MAX: Advanced 3D Aircraft Texture Mapping

[Tango Mike]

(All credits go to the original topic creator: SkyRaider3D)

Which technique is better?

You might wonder now what technique is better. There is no clear answer, it all depends on your model, its purpose, and the time you have to make it. I will try to explain you the benefits and drawbacks of each technique:


Technique 1 - Benefits

* Easy mapping
Unwrapping is kept to an absolute minimum and requires little skill. You just have to be careful with areas that "fold back on itself" such as airscoops. Also, if you have engine nacelles you may have to break them off and map them on a different location in your UV layout. In the case above, you notice that the wing overlaps the rear fuselage. This is wrong of course, and in a later stage the wing was detached in UV1 and moved into one corner so it wouldn't receive the fuselage mapping. But that's about all unwrapping you will have to do. Other than that it's simple planar mapping.

* Smaller texture sizes
Since you're using MAX materials anyway, texture resolution doesn't matter and you can optimise the textures to the shape of your model. This way you can keep each texture relatively small compared to when you would unwrap the entire model on one sheet. Smaller textures are of course much more pleasant to edit in Photoshop.

* No hard seams
By having waved, blurry or speckly edges on your mask, you can prevent seams on your texture really easily. Top/bottom and side views will blend in perfectly and make a neat skin for your mesh. If you keep the length of your top/bottom and side view textures equal, it's also really easy to match up panel lines and such. Also you can generate your right view from your left vuew, and your bottom view will have a lot in common with your top view. So creating your textures is quite a bit easier too.

Technique 1 - Drawbacks

* Lots of textures!
In case of a very complcated model, you can end up with many textures. In case of my Mustang, I wanted a high texture resolution, but not the burden of big textures. So I spit up my model in quite a few (too many!) different textures:
fuselage_top
fuselage_left
fuselage_right
fuselage_bottom
wing_top
wing_bottom
wing_front
tailplane_top
tailplane_bottom
tailplane_front
...and of each one a diffuse, specular and bump variant. As you can understand, the material structure got kinda complicated in the end, with lots of Composite and Mix Maps.
With the example above I was able to keep it much simpler. Top, bottom, left, right (which I never bothered making in the end! ).
Many textures can be a pain when you want to change something, such as the positioning of a marking that goes across the fuselage. This can easily mean you have to edit 12 textures to adjust one item on it.

* Complex material setup
As mentioned above, material structures can get a bit complicated this way. This also means that MAX has a harder time calculating the materials, which can make rendertimes longer.


Technique 2 - Benefits

* Simple material setup
If you manage to map everything onto one page, it's really easy to set up your material. All you need it one Bitmap and no hassle with Mix Maps and Composite Maps. This also improves rendertimes.

* Fewer textures to edit.
If you manage to unwrap everything to one texture, you will have an absolute minimum of textures to edit. Quite convenient when you want to make many variations of your model (different camouflages etc...).

* Stretchless.
You can unwrap your model in such a way, that there's no noticeable stretching. With Technique 1 stretching will be limited too, however, never exceeding 45 degrees if you do it properly.

* Trouble-free export & real-time use
Having unwrapped your model properly means that applications that don't support MAX materials can still display your model properly. So you can export your model to a different application and the materials will be preserved. This will allow your model to be exported to Shockwave 3D, for example, so you can view it in real-time 3D. There are several examples of this on my website.

Technique 2 - Drawbacks

* Tricky & tedious unwrapping
For many the unwrapping process is a complicated one, and for all it's elaborate. It simply takes time to align everything the way you want it. And mistakes are easily made if you lose track of where everything goes. It's also more complicated to edit the texture once it's unwrapped, as all parts have to meet up in the end. This brings us to...

* Lots of seams!
Breaking up your model creates countless seams which have to be cleverly edited on your texture in order to not show up. Here texture baking can be a great help. Imagine having a Tiger Meet F-16 to texture... matching up all the tiger stripes with this technique can be a nightmare. The best option is to make a new, planar, mapping set for this (UV2) and bake it to your unwrapped texture (UV1). But this is a completely different story altogether!

* Big textures/low texture resolution!
For me, the main problem with this technique is the huge texture sizes required to keep a decent resolution for your model - if you unwrap to one single texture that is. Anything over 2048x2048 will really make Photoshop struggle unless you have a super PC. The mentioned resolution is rarely sufficient for mapping an entire aircraft. One solution to this is to scale down unimportant parts and thus free up space for more important parts. For example the bottom of a tailplane will rarely be viewed up-close and can be scaled down, allowing the textures of the nose and cockpit area to be bigger to add detail which will show in close ups. But having parts at different scales also brings trouble: seams will be harder to hide and panel line thickness will vary, which might become noticeable if you're not careful.

