Welcome to! Click one of these buttons to join in on the fun.

3DU.S.S. Coronado, Katana Class Starship



  • evil_genius_180evil_genius_1801854 Posts: 9,999Member
    That looks awesome.

    How do you get those cool glow effects? Is there a node for that? (please don't tell me it's post processing)
  • Vortex5972Vortex5972206 Posts: 1,108Member
    Looking good.
  • McCMcC347 Posts: 697Member
    edited June 2019 #124
    Thanks, evil_genius_180 and Vortex5972!
    How do you get those cool glow effects? Is there a node for that? (please don't tell me it's post processing)
    It is post-processing, but it's internal to Blender, so it's all just part of the render. Here's the compositing node setup:


    This probably looks super intimidating, but it's actually really straight-forward and the only pieces that really matter to the glow is the Glare node, which I rolled open so you can see the settings.

    Quicky explanation:
    • Add Diffuse Direct and Indirect to each other (makes "Diffuse Light").
    • Add Glossy Direct to a blurred version of Glossy Indirect (helps cut down on Cycles noise; makes "Glossy Light")
    • Add Transmission Direct and Transmission Indirect to each other (makes "Transmission Light")
    • Multiply Diffuse by AO to enhance crevices and such (now "Diffuse+AO")
    • Multiply the Diffuse Color by the Diffuse+AO (makes "Diffuse")
    • Multiply the Glossy Color by the Glossy Light (makes "Glossy")
    • Multiply the Transmission Color by the Transmission Light (makes "Transmission")
    • Add Emit to Transmission (makes "Luminous")
    • Add Diffuse and Glossy (makes "Lit" or "Non-Luminous" or whatever you want to call it)
    • Plug Luminous into Glare so that only the lit-up things glow (makes "Glow")
    • Add Glow to Non-Luminous to get Composite A
    • (Optional) Plug Composite A into Lens Distortion to get some subtle chromatic aberration and use the output of this as your final composite.

    A lot of components, but splitting them out like that gives you tons of control.
    Post edited by McC on
  • evil_genius_180evil_genius_1801854 Posts: 9,999Member
    OK, thanks. I'll have to give that a try sometime. :)
  • McCMcC347 Posts: 697Member
    edited June 2019 #126
    I'm still here! :D

    Finally got up enough steam to get back to work on this lady here and found myself floundering as to what I had last been planning to do. While aimlessly spinning the model around, trying to decide what to tackle next, I noticed a really ugly distortion on the secondary hull, a bit aft of the deflector. Some kind of weird bulge or pucker had formed in the hull panels there, probably due to a minor distortion in the underlying cage mesh that compounded as I built the panels up on top of it.

    Thus began many hours of manual tear-down and rebuild of the affected polygons, making liberal use of LoopTools > Curve and Faces > Grid Fill to reconstruct corrected geometry and re-create the area by hand. I also used this as an opportunity to clean up the blend of the torpedo "nose" delta into the surrounding hull, which had been pretty gnarly before, even if it looked okay.

    Here are a pair of before/after OpenGL viewport captures showing the old geometry and the rebuilt new geometry.

    The above turned into an animated gif:

    Here's the same area from a slightly different angle, showing the area free of distortion as the specular highlight falls across it.

    Now to decide what comes next...
    Post edited by McC on
  • Vortex5972Vortex5972206 Posts: 1,108Member
    Woo! Good to see you back at it.

    Looking forward to what comes next.
  • publiusrpubliusr286 Posts: 1,403Member
    This is one of Sternbach's idea for the NOVA. I think--in the opening of Voyager--a full Federation starship--and not just a Raider--was going to be the Val Jean
  • StarscreamStarscream229 Posts: 1,049Member
    publiusr wrote: »
    This is one of Sternbach's idea for the NOVA. I think--in the opening of Voyager--a full Federation starship--and not just a Raider--was going to be the Val Jean

    Specifically, it originates from the old Nova-class concepts in the TNG Technical Manual, postulating what a future Enterprise-E might have looked like. The more angular of those four concepts probably informed the subsequent Defiant Pathfinder design in the DS9 TM, which in turn was modified into the actual Nova-class USS Equinox for Voyager.

