@Professor Moriarty - Now that you're getting some great experience with sub-D modeling, how are you getting along with texture mapping the objects? It looks like you've got a handle on the materials.
So, I made this.
- Thread starter Professor Moriarty
- Start date
By cheatin' like the sleazy sumbitch that I am! I will save learning how UV mapping works in Octane for later; right now, everything I'm texturing can be accomplished with one of the geometric mapping projections that Octane has (mostly Box, Spherical, and Cylindrical). That doesn't mean that the materials themselves are simple. This one was especially tricksy:
This is the material that covers the hull on the upper port side forward of the entry hatch, which I applied with a "Box" projection that was tilted a few degrees to place the PNG image file decals perpendicular to the slanted surface. There is a scene I'm planning where you'll see multiple shuttlecraft in the hangar bay, so I have to keep the shuttle name (the Galileo II is shown here) and the "/x" part of the registry ("/7" for the Galileo II) as separate PNG images that must be combined together with the parts that don't change (e.g., "NCC-1701"). This becomes even more hairy when different materials are needed for each decal. For example, the gold arrowhead pennant is a metallic material, while the painted-on shuttle name and red striping are a glossy material and the hull itself is yet another material (it'll actually be two more materials when I go back to add some dirt and streaks). Octane's composite material makes this easy enough to accomplish, but wow it makes for some spaghetti texture node connections!
ETA: The Galileo II is 100% sub-D modeled, by the way. Not necessarily the best approach for some of the parts that are in separate layers (the way Octane works, everything that's in a layer must either be sub-D or not sub-D), but since this was partly a learning exercise I set out to make her 100% sub-D to get experience, and that's what I did. I haven't frozen the mesh, so when I render in Layout I choose the number of Catmull-Clark subdivision levels to use. "3" seems to work no matter how close I get to the model, so I leave it at 3. Because of the way that Octane works, it's not worth the hassle to go through all 26 layers of the model to change the subdivision levels to a lower number when the shuttle is further away from the camera; level 2 renders only a couple of seconds faster than level 3, so I just leave it at 3 all the time. That did take some getting used to... at level 3 the shuttle weighs in at an eye-popping 5.6 million polys! But Octane is so stupid fast, who cares? Those 4K frames rendered in 2 minutes apiece, and so did this one.
This is the material that covers the hull on the upper port side forward of the entry hatch, which I applied with a "Box" projection that was tilted a few degrees to place the PNG image file decals perpendicular to the slanted surface. There is a scene I'm planning where you'll see multiple shuttlecraft in the hangar bay, so I have to keep the shuttle name (the Galileo II is shown here) and the "/x" part of the registry ("/7" for the Galileo II) as separate PNG images that must be combined together with the parts that don't change (e.g., "NCC-1701"). This becomes even more hairy when different materials are needed for each decal. For example, the gold arrowhead pennant is a metallic material, while the painted-on shuttle name and red striping are a glossy material and the hull itself is yet another material (it'll actually be two more materials when I go back to add some dirt and streaks). Octane's composite material makes this easy enough to accomplish, but wow it makes for some spaghetti texture node connections!
ETA: The Galileo II is 100% sub-D modeled, by the way. Not necessarily the best approach for some of the parts that are in separate layers (the way Octane works, everything that's in a layer must either be sub-D or not sub-D), but since this was partly a learning exercise I set out to make her 100% sub-D to get experience, and that's what I did. I haven't frozen the mesh, so when I render in Layout I choose the number of Catmull-Clark subdivision levels to use. "3" seems to work no matter how close I get to the model, so I leave it at 3. Because of the way that Octane works, it's not worth the hassle to go through all 26 layers of the model to change the subdivision levels to a lower number when the shuttle is further away from the camera; level 2 renders only a couple of seconds faster than level 3, so I just leave it at 3 all the time. That did take some getting used to... at level 3 the shuttle weighs in at an eye-popping 5.6 million polys! But Octane is so stupid fast, who cares? Those 4K frames rendered in 2 minutes apiece, and so did this one.
Last edited:
The Tab key is your friend when doing this. Figuring out how the edge polys flow when they're subdivided (so that you can correct issues like pinching) is still a bit of trial and error for me, so I'm constantly tabbing back and forth in Modeler to check my work. Unfortunately I am finding that Octane isn't as forgiving as I originally thought. It really doesn't like ngons on curves and when it sees something like that it usually will leave the offending geometry undivided (e.g., a smooth fillet in Modeler will sometimes have a sharp edge in Octane). I sometimes have to add extra edges to get to a flat spot so that everything works.
Thanks I'll have to watch out for that! I had been freezing things that I had subdivided before going to Layout (like the Burke chairs and the Class J) but will experiment with leaving them subd in Layout.
The great thing about letting Octane do the subdivision on the fly in Layout is that you can leave the simple mesh undisturbed in Modeler... invaluable for rendering test shots while you’re still modeling your objects. It does add a little overhead on every frame, though, while Octane computes the subdivided polys.
