness or intensity of the color. Think of fire engine red, and then think of brick red. Fire engine red is highly saturated (very vivid or intense). Brick red is a low saturation color (not vivid or intense). Tint: Adding a very small amount of one color to SLIGHTLY change another color. Atmospheric perspective: The atmosphere is not 100% transparent – water vapor, dust, etc in the air is the primary cause of color change. Atmospheric perspective will cause colors to appear lighter in value and lower in saturation. Distance between the object and your eyeball is the main variable – the more air the reflected light must pass through, the more the color will change. Scale Effect: The same color applied to two different objects can appear different if there is a significant difference in size of the objects. Generally, the larger the object, the less saturated the color appears to be. Simulating real world changes to a color is all about making the color lighter (in value) and duller (lower in saturation). THE SAME ONLY DIFFERENT: All color is relative. This is fancy way of saying that a color's appearance is affected by the other colors around it. Brick red appears red when placed next to brown, but appears brown when placed next to red. Adjusting the way a group of colors interact with each other is probably the second most important thing to learn (after lightening and dulling the colors). Think of the faces of the brothers and sisters in a family. There is a thread of family resemblance running through those faces even though each face is unique – they are "the same, only different". A color system can have a wide variety of colors but they will have a thread of similarity running through them – a kind of family resemblance. The thread can be any one of the three color components: hue, value, or saturation. Really complex color systems can have a combination of components, but let's keep things simple for now. ANALYSIS: Figure 1 shows an example of colors that lack family resemblance – the airfield - when compared to the surrounding colors. FIGURE 1 Notice the wide variety of colors, but they all seem to be part of a piece – there is a thread of similarity running through all of them (see THE SAME ONLY DIFFERENT). Except for the airfield above the wing – it's "too different". To analyze why the airfield seems too different, think about the three components of color: hue, value, and saturation (see TERMINOLOGY). While looking at the image, squint your eyes until they are almost shut. The green portion of the airfield partially fades into the surrounding colors, while the beige portion of the airfield (bare earth) remains distinctive. Squinting your eyes is a good way to get into the ballpark quickly. Minor differences in hue, value, and saturation tend to fade away. However, major differences that are "too different" do not fade away as much. 1) Hue: Both the beige portion and the green portion of the airfield do not fade away as much as the surrounding colors. The beige portion really stands out because - except for a sprinkling of houses - it is the only example of that specific color. You have just identified a mis-match in hue. Now look at the green portion of the airfield. While less pronounced than the beige mis-match, there is a difference between the airfield green and the surrounding green colors. You have just identified another mis-match in hue. 2) Value: The beige portion of the airfield does not fade away as much as the surrounding colors because is lighter in value. As the bare earth visible elsewhere is not as light as the airfield patch of bare earth, you have just identified a mis-match in value. Now look at the green portion of the airfield. The value is not wildly different from the surrounding colors, so there is no mis-match in value. 3) Saturation: Both the beige portion and the green portion of the airfield do not fade away as much as the surrounding colors. Ask yourself: do they stand out because they are more, or less, vivid than the surrounding colors? Remember, you are looking for the difference between brick red and fire engine red. They are not significantly different than the surrounding colors, so there is no major mis-match in saturation. Bottom line: the airfield stands out from its surroundings because of differences in hue and value, but not saturation. HUE: The beige mis-match is pretty obvious, but the green mis-match is more subtle and thus harder to understand, so let's talk about greens. Remember "tint"? (see TERMINOLOGY). To over-simplify a bit, there are three different tints to green: - Red-ish tint: Think of forests as found in temperate climates. - Blue-ish tint: Think of jungle as found in tropical climates. - Yellow-ish tint: Think of early Spring grass as found in temperate climates. Figure 2 shows the three different tints: left is red-ish, center is blue-ish, and right is yellow-ish. Figure 2 Note that in each example, there is a wide variety of greens but they all share the same family resemblance (see THE SAME ONLY DIFFERENT). All colors can be lumped together into families, though the actual tints will vary. To "look realistic", all colors should be part of the same family. VALUE: The following will blow your mind. Figure 3 is an IDENTICAL COPY of Figure 2 except that it has been de-saturated (converted to gray) to show value (see TERMINOLOGY). FIGURE 3 Figure 3 illustrates how deceptive some colors – particularly yellow – can be when analyzing a screenshot. Converting a screenshot to gray (de-saturation) is a good way to identify a mis-match in value. By converting your screenshot to gray, you can avoid being deceived by color. More deception – black and white are NOT colors. They are values. Remember, color is light (see THEORY). True black represents the total absence of any light. True black is seldom found in nature, except inside enclosures that block all light from entering. Also, true black and true white are absolutes – none of this stuff where colors are affected by surrounding colors. Once again – true black is the total absence of any light, and does not usually occur in nature. What you see when you look at "black" markings on a real aircraft are actually EXTREMELY DARK GRAY markings. Conversely, "white" markings are actually EXTREMELY LIGHT GRAY – except in areas of greatest reflectance. http://www.webopedia.com/DidYouKnow/Comput.../2002/Color.asp Let's call this another optical illusion. Figure 4 reveals the optical illusion by showing true black and true white stripes drawn across "black" and "white" markings. FIGURE 4 This is why Classical Painters (capital letters intentional) don't use true black or true white as general colors, such as "black" or "white" markings on an airplane. Your terrain tiles, skins, or decals will not "look real" if they use true black or true white. Use extremely dark gray ("off black") or extremely light gray ("off white") instead.

