P-47 Progress Thread

MajorMagee said:
The original intention was for r to map out the specular reflection patterns, and the s to determine the shadowing to simulate ambient occlusion. The naming convention is actually only relevant to help identify which is used for what. ...
Click to expand...

Thanks, that clarifies their function a bit - but for me it still doesn't answer why I should choose one over the other. After looking up what ambient occlusion actually does, it seems to me that both files are useful. Unfortunately CFS3 doesn't allow for both files to be used simultaneously...

BTW, the intake duct to the turbo on the P-47 model doesn't darken much when looking along it's length. Would a *_s. map help there?

MajorMagee said:
... The issue with using any sort of texture map to define the lighting qualities of a 3D object that moves through space like an airplane (versus trucks, buildings, trees, etc.) is that they are static, when the lighting is not. Paint in a dark to bright gradient along the top of a curved surface to emphasize the sun falling on it from above using an s file, and it will look wrong when the aircraft banks. Give the surface material a uniform brightness with AnKors shaders and they will automatically adjust the brightness gradient on the surface based on the relative direction of the sun. ...
Click to expand...

This sounds to me an awful lot like how PBR texture maps are set up and what I was referring to in my earlier post; no 'painted on' shadows or highlights/gradients to simulate or emphazise shapes but let the shaders do their work. The difficulty and challenge for me lies in finding correct colours and material values for both the maps and gmax. That's why I have been looking at PBR a lot while doing this - hoping to find/create a translation to the CFS3 world.

MajorMagee said:
... AnKor's shaders also allow us to use normal maps to assist with the dynamic lighting in game. The intention here is to create the illusion of bumps using height difference information that changes the local direction vector of the light bouncing off an otherwise flat polygon surface. Again this is predetermined in a static texture map, so it may give erroneous results as the aircraft moves and turns in space, but if we're dealing with small features like panel lines and rivets the illusion works well enough most of the time.
Click to expand...

Noted - so far parts seems mostly quite flat when lighting is coming from the 'opposite' or 'wrong' direction.

A second question that might be related to this came up: I have parts that should be shadowed by other parts but are lighted as if they receive direct light. What is causing this? I am not sure it is the modeling or not: I have placed 3D bolts in the cockpit (as separate objects placed against the canopy frame object) that should be in the shadow cast by the canopy frame and I have modeled ribs in the wheel wells as one single object (so no separate objects but part of the bay) that both shows this phenomenon...
 
I didn't think we had ambient occlusion in CFS3 (which would be why the intake duct doesn't darken), where did you get that info Andy? If no ambient occlusion is possible, the intake duct may be an appropriate place to paint on some shading, since the times when the sun will be shining into it will usually be very fleeting, and the fact that when looking into a deep narrow hole when the sun is shining brightly on everything else, you really can't see into it.

A second question that might be related to this came up: I have parts that should be shadowed by other parts but are lighted as if they receive direct light. What is causing this? I am not sure it is the modeling or not: I have placed 3D bolts in the cockpit (as separate objects placed against the canopy frame object) that should be in the shadow cast by the canopy frame and I have modeled ribs in the wheel wells as one single object (so no separate objects but part of the bay) that both shows this phenomenon...

This may be a modeling problem, in which case I don't know the answer, but I have also seen this when something is wrong with a bump map. If the format is wrong or the colors are inverted you can get some strange results due to the bump map telling CFS3 that the surface is facing the wrong direction.
 
Going back to the s, and r.dds, I can only speak from an artistic point of view, I don't get into the technical part of it unless I really have too,( I think Andy covered that quite well anyway) but I would include both r.dds and s.dds in your model Joost. One reason, yes this P-47 did mainly use bare metal, but they were also camouflaged too. In past skins of this type I have simply blacked out the r.dds, and used it's s.dds for the none reflective camo.
My main use for the s.dds since Ankors shaders is for reflective adjustment, very subtle panel changes, weathering, and paint chipping. I've found the s.dds also works to make the bump map look a lot better too. I've no idea why, it just does!

Regards bump map textures, I usually use a neutral grey for my backgrounds, with no weathering or chipping, you don't need any of that in there, and don't forget, dark shades raise the bump, light shades lower.

Ankors shaders have changed the way it all works, I no longer paint any major shading onto my skins now.
The modeller has to get their reflectivity percentage just right, if it's too shiny, or too dull, it will effect the final texture quality in a big way. I'm not sure what percentage Steve set the Spitfire we are currently working on, but (for the sim), to me that was right on the money shine wise, gave me nice control via the s.dds. I know that in real life it was probably not as reflective, but we're talking a very old sim here.

Regards your intake, yea I still find the odd small bit that will need some manual shading lol

I've probably not been much help to you, as I've said it's just my artistic view, not technical.

You picked the hardest scheme to get right Joost, a wartime reflective bare metal skin is a pain to get to look natural in this sim, well it is for me!:biggrin-new:
 
gecko said:
... If no ambient occlusion is possible, the intake duct may be an appropriate place to paint on some shading ...
Click to expand...

I was thinking the same thing here

Bravo/4 said:
Going back to the s, and r.dds, I can only speak from an artistic point of view, I don't get into the technical part of it unless I really have too,( I think Andy covered that quite well anyway) but I would include both r.dds and s.dds in your model Joost. One reason, yes this P-47 did mainly use bare metal, but they were also camouflaged too. In past skins of this type I have simply blacked out the r.dds, and used it's s.dds for the none reflective camo. ...
Click to expand...

Okay, I'll have to study that part a bit more and see what effects I can obtain using both those files!

Bravo/4 said:
Regards bump map textures, I usually use a neutral grey for my backgrounds, with no weathering or chipping, you don't need any of that in there, and don't forget, dark shades raise the bump, light shades lower.
Click to expand...