    (Shame, I always preferred the name 'Nova' over Sovereign, which to me seems rather backwards for an interstellar republic. But whatever.)

    Not sure what any of those concepts has to do with the Val Jean though; and there's no record of any full size Starship designs being intended for Caretaker (and it's not really to be expected that the Maquis would be able to fuel and run and full-sized vessel anyway, surely!)
  • McCMcC347 Posts: 697Member
    edited June 2019 #130
    If you're just joining this thread, hi! Please read the first post for some background info. Many questions have come up over and over again that are answered there. Thanks!

    Well. It's been a minute, hasn't it? :D

    It's been a struggle to muster up the energy to work on Coro for the last long while because every time I've opened her I've been faced with the prospect of rebuilding a bunch of too-low-res hull panels on the stardrive. The issue I talked about in my last post turned out to be just the tip of the iceberg. I couldn't progress with detailing there at all until I slogged through rebuilding that section...which I finally have!

    Here's an in-progress open GL shot of what I'm referring to:

    The stardrive after finishing the rebuild (sadly, the differences are practically invisible at this distance):

    But with that done, I could finally start in on the windows (and the rooms behind them) that I first mocked up back in February after posting my last update. Powered through those all day today.

    Now I can finally move on to some actual hull detailing, after all this time! :D
    My goal is to have the model render-complete by the end of August, 2014.
    Ah, the naivete of early project enthusiasm. :rolleyes:
    Post edited by McC on
  • evil_genius_180evil_genius_1801854 Posts: 9,999Member
    It's good to see you back and that you're still working on this. :)

    Wow, that looks like a big undertaking, redoing all of those panels. I probably would have just started over and been nowhere near as far along as you are. :shiner:
  • Vortex5972Vortex5972206 Posts: 1,108Member
    Whoa! Nice to see you still at it.
  • McCMcC347 Posts: 697Member
    edited June 2019 #133
    If you're just joining this thread, hi! Please read the first post for some background info. Many questions have come up over and over again that are answered there. Thanks!

    Thanks, evil_genius_180 and Vortext5972! :D

    Decided to reduce the window count on the stardrive a bit after looking at it from various angles. The below renders aren't exciting; just self-illumination-only to show the revised/reduced window layout.
    coro_2017-09-05-1435.jpg coro_2017-09-05-1440.jpg

    Also added a little shuttle observation/control room at the end of the spine.

    Finally, been hard at work on the ventral "notch"/fantail and the underside of the pylons. This includes a very quick color-only overlay texture UV projected from a plane placed beneath the fantail geometry. I got fed up with how bland it was looking without the intended variations in hull panel coloring, so I took a screenshot of the bottom side of the model and then just painted over it. Throwaway work, but it made me feel better. :D


    I think I need to rebuild the trailing edge of the fantail; seeing segmentation there, too. That, at least, shouldn't be nearly as nasty to rebuild as the hull panels were.
    Post edited by McC on
  • evil_genius_180evil_genius_1801854 Posts: 9,999Member
    That's looking fantastic, bro. I like the reduced windows. Some Trek ships have too many for my taste.
  • RekkertRekkert2432 Buenos Aires, ArgentinaPosts: 2,128Member
    Very nice progress! I looove self illumination only renders. :D
    For my finished Trek bridges and other works of mine, visit my portfolio
    Feel free to contact me if you're interested in commissions!​
  • McCMcC347 Posts: 697Member
    edited June 2019 #136
    If you're just joining this thread, hi! Please read the first post for some background info. Many questions have come up over and over again that are answered there. Thanks!

    Still slowly plinking away on this. :)

    Updates from September 2017

    I did a few more things in September since my last post before putting the model down entirely for a while. As mentioned, I did indeed end up rebuilding the trailing edge of the fantail (and also tossed down some stand-in shuttle landing markers for fun!), as well as adding some more surface detail to the underside of it.

    coro_2017-09-08-1218.jpg coro_2017-09-08-1656.jpg coro_2017-09-08-2316.jpg

    I also added an arboretum just above the aft shuttlebay.

    coro_2017-09-09-2104.jpg coro_2017-09-09-2110.jpg

    Pretty happy with the state of the hull, it was time to turn my attention to the nacelles. Oh, the nacelles. How little did I realize these would plague me!