Awesome, maybe (thanks!) Undersized, definitely.
I'm now trying to decide what to do here: When I redid the shuttlecraft model, I used blueprints that appear to have been based on the human-sized studio model... except the human-sized studio model was not full-sized! It's only a little over 21' long, but onscreen, unassailable dialog says that the shuttles are 24 feet long. And 24 feet long makes a lot more sense; at 21' a six-foot human being would really have to squnch up to sit in the seats, and the eyeline of the forward viewports would actually be beneath the eyeline of the seated pilot (unless the pilot was bite-sized).
I know the interior model was comically oversized so that the actors could stand straight up, so I'm not enlarging the shuttles to fit the interior. But the 21' shuttle is just too darn small. What have other modelers of the class F shuttle settled upon for its length when you've modeled it?
I'm now trying to decide what to do here: When I redid the shuttlecraft model, I used blueprints that appear to have been based on the human-sized studio model... except the human-sized studio model was not full-sized! It's only a little over 21' long, but onscreen, unassailable dialog says that the shuttles are 24 feet long. And 24 feet long makes a lot more sense; at 21' a six-foot human being would really have to squnch up to sit in the seats, and the eyeline of the forward viewports would actually be beneath the eyeline of the seated pilot (unless the pilot was bite-sized).
I know the interior model was comically oversized so that the actors could stand straight up, so I'm not enlarging the shuttles to fit the interior. But the 21' shuttle is just too darn small. What have other modelers of the class F shuttle settled upon for its length when you've modeled it?
I have a similar quandary in that I had modeled my interior flight deck on the 21'-22' shuttle so when I scale it up the shuttle the size of my Enterprise goes up too
But I'll end up scale it up to 24'. It'll look fine and I bet yours will too scaled up
I've noticed the same thing with the shuttle model, although mine is a "reimagining." The larger size pushes the limits of what the JR-Prise can hold, especially the bay behind the bridge, and the JR-Prise is already considerably larger than the Jefferies Connie. I decided to go with the smaller size for the shuttle. IRL I don't think aircraft offer standing room until you get up to the bigger business jets.
24 / 21 = 1.1429 times larger.
Hmm...
1.1429 x 947 = 1082
aaahhh crap.
Hmm...
1.1429 x 947 = 1082
aaahhh crap.
Almost perfect...
Ah yes, the good old days when a material setup could fit on one screen. Well, it still can... sort of...
Each of those nodes in the third column contain on average about as many nodes as that setup I posted a few weeks ago. And this is still not the most complex material I've achieved on the Galileo II (that would be the material used on the casing of the warp nacelles). Organization (and liberal use of Lightwave "Compound" nodes) is definitely a must when you get to this stage. Hopefully the end result will be worth it...
Excellent-looking work!
This is very cool and educational @Professor Moriarty ! Thanks for sharing 
It's going to be a while before I get to that level of sophistication.
It's going to be a while before I get to that level of sophistication.
Last edited:
but awesome
So I decided last night that the 2013 standalone hangar deck model was gonna get the ole heave-ho too. Too many errors needing fixing, and because it's a standalone maquette and not part of the Enterprise, I wouldn't be able to use it for shots where I'm outside of the ship peering into the hangar. So, I bit the bullet and starting finishing the hangar deck built into the Enterprise. It's actually in a separate *.lwo model that's pieced together with the rest of the ship only when needed, so that when the doors are closed it's not part of the exterior Enterprise model. The reverse holds true too; when I'm shooting from angles inside the hangar where you can't see the warp nacelles (or when the clamshell doors are closed), the hangar_deck.lwo model can also be used standalone.
So, I have a little more modeling to do, and then some detailed texturing to match the level of detail on the Galileo II, and then I'll be completely finished with the Enterprise. But for now, here's a quickie test work-in-progress shot. I call this, "Ensign Ricky Has a Brown Alert".
"Shuttle One, execute an immediate ninety degree turn to port! We still have a craft on the pad in here!"
(Morgan Freeman V.O.): Ensign Ricky would later realize this was the moment his career as a Starfleet shuttle pilot was over.
So, I have a little more modeling to do, and then some detailed texturing to match the level of detail on the Galileo II, and then I'll be completely finished with the Enterprise. But for now, here's a quickie test work-in-progress shot. I call this, "Ensign Ricky Has a Brown Alert".
"Shuttle One, execute an immediate ninety degree turn to port! We still have a craft on the pad in here!"
(Morgan Freeman V.O.): Ensign Ricky would later realize this was the moment his career as a Starfleet shuttle pilot was over.
Besides the trick retracting tailgrear and the navigation lights, I tried to keep the Galileo II as screen-accurate as possible. Her sister craft (piloted by the hapless Ensign Ricky) has a few enhancements that I’m keeping under wraps until June 16th.
Similar threads