Nice colors.

Thanks for your comments guys. It initially looked like no one was very interested in the subject, but the spate of recent comments tells me I should get back to work. Another factor in realistic appearance is lighting - the amount of sunlight flooding a landscape can also affect how we see colors. This gets us into areas that TW sims can't simulate - yet. All we can do is be aware of the effect and try and balance colors for the best compromise possible. Shot below is a developer's shot for a popular shooter - it shows the photo the devs worked from, and the game level the devs made that was based on the photo. The "real" photo shows color variation in value (lightness/darkness) and saturation (vividness/dullness) due to some parts of the landscape being in shadow, and some parts being in direct sunlight. The game level does not attempt to simulate this extensive color variation, and pretty much shows a uniform amount of sunlight everywhere. This is an old dev shot that does not depict what was eventually released - the shooter DID simulate this kind of color variation when it hit the store shelves. Unfortunately, TW modders can't simulate this variation in lighting without sometimes getting some weird results. This is a judgement call - that's why they call it "art" - but I suspect that a uniform system of colors would look the most realistic, most of the time. In plain English, without the ability to simulate the light variation shown in the photo, TW game levels should probably have the uniform look shown in the game level shot. Any comments?

Started fooling around with the cockpit and pilot again.

It'll be a while before this is ready, as I will also make 512 versions for guys with older rigs. I also plan on releasing the original 1024 Photoshop art with color swatches so guys can make squadron markings and personal insignias. Then, I hope to expand the 1930s ground objects by adding more WW1 stuff. It will then be the "Golden Age Ground Objects" that will cover the era of the biplane, from 1916 to 1940.

More progress shots.

That's fascinating stuff that I did not know about - the Albatros engineers kind of backed into a workable solution. It probably didn't "fix" the problem but delayed the onset of the torsional flutter to a speed that was higher than normally encountered in combat. The problems with installation of the field mod kits are also very interesting, and illustrate the vast gulf between guys working in an orderly factory environment, and what happens in front line units experiencing the pressures of daily operations. At work, I am one of the few guys who has actually served a hitch in aircraft maintenance so we often have heated discussions when I attempt to introduce a reality check. Once again, I cheerfully admit this info may be of little use, but I am glad that you appreciate it. In the article there was also a finite model of the SPAD 7/13 wing cellule for comparison, and the SPAD's wingtips did not fail due to torsional flutter until a theoretical speed of over 300 mph. As terminal velocity of a SPAD would be less than 300 mph, the SPAD was - for all practical purposes - immune to flutter. The article also singled out the Pfalz D.3 for having similar qualities - an extremely strong aircraft that could dive at high speeds for prolonged periods of time. Thanks for your comments - now it's back to the Fokker textures.

That kind of distortion has often driven me crazy, and I am amazed that you were able to overcome the problems. VERY nice work!