Conclusion?

Personally I seem to be using these two techniques in alternating fashion. Whatever I fancy at a given time!
Once I've done a proper unwrap I've become tired of the large texture sizes and poor texture resolution. So I switch back to the first technique. But then I get annoyed by the many different textures I need to edit in order to make changes. Not very pleasant when having to do many different variations of the same texture as well!

However, unless you're modelling for real-time use (dictating technique 2), there is no reason to stick to one technique. In fact, I have found out that combining both techniques works well too. The Me P 1070 uses both mixmaps, but also is unwrapped semi-properly. This way I've managed to keep both texture sizes and quantities to decent levels, without sacrificing quality. At the same time unwrapping was kept to a healthy minimum, speeding up the process dramatically. The best of both, so to say. But it also means the worst of both. The material setup is not suited for export to real-time engines, for example. Additionally I still have to deal with texture seams here and there - tough very few!

The ideal solution? I guess that depends on your personal wishes. If you have a super fast computer that can deal with 10,000x10,000 textures easily in Photoshop, use technique 2 for versatility and ease of texture/material management. But if that's not the case, your time is limited, and real-time use is not required, definitely go for technique 1. For variety's sake, I personally just keep alternating these techniques on a per-project basis!

Well, I hope this "mini-guide" was of some use to those who struggle with texture mapping their aircraft models!

Happy texturing!

Ronnie Olsthoorn

[Tango Mike]

(All credits go to the original reply creator: kevjon)
As I posted this in another thread, it might be useful to post this info here instead. This is the technique I use which I learned from this thread. It is not the only way to uvmap a aircraft but it works for me.

I generally uvmap a high poly aircraft which has 2 iterations of subdivision with planar uvmapping. I break the model down into seperate planar pieces by selecting polygons and then rotate & move those pieces into a square uvlayout as per the attached meteor file. It is important you keep pieces lined up which makes it easier to draw your panel and rivet lines etc across the different uvmappings.

Step1
Before you break any of your model down into seperate pieces and move and rotate them you must set their pivots to world and 0,0,0.

Step2
Once that is done you can break the model into planar pieces by selecting polys and detach and then move and rotate those pieces into a square uvlayout (as per my meteor image below)

Step3
Select all pieces and uvmap with planar option & collapse the stack to bake those uv's in.

Step4
Select rotate tool, Select all pieces, hit F12 to bring up dialog box and set absolute world to 0,0,0.
Select move tool, select all pieces, hit F12 and set absolute world to 0,0,0.
All pieces will be back together again for welding.

Step5
Attach all pieces together into one model and then weld the vertices with a low tolerance value which is dependent on the scale of your model. For me I usually use 0.1.

Step6
Add unwrap modifier to your model and move, scale and arrange the various pieces to in the uveditor to suit your requirements & then export you uvlayout to jpg, tga or tiff collapse the stack.

Step7
Any parts of your uv'd model that you want to move or animate such as ailerons, elevators etc can then be detached by using the element option.


I would use the same technique with a low poly model, the only difference is that you are working with fewer polygons so in uvunwrap you can stitch various parts together to reduce the amount of seams in your model.

I suggest you try this on a few simple shapes to understand it before you commit to a model.

[Tango Mike]

(All credits go to the original topic creator: SkyRaider3D)

Before Max would support UV Unwrap on multiple objects, another way to go about it was to break up your model in planar areas, but detach to elements rather than to parts. The animate them into place. Once all is flattened that way, add a square UV Map, then a Unwrap UVW. Next Cut the Unwrap UVW, delete the UV Map modifier, go back to frame 0 and delete all animation keys, then lastly paste the Unwrap UVW back onto your model and you have the same end result.

The benefit of animating is that you can also manually tweak areas as you go along, such as pulling out leading edges on wings (especially when unwrapped before subdividing or with low-poly models). The downside is that it's less robust. If you want to detach or attach parts that are already animated, the mesh will screw up as you're re-ordering the internal vertex numbering of the model and things will get messy So it takes a little more planning, but it also has some benefits. I use both the animated elements and non-animated objects techniques, depending on how much tweaking I need to do.

Attached an animated sequence from an old model I unwrapped using the animated way. Note that the pieces are not fully planar and I am in fact mixing technique 1 and 2 from my tuturial above. This unwrap is voor de sideview (UV Channel 1, shown on the top-left of the animation) and I did another similar simplistic unwrap for the top/bottom view (UV Channel 2), then blended them with a Mix map.