The YF-19: Mapping update #2

September 1 - Mapping of the hidden geometry

September 1 – Mapping of the hidden geometry

There were a few logistic things I needed to figure out here. My whole model is mirrored, that is, only the left half exists, and the other one is a mirror projection. That’s fine and good, but when mapping it, there are parts of it like the nose cone, the upper cockpit or the wings that I don’t want to be a mirrored thing, especially since there are going to be so many labels and warning signs and all that which I don’t want to show as mirrored. But there are things that I don’t mind are the same on both halves, like the inside of the fingers, internal mechanisms and borders that are just one colour or have no details that look wrong when mirrored. If I had those ones not mirrored that would be twice as much work, and also twice as much space spent, and as far as I’ve learned from UV Mapping, you need to use space as wisely as you can.

So the way I ended up solving this dilemma was: To map everything when still mirrored, then packing the parts inside the UV square, so when applying the mirror transform to the model so it’s a full object those maps remain in the same place and I don’t have to manually move and mirror their axis and put them on each other. Then, I’ll select only the parts that are not mirrored and apply the UV unwrapping again. Finally, when I have all the parts not mirrored, I’ll pack them on the square and simply fill in the empty spaces with all the unmirrored (and smaller, sometimes) parts, without moving them from their mirrored versions. It’s the approach of filling a bottle with pebbles and sand. You don’t start with the sand because then there’s no space left for the pebbles; you start with putting the pebbles in (the bigger parts) and then pour the sand in. The sand will fill in all the empty spaces efficiently. So I came to that conclusion and started working.

September 2 - Hand mapping

September 2 – Hand mapping

There are parts of the mecha that are not visible when in plane mode. They belonged to a different object, which I had added to a driver that turns it visibility and renderability when you begin the transformation. That way, when it’s in plane mode those geos are not visible and Cycles won’t render them, but the moment you start transforming the rig, they become visible automatically. Same goes to the landing gear.

Well, in order to map this efficiently I had to track back from that. Everything had to be in one single object so the mapping could be done as efficiently as possible. So I had to leave a few polygons from the object with the constraint, so I can revert back to it, save the polygons as vertex groups, then join them to the main body so I can project the UV map of everything in the square.

September 3 - More hidden geometry

September 3 – More hidden geometry

September 4 - Testing the map on transformed model

September 4 – Testing the map on transformed model

After a while, it became apparent I wasn’t going to fit in EVERYTHING in one single square map. I would need to divide the map by parts, or materials. For example, the landing gear would get its own map. Then, the fuselage with all the things I wanted as big and detailed as possible like the wings and legs, the cockpit inside would get another one, and finally, everything mechanical, small and usually hidden.

This brings another advantage: If I manage to make future paint schemes in the future, I think it’s wise to just have one map with the things that change colour, that is, the fuselage. The mechanical parts, cockpit and those small gray and metallic bits will always remain the same. The landing gear won’t probably change from the blue-white version of the first Alpha One testing plane to the tan and red one from Macross Plus. So I’m essentially thinking about the future iterations when deciding how to map this.

Also, the renders were looking a bit washed out. Why was this? I made the wrong material, I had a mixed shader between the uv map in diffuse, and a white glossy material. Now both glossy and diffuse have the same texture. Boom.

150904-3

September 5 - Plane mapping finished

September 5 – Plane mapping finished

This week was slow and boring, I don’t think I’ve enjoyed mapping as much as I thought. I needed to step aside from the tedious part and get into a bit of experimentation with actual texturing, so I decided I would start painting the landing gear first. That way, I can have some practice and test my methodology of mirrored/unmirrored maps before the main attraction.

150905-2

September 5 - Front wheel mapping

September 5 – Front wheel mapping

Yikes, I just realized I’m rusty with Photoshop.

150905-4

The rear wheels are done after a hi-res photo of the Tomcat landing gear I found online. The front one was a bit of a Photoshop creation. After this small experiment, I was satisfied and with enough energy to finish the main fuselage division of mirrored vs unmirrored. Now it’s all about making this puzzle work and do a first rough colour of the whole plane.

September 7 - Mirroring fixes done

September 7 – Mirroring fixes done

Advertisements

About Eduardo

I'm some dude with a huge love for animation and film. I've been working in TV animation for the past 6 years in New Zealand. I love dragons, pointy metal things that can cut dragons, and flight simulators.
This entry was posted in Blender Modelling, The YF-19 Project and tagged , , , , , , , , , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s