3DU.S.S. Trafalgar, Ambassador class

McC

After posting a question in Fractalsponge's thread, I realized I'd never actually posted any of my work to SFM. I suppose that makes some amount of sense, since I haven't really done a lot of hobby work for the last several years, but that has changed and here I am!

The Ambassador holds a special place in my heart, since I used to run a Trek RPG based on an Ambassador-class ship called Trafalgar. For this model, I'm using a bunch of different schematics and reference shots of the studio models (both Enterprise-C and Yamaguchi variants). My inspiration is dwl's Intrepid-class variant, U.S.S. Horizon, and I plan to take a similar approach to retaining what I like about the design and changing what I don't.

I started this back in May '11, but didn't make much progress before setting it aside for a while. I just picked it back up again last month and have been slowly chipping away at it ever since. Along the way, I discovered Blender's new Cycles rendering engine, and have spent no small amount of time playing around with material and render settings alongside modeling.

8/20: Saucer panel lines are a pretty close match with the studio model, and share the exact count.


9/1: All of the luminous parts of the ship have refractive, transparent covers over actual light sources, which should make for some neat visuals for any animations.


9/3: Among other things, this shows my first crack at the warp grilles, which treated them like large neon tubes. I would later realize this was wrong.


9/3: Windows cut into the dorsal surface of the saucer. I'll have to re-do all of this, including the panels, since I am unhappy with the level of segmentation.


9/4: Removed the neon tubes and made the warp nacelles hollow, placing actual warp coils inside.


9/10: Put in proper warp grilles, based on this shot from TNG. The warp coils can be seen behind the grilles.


Feedback, especially critique, is more than welcome! It's been a very long time since I actually finished a model, and community prodding might just be the trick to getting across that finish line.

VALKYRIE013

Yay! Another Ambassador! good to see this ship get so much love!

Link to all my Ambassador refs I have, hope they help!
Link

Looking great! hope you get plenty of time to finish her!

evil_genius_180

We can never have too many Ambassadors. It's one of my favorites.

Really nice work so far. I like the level of detail you're putting into this, especially with things like the warp coils.

McC

Thanks, VALKYRIE13 and evil_genius_180!

VALKYRIE13, I see we "shop" at the same reference image store. Thanks for posting those, though! There were a couple I hadn't come across before.

I've been modeling for about 15 years now, and still feel like a complete noob most of the time. Take tonight, for example. I just learned tonight that edge weights--a term I'd heard many, many years ago, experimented with briefly, and dismissed--are incredibly useful for sub-D modeling. :argh: I'd just used them wrong when I first tried them. I thought they were a poor man's way of adding edge-refinement/control cuts. Little did I realize their utility when applied to said edge control cuts! The prospect of cleaning up a frozen mesh just got way less nasty, and the amount of fine tweaking control loops just got way less annoying! (Thanks to MadKoiFish's post in nivao's thread, which got me thinking about their utility once again.)

That said, there's hardly anything new to see in these renders, despite having made modifications to every single component of the ship. Every tight-corner edge now has edge weights on its control loops, which provided the whole model with much better topology. I deleted and rebuilt the old saucer, this time as a sub-D cage rather than polygons. As a result, the neck doesn't quite match up right at the moment. I also made each of the warp grille elements 25% larger, to tighten the gaps between them. The warp engines in general got a fair amount of TLC, but most of it is too subtle to see.

Beauty shot


Beauty shot, ship self-lit


Warp nacelle detail shot


Sub-D wires, for fun


Though this is almost certainly subject to whatever whim strikes me next time, I suspect I'll tackle freezing, cleaning, re-paneling, and re-windowing the saucer next. Or, I might spend more time noodling with the warp engines, since I seem to spend so much of my time on them anyway. :rolleyes:

On a separate note, I decided to dig up the renders of the Ambassador model I made in LightWave...11 years ago!

BolianAdmiral

Very nice... I love the coils inside of the nacelle... nice detail touch!

Brandenberg

Simply Fabulous! Please finish it!

killakanz

You said in the first post that you're taking reference from both Ent-c and Yamaguchi, but looking at it it seems you're basing heavilly on the early varient. The only part I see of the latter is the sensor module under the saucer, so are you aiming for this ship to be from just before the major refitting?

It's a great start though. Always nice to see new Ambassadors appearing. Even if they do keep reminding me that I need to sort out my own

McC

Thanks, BolianAdmiral and Brandenberg!

killakanz wrote: »

You said in the first post that you're taking reference from both Ent-c and Yamaguchi, but looking at it it seems you're basing heavilly on the early varient. The only part I see of the latter is the sensor module under the saucer, so are you aiming for this ship to be from just before the major refitting?

I'm going for a sort of sensible hybrid version of the two, with a healthy dose of my own take on the detailing thrown in for good measure.