    First, I mocked up the bits I wanted to steal from the Sovereign nacelles and what I wanted to change.


    Several weeks of fiddling, tweaking, rebuilding, tweaking some more, rebuilding again got me here (apologies for the poor render quality):


    And then...I kind of went dormant for a while.

    Fast-forward: July 2018 to present

    After 9.5 months of very infrequent tweaking, I made it this much further:


    Wow, such progress :rolleyes:

    But also around this time, I started fooling around with Blender's viewport matcap stuff. Applying a bright, cherry-colored car shader to the whole model, I started seeing just how bad the geometry underlying the nacelle was. A whole lot of time went into taking the nacelle cap from what you see on the left to what you see on the right (or top/bottom, depending on your browser, I guess!):

    coro_2018-07-17-1906_matcap.jpg coro_2018-07-18-1209_matcap.jpg

    Now heavily afflicted with matcapitis, I started obsessing over correcting my nacelle topology and spent days and days and days tweaking and rebuilding them.


    The resulting nacelle:

    coro_2018-08-02-1743.jpg coro_2018-08-02-1755.jpg coro_2018-08-03-2205.jpg

    Next, I turned my attention to that aft shuttlebay. I had originally planned to do an Intrepid-style vertical big door, but throughout the life of the roleplaying game, we had always conceived of (and my previous models had featured) that aft bay being a curved, overlapping warped clamshell door, like the Sovereign or Excelsior.

    I knew from screwing around with (and eventually abandoning, in favor of a refit Constitution-style) such doors on my Ambassador, though, that they didn't really "work" from a mechanical point of view. Not in the graceful way that a true hemispherical clamshell does, as found on the original Constitution, but a strict clamshell just didn't look right on this ship. An angled-slats-on-tracks (a la Constitution refit and my Ambassador solution) wouldn't look right on this ship, either. What to do?

    I ended up trying a whole mess of concepts. (These are webp images, converted from animated gifs; I originally tried to embed them as much-much-much smaller mp4s, but neither the video nor img tags seem to support mp4. The mp4s are about 1/3 as large as the webp images, which themselves are about 1/4 the size of the gifs!)

    Excelsior-style, individual slat retraction. Consumes a lot of interior volume for slat storage.

    Excelsior-style, single door hinge with slat-to-slat hinges to allow curvature matching to the hull. Requires less interior storage, but still a lot.

    Sovereign-style, multi-depth layered clamshell with "cheated" pivot points.

    Similar to above, but with the "bottom" of the clamshell wedges constrained to follow a consistent curve. Pivot points are obviously all over the place.

    Excelsior-style, splitting in half and following the upper bay contour, but allowing the slats to "sink" into the hull. Interesting byproduct where the slats stuck into the hull also follow a pretty consistent curvature, which makes this kind of promising. That they stick out of the bottom isn't as big an issue as it seems, since they can be retracted up into the hull a fair amount to avoid this. The downside is that, in profile, these straight slats don't look consistent with the curves of the rest of the ship.

    ST'09-style "Delta" doors, opening in the best way I can figure that they'd open, based on their appearance in both open and closed positions. I mostly did this as a gag, but the idea of the door having a oouple of major parts, each which retract into the hull in their own way, has some merit. But why would you ever build a door like this, anyway?

    That's where things stand as of right now. I still haven't solved the bay door issue and I want to find a feasible mechanical solution that also looks good and befits the ship.
    Post edited by McC on
  • SchimpfySchimpfy171 Posts: 0Member
    I think your delta door layout shows the most promise. it's visually interesting and from an engineering standpoint it seems to be the simplest yet functional design.
  • McCMcC347 Posts: 697Member
    edited June 2019 #138
    If you're just joining this thread, hi! Please read the first post for some background info. Many questions have come up over and over again that are answered there. Thanks!