Thanks for responding. I really don't have a position one way or another, I simply thought that you guys might find this information of some interest. FYI, the techniques employed by the author - an aeronautical engineer - for modeling aero elasticity have only been in use since the 1960s, as the phenomenon is extremely complex and requires extensive computer horsepower to model accurately. The finite element model was constructed using industry standard techniques that employ MSC NASTRAN, a program originally developed by NASA, and referenced semi-empirical flutter analysis by the well known French engineer Professor Y Rocard. The original static load factor of the Albatros D.5's lower wing was only 1.2, and Idflieg requested an increase in strength to 1.7 which was confirmed in tests conducted in April 1918. Additional modifications improved the static load factor further to 2.5, but this was still inadequate when compared to the 5.3 static load factor of the Albatros D.1 of 1916. The problem however had nothing to do with the inherent strength of the sesquiplane wing spars or attachment fittings. Idflieg torsion tests on the Albatros D.5 were conducted in May of 1918 and revealed that the plywood lower wing spar could twist as much as 79 degrees without splintering. Flutter of a hinged control surface is different from the aero elastic deformation of airframe structure - the structural engineers I work with often get sloppy and use the term interchangably and I confess I made the same mistake when I summarized the article. It would be better to describe the problem as vibration. The problem with the Nieuport and Albatros lower wings was that they were single spar designs, which have a natural tendency to twist around the horizontal axis that extends along the length of the spar itself. The small auxiliary strut introduced on the Albatros D.5a was an attempt to damp this twisting motion caused by inflight vibration. Unfortunately, because vibration-caused deformation was NOT understood in WW1, the auxiliary strut had no effect whatsoever. The problem was that the lower wing's center of gravity for both the Nieuport and Albatross was located aft of the spar's horizontal twist axis. In this case, aerodynamic loads interact with airframe vibrations and cause the lower wing to both twist and flex in flight - the faster the aircraft's speed, the faster the wing will twist and flex until it fails. The shot of the finite model demonstrates this fatal twist and flex that is caused by speed and vibration interacting to produce dynamic loads that no structure can withstand - the wing breaks off DOWNWARDS. You mentioned that some WW1 aircraft could be dived faster than 150 mph and you are absolutely correct. All aircraft experience vibration-induced aero elastic deformation, but the twisting and flexing is generally minor thoughout the aircraft's normal speed range - the deformation only becomes a problem at speeds greater than what the aircraft usually flies. As you correctly pointed out, the drag of most WW1 aircraft prevented them from reaching this critical speed range. Most, but not all. Some WW1 aircraft designs had inherent aero elastic flaws that were not apparent until the aircraft was dived at higher-than-usual speeds. This was the starting point for the generalization about 150 mph - the author stated in the article that this estimate was a ball-park guess because each aircraft would have to be extensively analyzed and modeled to determine it's unique critical speed. Because the author had done precisely that with the Albatros structure, he could accurately predict failure somewhere between 118 and 135 mph. So...my statement that all aircraft would experience these problems above 150 mph was correct, but I should have added that it is not clear HOW MUCH above that figure. I am not trying to create a controversy here, and I readily admit that my summary of the article may be of no real use to you guys. I simply hoped that you would be interested in an analysis of WW1 structures by a modern aircraft engineer (the author's credentials: Staff Engineer, Lockheed Martin, 1994-1995. Structural Engineer, Space System Division, General Electric Company, 1985-1994. Structural Engineer, RCA, 1983-1985. Product Research Engineer, The Budd Company, 1978-1982). Interestingly, the fix for the single spar lower wing would have been cheap and easy, had it been understood in WW1. By simply attaching a few pounds of lead weights under the leading edge, the wing's CG would have been shifted forward AHEAD of the wing's twist axis, and the problem would have been eliminated completely. I love what you guys are trying to do with FMs, and I wish you guys success.

I have followed this thread with great interest and finally found an old Air Enthusiast article that might be a LITTLE help. It might also muddy the waters too! The article was an engineering analysis of WW1 single spar sesquiplane designs, with particular emphasis on the Albatross D.5. The article examined the dynamic loads imposed on Nieuport and Albatros aircraft and concluded that they were little better than death-traps in anything more than gentle dives. The author constructed a finite element model of the Albatros D.5 wing structure to illustrate the problem of lower wing flutter that was inherent in both designs. Note that flutter was not even recognized as a problem in WW1, let alone understood by engineers of that time. Basically, the single spar sesquiplane wings failed due to SPEED, not G-loads. Acknowledging that his finite model was a bit of a WAG, the author estimated the failure speed to be anything from 118 mph to 135 mph for the Albatross D.5, and also stated that the Nieuport would have been roughly the same. The wings would flutter and literally twist off the airframe before exceeding their load factors (shot of finite element model below). In one sense this only makes your efforts harder, but in another it may help you. This modern analysis confirms the conclusions you have come to about diving speeds, based on your reading. Very few WW1 aircraft could be dived at high speed - dynamic loads such as flutter would literally tear them apart before reaching their theoretical maximum load factor. Two exceptions specifically mentioned in the article were the SPAD 7/13 and Pfalz D.3 which would only develop flutter problems when their speed approached 300 mph. Even the aircraft that were considered structurally strong in WW1 (Fokker D.7, SE.5, etc) would experience flutter problems above 150 mph. I don't know how much help this is, but I thought you might appreciate the information.