Things like the Yamaguchi Bussard covers don't make a lot of sense, for example, because you're going to want as much collector surface area as possible. Cutting that down doesn't make sense in the context of the component's function, and I'm one of those spoil-sports that insists that science fiction make sense at least within the context of itself.

On the flip side, I was never a big fan of the unadorned sensor dome that the original Constitution and Enterprise-C both had on their ventral saucer surface. I much prefer the more elaborate one seen on the Enterprise refit and Excelsior, which also makes an appearance (albeit rotated 90 degrees) on the Yamaguchi variant. Similarly, I keep ripping out and rebuilding the bridge, because I keep forgetting that I don't want to follow the diagrams or studio models, but rather want to try to model an exterior that makes sense based on the actual design of the bridge. This will probably result in a low-res model of the bridge interior, which might turn into a high-res model of the bridge interior, depending on my gumption level.

I also haven't decided what I want to do with the neck/impulse engine area yet. The Yamaguchi's under-the-saucer version is simpler to do, but I kind of like the oddness of having the neck zig-zag back out from under the saucer the way the Enterprise-C model does it. My uncertainty here is part of why you haven't seen renders from this angle yet.

The big one is going to be the tail, though. I don't think either of the variants has a shuttlebay that was well thought-out. The Enterprise-C version is very steep and narrow, and the Yamaguchi version doesn't seem like it would actually have a working door. I will probably just do my own thing here. I also want to put a proper aft torpedo launcher in somewhere, space for which doesn't exist on either of the two studio models.

It's a great start though. Always nice to see new Ambassadors appearing. Even if they do keep reminding me that I need to sort out my own

Doooo eeeet! I've been keeping an eye on a number of the more-recent Ambassador threads for inspiration, yours included.

Starscream

McC wrote: »

Things like the Yamaguchi Bussard covers don't make a lot of sense, for example, because you're going to want as much collector surface area as possible. Cutting that down doesn't make sense in the context of the component's function, and I'm one of those spoil-sports that insists that science fiction make sense at least within the context of itself.

In that case I'm going to be an equal spoilsport (and not just because I happen to like the Yamaguchi's Bussard collectors!) and suggest you move the nacelles far enough down so that the collectors aren't obscured by the saucer. :P

I'll never understand that move on the original model -- as far as I'm concerned that's as nonsensical as placing the Impulse exhausts directly in front of the nacelles...

Can't wait to see more!

evil_genius_180

Your old model was good for its time. I've seen some of the public release and some others from back then and most weren't that accurate. However, your new one is clearly better so far. Isn't it fun revisiting old builds and improving them?

McC

Starscream wrote: »

In that case I'm going to be an equal spoilsport (and not just because I happen to like the Yamaguchi's Bussard collectors!) and suggest you move the nacelles far enough down so that the collectors aren't obscured by the saucer. :P

I'll never understand that move on the original model -- as far as I'm concerned that's as nonsensical as placing the Impulse exhausts directly in front of the nacelles...

...okay, deal. I actually hadn't even thought about that until re-reading the EAS page on the differences in the two models, and as soon as I read it my first thought was, "...damn, given what I just said about the Bussard collectors, maybe I should look at doing that, too!"

Can't wait to see more!

Thanks! I should have a chance to work on it more tonight.

evil_genius_180 wrote:

Your old model was good for its time. I've seen some of the public release and some others from back then and most weren't that accurate.

That old one hewed pretty close to the Mike Swantak schematics. I think I built most of it from profile curves and hard polygons, rather than starting with sub-Ds. Ah, the hallmarks of an earlier era.

For this build, my background images are the "Gus" schematics, though I'm using them largely for overall proportion rather than exact alignment, with frequent consultation of studio model photographs.

However, your new one is clearly better so far. Isn't it fun revisiting old builds and improving them?

Thanks! And yes, indeed!

McC

Another night of work with few discernible differences. In no particular order:

• More tweaking of the nacelles, largely to fix smoothing issues in the rear and to put in some edge weights I'd missed before.
• Created an Empty that "echoes" the nacelles, which I can then lower so that the nacelles are below the saucer. I'm doing it this way so that I can continue using my reference images as a guide without having to modify them, and not have to remember to move the nacelles each time I render.
• Rebuilt the entire front part of the engineering hull to better-match the studio model's shape.
• Rebuilt the shuttlebay area. Still not sure how I feel about it. I might leave it for last.
• Gave the front of the neck a gentle slope into the saucer and engineering hull.
• Made a decision and went with the Enterprise-C-style neck overhang, rather than the Yamaguchi-style neck. Fleshed out the shape in this area a bit.
• Increased the height and depth of the impulse assembly by about 15% over what it was before.
• Reduced the internal impulse thruster count from 4 to 3. (These are internal bits behind the impulse exhaust grille.)
• Futzed with the saucer command superstructure/island. Do these things have a common name? I'm talking about the bit that includes the bridge, and the decks the bridge sits on, above the saucer proper.