    Been a minute, hasn't it? :sweat_smile:

    I haven't done much on this since October, but I did end up solving the bay doors in a way I was satisfied with, and then started in on the shuttlebay interior. I've returned to lurking ever since the forum revamp and doing so has gradually ratcheted up the desire to get back into the mix of things, despite the craziness of my schedule these days--currently running four D&D campaigns, including a game for my coworkers, as well as playing in a campaign my wife is running; my wife and I have also taken up indoor rock climbing and try to make it to the rock gym three times a week if we can. Doesn't leave a lot of time for modeling :grimace:


    Still unhappy with any of my bay door solutions above, I started looking outside the box I'd put myself in for completely alternate solutions. One of the old Coro RPers suggested looking to the Luna-class for inspiration, which has a more conventional spherical clamshell door (and thus doesn't have the mechanical issues that non-spherical doors do), but nestled inside some long hull extrusions. Combining this idea with the fantail of the Intrepid-class that had formed the original basis of the ship's aft bay, I ended up with something I rather liked:

    coro_2018-09-03-1440_ogl.jpg coro_2018-09-03-1441_ogl.jpg coro_2018-09-03-1442_ogl.jpg

    Test animation to verify the doors fit/open nicely:

    Here's what she looks like when dressed up for the camera (still with only temporary textures, though):


    Cool! Finally time to move on--or rather, in!

    coro_2018-10-13-2234.jpg coro_2018-10-14-1222.jpg coro_2018-10-14-1456.jpg coro_2018-10-14-1938.jpg coro_2018-10-20-1520.jpg

    For the shuttlebay, I'm taking a lot of design cues from the Intrepid-class bay interiors, while retaining some of the design elements I used on my Ambassador, despite pretty wide technological gulfs between the ships. I figure a shuttle tractor beam is a shuttle tractor beam is a shuttle tractor beam. :grin: The overall bay layout comes from some 2D floorplan work one of the Coro RPers did back in the day, slightly adapted to fit the hull shape of this model.

    I also started messing around with Blender's filmic rendering and redid a bunch of shaders.


    I'm not sure if any of the things that went into this image are actually all that noticeable (other than "it's darker/shinier" or similar), but the above represents the following:
    • Switching from standard to filmic rendering, dramatically expanding the range of virtual f-stops
    • Redid the underlying shaders so that every single piece of hull/metal shares a single "hull shader" master node, which uses Principled BDSF at its heart.
    • Cooked up a shader for the running lights that will automatically color them red/green/white based on where they are relative to the centerline of the model.

    Another angle, with an older render for comparison.


    None of the shader settings are final, and the textures continue to just be quick idea mock-ups, nothing remotely final.

    Sadly, I haven't done much since then, though. Hopefully will find some time to get back to it soon!
    Post edited by McC on
  • evil_genius_180evil_genius_1801854 Posts: 9,999Member
    It's nice to see you are/were still working on this. (it's still an ongoing WIP) I like your solution for the bay doors. Clamshells are the best.
  • RekkertRekkert2432 Buenos Aires, ArgentinaPosts: 2,128Member
    edited June 2019 #140
    Very cool update, the changes on the shaders make the ship look even better than it did before.
    Post edited by Rekkert on
    For my finished Trek bridges and other works of mine, visit my portfolio
    Feel free to contact me if you're interested in commissions!​
  • McCMcC347 Posts: 697Member
    If you're just joining this thread, hi! Please read the first post for some background info. Many questions have come up over and over again that are answered there. Thanks!

    Thanks, evil_genius_180 and Rekkert! Certainly been a long road since that first post at this point. :sweat_smile:

    Today's post brought to you by adventures in rabbit holes.

    Since October's renders, I did some more reading about the filmic lighting model and various aspects of trying to cultivate a more physically-based approach to surfacing and lighting generally. One of the big shifts is the use of correct, high-intensity lights where appropriate...such as when directly lit by the sun. I decided to remove the flat environment light that I'd been using (instead replacing it with an emissive value that more or less matched actual starlight) and cranked up the intensity on my main sunlamp to match the intensity of the sun on the surface of the Earth in broad daylight. The result was quite striking:

    coro_2019-06-11-1645.jpg&size=320 coro_2019-06-11-1716.jpg&size=320

    I had some obvious sampling noise that I was going to have to deal with at some point, but as a first experiment, I was pretty happy with the result.