I think we have a similar approach - I inspected some of your *.+nm.dds files: they look quite similar to mine. I fill my background with RGB 128,128,255 (seems to be the background colour in my filter) and 'bump' each object individually to have more control over depth/height or direction of the 'bump' (depending on the object's orientation on the map). In the Normal filter I use in PS it's also possible to give the "bump amount" a negative or positive value to make dents or bumps using the same base colour (though darker colours have a stronger effect indeed)

Bravo/4 said:
... You picked the hardest scheme to get right ...
Click to expand...

Story of my life... :biggrin-new:
 
Here's an example of combining painted and bare metal surfaces.

Make sure the video is playing in HD, but the resolution is still far less than it was in game.


If you watch during the turns the ocean is lightly reflected off the weathered bare metal skin, but not the painted surfaces. Also, the highlights and shadows inside the engine cowling shift as the angles change.

This was done referencing an _s.dds file. As you can see the lightest tones are only at 12-16% brightness to produce a bare aluminum finish (but not museum mirror polished). The painted areas are at about 4-6% brightness.

LYUdAV1.jpg


Hex editing the m3d file must also be done to activate the appropriate drawing mode (adding FF in the 27th of 36 columns) so AnKors shaders can take full advantage of the reflectivity map. This provides better grading/shaping of the highlights on curved surfaces than the standard drawing mode. This is also how the glint on the glass and prop disc is activated.

5TnNUm2.jpg
 
MM,
Great stuff.

Just watched the vid. Very nice the subtle differences between the shine of the metal to the lower sheen of the paint, a good shot is when the reflection moves up the metal sides to the painted anti-glare panel on top.

But even Ankors shaders and your variable shine/sheen application, can't change how bloody awful that original P47 looks...ha ha!!..I can't wait for Frosty's stunning model :jump:

Cheers

Shessi
 
Now here's an interesting tidbit (though the explanation is rather lengthy):

When I started this project, I didn't know how to add a *_r.dds or *_s.dds map to the model as simply copying one of those files into the CFS3 folder of the model clearly didn't work (I was hoping it would be picked up automatically by the sim, as *.+nm.dds files are). Then I remembered hexediting .m3d files to create standalones, so the reference to those files was in the .m3d file, which in turn meant that they should be added in gmax. Looking through gmax I found a series of map options in the material editor. Adding my *_r map to the material did work and the texture was picked up ingame. But adding the *_s.dss to this material resulted in an error when I exported the P-47 model, telling me that the '4th channel couldn't be used and this texture would be ignored'. That seemed consistent with the SDK that stated that having both files simultaneously wasn't possible and with the stock P-47 that had both textures in its folder but with just one reference in its .m3d file. But that was on my laptop...

When I tried the same thing on my desktop, I was able to export a .m3d with both files added to the model without the error message. I checked the .m3d with a hexeditor and both references are there indeed. Weird, isn't it? Maybe it has something to do with the gmax settings or other alterations I made to the two installations?

Now I haven't tried what the model looks like ingame as I created my *_s file quickly by copying the *_r.dds and renaming it to *_s.dds (so they are basically the same - but so do the stock P-47 files appear to me). Also, gmax quickly runs out of memory when adding the *_s file to all materials - having three 4096 files per material (*_t, *_r and *_s) clearly is too much but I think that could be resolved by replacing the *_r and *_s files with small placeholder files for export purposes, as they are not practically used in gmax anyway. But it also has me a bit confused because if this works it would appear I now have four variables to control to get the looks quite right: The gmax material settings and the *_t, *_r and *_s. texture maps!

To be continued?! (Hopefully...)

Oh, and I could be going about this completely the wrong way but then I'm sure someone will correct me :untroubled:
 

Attachments

  • grab-000.jpg
    grab-000.jpg
    37 KB · Views: 419
Well there's only one way...suck it and see.....;)

If MM doesn't mind me saying, and up for it, I'd send it to him to assess.


Shessi
 
There is also the possibility of doing a separate version for skins that use full paint, either in gmax or with a hex editor later, though I suppose the ideal is to not have to do that.
 
OK guys,

Right now I am doing three things simultaneously: a little (re)modeling, unwrapping and playing with textures and material settings to keep things both interesting and in motion. I managed to create some short clips showing where I am with the model:

https://vimeo.com/493129138

https://vimeo.com/493129137

https://vimeo.com/493128990

Comments are always welcome but be prepared that you probably will get a "It's already on my list" or a "It's a WIP" -reply... :untroubled:
 
Small update:

I now have properly unwrapped nearly every part of the external model and managed to place them on a single texture sheet, so that means that I can start painting the exterior in full earnest - what you have seen so far was mostly try-out and me being impatient / wanting to see what things would look like ingame. It also means that there's going to be a paintkit!


sixstrings5859 said:
... Can't wait to see this one in action !...
Click to expand...

Just for fun, a 30s. video of the flaps being deployed in gmax. In the real aircraft, the maganta parts are linked to a torque tube that runs the span of the flap - see posts #59 & #72 of this thread. There are slots in the trailing edge of the wing for the green arms. These are already cut but I can't decide whether the slots were covered by spring-loaded covers (such as the one nearest to the fuselage) or left open. Museum and flying examples seem to have them seldom installed. The flap deployment starts about 3-4s. in and is repeated once:

https://vimeo.com/501810036
 

Attachments

  • Teaser32.jpg
    Teaser32.jpg
    123.5 KB · Views: 317
Frosty said:
Small update:

I now have properly unwrapped nearly every part of the external model and managed to place them on a single texture sheet...
Click to expand...

Tidy work, very tidy work; there's little space wasted there! :applause: I approve of the fuselage unwrap, they're a bear to get right, ain't they?
 
You must log in or register to reply here.
Back
Top