Lots of little tweaks, but few of them really visible. I think I'm partly just avoiding having to freeze my sub-Ds at this point.

The nacelle pylons still need some shape work and I need to put the spine back in, but other than that, is there anything obvious in the sub-Ds that I should refine before freezing and beginning the arduous process of paneling?

Also, I need to come up with a better-looking lighting rig.

evil_genius_180

Everything's looking great, nice and smooth.

McC wrote: »

Futzed with the saucer command superstructure/island. Do these things have a common name? I'm talking about the bit that includes the bridge, and the decks the bridge sits on, above the saucer proper.

A lot of modelers refer to that structure/area of these ships as the B/C Deck because it's directly below Deck A, which is the bridge.

Aresius

nice.

Starscream

^ what I was gonna say.

I have just noticed after looking more closely at the aft view though, your warp field grills at the ends of the nacelles appear to be bulging outwards toward the middle -- as I recall from the reference model, those should be creating a flat profile.

McC

Thanks, Aresius and Starscream!

Starscream wrote: »

I have just noticed after looking more closely at the aft view though, your warp field grills at the ends of the nacelles appear to be bulging outwards toward the middle -- as I recall from the reference model, those should be creating a flat profile.

Yeah, I caught that last night as I was posting these. Will definitely get that fixed. I'm also trying to decide what to do about making the nacelles look less flimsy, given how hollow being able to see through them makes them feel. Might just have to build out more stuff inside the nacelle, so it feels like a complete thing.

McC

Removed the entirety of the interior of the nacelles, added additional coils to fill in the space. Still need to do something about how flimsy it looks in there, but it's a step. Also, fixed the profile of the grille, so that it's curved at the front, but straightens to be vertical as it approaches the back of the nacelle and goes into the curve. Finally, replaced my "two giant panels" lighting rig with a three-point spherical emitter rig.



More to come!

McC

I took the plunge and, after a few tweaks to the rim, froze the saucer!



Initial freeze polycount: 325,632 faces (651,264 tris)
Post-cleanup polycount: 122,880 faces (244,224 tris)
Radial segments: 768
Polygon normal tolerances within 0.25%

The 768 radial segments may seem somewhat excessive, but if you think about it, it works out to ~2.1 segments per degree, or half a degree per segment. For really close detail shots...eh, probably not necessary even then. I don't know; worth reducing, or should I just leave it?

And now, I'm going to bed!

killakanz

McC wrote: »

...okay, deal. I actually hadn't even thought about that until re-reading the EAS page on the differences in the two models, and as soon as I read it my first thought was, "...damn, given what I just said about the Bussard collectors, maybe I should look at doing that, too!"

That's a good point, one that came up in my thread too. In the end, like you, I decided to bring them down just enough for the bussard collectors to clear the saucer, but that didn't include the caps! So yours may have to come down a bit more.

As for the aft torpedo bay... well that's upto you. While making mine I had suggestions like the neck, the pilon horizontals, lower secondary hull where it arches up... However on my model I have an extrusion running down the spine and I think the end of that, just above and slightly in front of the shuttlebay door, would be a nice place to put it.

VALKYRIE013

I put my rear torpedo launcher in the eng hull warp cutout.. tryied to fit it above the rear shuttlebay, like the Ent D stinger.. but not enough room..
Looking good!

Ortho

Starscream

The eventual location you picked Valkyrie was perfectly fine, IMO at least. It helped prevent the cutout from looking too bland also.

My only minor nits about your mesh at this point would have been the number of windows in the secondary hull (yes, I know the original had them!), and that you've got light being emitted from the internal nacelle field coils (IIRC it should be coming from the field emitter grills instead). You've already mentioned the issue of the Dreaded Bussard Caps...

McC Your engineering hull is looking mighty fine from that angle.

McC

Thanks for the feedback, everyone!

I had no time whatsoever to work on it this weekend, sadly. Friday, my wife and I ended up deciding to go to a party that we hadn't originally planned to attend, Saturday we tried to recover from that and then had more people over for another round of chicanery, and then Sunday we spent the day shepherding a friend of ours through the David Tennant-to-Matt Smith transition (all four specials and then the first episode of series five!). Not sure what my free time looks like this week, but this coming weekend isn't currently booked solid, so hopefully as the week goes on, I'll have increasing amounts of time to devote to this!

On the topic of windows, I was thinking about abandoning my reference images (fan-created orthos and studio models alike) when it came to windows, and instead using (fan-created) deck layouts as a guide. The general arrangement of the windows would still be the same, but with the quantities and specific layout governed instead by the internal structure of the ship. This might also help inform a good placement for the aft torpedo tube.

McC

Not much of an update this evening. This week remains ridiculously busy, but the end is nigh! Can't wait to get back to doing some actual modeling on this.



These are the "corrected" shield grid ring locations, based on comparing images of the studio model. I'll use these as my template for modeling my gridlines.

VALKYRIE013

Grid looks good! yeah.. eng windows.. for some reason I didn't check the plans I was using to the refs.. and put WAY to much windows in the eng hull... owell.. use the plans as a good start, and cross check on the refs of the model

Good luck!

ryo80

Agreed, the plans you have look pretty good. And very nice work on the nacelle grille/coil configuration. I looked at the filming model so long when I built mine, that I just threw it together without thinking about how to render it without being "clay" shaded. As for the stardrive windows, I tried to keep my setup as close to the model as possible, but when I tried to straighten them all out, a couple fell off of the top of it, below the "neck." I'm very interested to see what you come up with in that area. Great work, so far!

Schimpfy

I've been following this thread and I just want to say I love what you're doing with the Ambassador. I'm really looking forward to see where you take this and, also, awesome job on the nacelles. I like when there's substance to the inner workings of things.

McC

Thanks very much for the encouragement, VALKYRIE013, ryo80, and Juvat! It helps keep me going.

Blah! This took far longer than it should have, largely because I got the fool idea in my head that trying to boolean shield grids was a smart idea. After 3-4 hours down that rabbit hole, I took a step back and did some poking around in the tutorial subforums here, where I realized that the inset poly group/extrude method would make way more sense and take way less time. Less than two hours later, we are here!



They aren't actually inset into the hull yet. Blender's Extrude Region tool seems hellbent on computing the average normal of the selected polygons and using that, while the Extrude Individual tool gives me the right normal vectors, but as the name implies extrudes each selected polygon individually, leaving me with a bunch of polygon "walls" that I don't want. Does anyone know a way to extrude a group of polygons, but along their local vertex/edge normals? Sort of a hybrid version of Extrude Region and Extrude Individual?

evil_genius_180

Dude, boolean shield grids are the way to go. That's all I do (but I also don't use Blender. )

McC

evil_genius_180 wrote: »

Dude, boolean shield grids are the way to go. That's all I do (but I also don't use Blender. )

How do you generally do it using booleans? I'm curious if the method could be applied in Blender.

In LightWave, I would use the Solid Drill or Stencil tools for hull detailing like this, which I've sorely missed in every other 3D package I've ever used. I think MAX has something equivalent to this, but I can't recall what it is. I know Maya does not.

It's not really useful to lament what your current 3D app doesn't have, unless you can actually do something about it (which, technically speaking, I probably could...if I had the time; I'm quite comfortable in Python). So far, this method is working alright, though.

One more update for tonight! I am very glad that this saucer is radial; means I only have to do 1/8th of the work (assuming I then mirrored the 8 radial wedges from one side to the other). If the saucer shape wasn't so regular, I'd have to do all the panels, and that would be incredibly tedious!



The grids are now properly inset (I chose a 15cm width and depth for them, which is an approximate compromise from the various measurements I gleaned from the studio model; ~5cm on the low end, ~30cm on the high end) and even microbeveled! Blender's Edit Mode Bevel is terribad (they're working on it!), but its Bevel modifier using Bevel Edge Weights is fantastic, so that's what I ended up using.



I actually made a small technical mistake on the lower grids, so I'll have to go in and re-do them. It's not actually something you're ever likely to see, but knowing it's there will bug me if I don't fix it. Fortunately, this method doesn't take very long to do, so that should be quick. I also flat-shaded the saucer for these renders, because the smoothing setup I was using looked all ripply. I'll go in and fix that next time, too.

And I should really model that bridge at some point...

evil_genius_180

She's looking great.

I use trueSpace, which is old and outdated, but it handles booleans better than some other software. In fact, booleans is the way to go on a lot of stuff with it. I've heard of other methods of doing lines in it, but the booleans is quick and easy and it doesn't suffer from a lot of the error problems that programs like Max and Blender have when using booleans. I have no idea if what I do in tS can be applied to Blender because I've only used booleans once in Blender and that was to cut a simple hole through something.

McC

evil_genius_180 wrote: »

I use trueSpace, which is old and outdated, but it handles booleans better than some other software. In fact, booleans is the way to go on a lot of stuff with it. I've heard of other methods of doing lines in it, but the booleans is quick and easy and it doesn't suffer from a lot of the error problems that programs like Max and Blender have when using booleans. I have no idea if what I do in tS can be applied to Blender because I've only used booleans once in Blender and that was to cut a simple hole through something.

Hah, no kidding! I actually cut my teeth on trueSpace3 and trueSpace4 before migrating over to Inspire/LightWave. This was more than a decade ago now, but still. Fun stuff.

She's looking great.

Thanks!