    Still, that shuttlebay wasn't going to model itself. To give myself some direction, I put together a collage of existing Trek inspirations I wanted to pull from to use as a point of reference. I drew from stills and unused concept art of the Sovereign-class, as well as stills from Voyager of the Intrepid-class and cooked up this:


    I next kicked out a render from a camera inside the shuttlebay with a wide FOV, capturing everything from the edge of the bay doors to the midpoint of the control room window to use as a background plate to paintover and plan direct features.

    coro_2019-06-15-0236.jpg&size=320 coro_aft_bay_paintover.jpg&size=320

    Thus armed with some specific items to work on, I set to work.

    coro_2019-06-16-1122.jpg&size=320 coro_2019-06-16-1428.jpg&size=320 coro_2019-06-16-1638.jpg&size=320

    In all of this, I also kept futzing with materials and rendering settings. That from-the-outside-in shot above took 40 minutes to render, but I noticed that several of the tiles in the center seemed to take much longer to render than the rest of the image, and I couldn't figure out why...until I pieced together that the CPU+GPU rendering mode of Cycles was on by default in my GPU Compute render settings, and the middle 7 tiles were being rendered by the CPU, not the GPU! Dialing down the tile size to something more CPU-appropriate like 32x32 brought the render time back down to ~6 minutes. Turning off the CPU and leaving the tile size at 256x256 was about the same, so I decided to keep things simple and leave the renders as GPU-only.

    Having stumbled across that, I wondered what other rendering and lighting settings I could futz with. I kept on seeing a lot of sampling noise in the renders, which I didn't want to just throw more samples at. Brute force can be the answer, but it should at least be a last resort! I did a bunch of A/B testing by changing one render parameter multiple times and then flipping between results in GIMP to see the difference. I used this approach to settle on a number of bounces for each ray type that seemed to be right around where adding more bounces made little to no difference.

    Even so, with the bay open, the full ship enabled, and the bright sun lamp blasting light into the scene, it was still noisy at 2000 samples (500-2000 is where I usually render these). I did a pass through all my materials, tweaking settings and making sure there weren't incorrect settings (There were some! Escape Pods definitely don't need Multiple Importance Sampling!), simplifying half-complete experimental shader networks down to simpler ones for the sake of simple WIP renders, and so on. Still noisy.

    Frustrated, I decided to really stress-test things by picking a small-but-noisy section of the image and crank the sample count up and up and up to see how high I had to go. Even at 8000, I still had noise! In a huff, I cranked it up to 12000 samples and let it render for two hours.


    Still noisy! Grr!

    I shared some of these images on the SFM Discord and several folks immediately questioned why I wasn't using the built-in Denoiser. In point of fact, quite a few of the earlier images posted do use the Denoiser, but I had turned it off when debugging the CPU+GPU slow render speed issue and pressed on with it off in case using it was masking some other problem(s) (like bad light bounce settings, bad material settings, etc.).

    I also wasn't wild about it because I like to render out each of the passes separately and use Blender's Compositor to assemble the image piecemeal, processing different layers to achieve different effects (e.g. bloom on bright surfaces, the glow on the emissive surfaces, the radial flares on the running lights, etc.) As far as I can tell, there isn't a way to get the Denoised passes from Blender, only the Denoised final image.

    However, Daemoria pointed me toward an alternate plugin that I hadn't heard about that I'm planning to give a try next!
  • McCMcC347 Posts: 697Member
    If you're just joining this thread, hi! Please read the first post for some background info. Many questions have come up over and over again that are answered there. Thanks!

    I have not been idle these lasts two weeks, but I sure have been frustrated!

    My issues fighting sampling noise continued and occupied most of my Blender attention, leading to slow (but steady) progress on the actual detailing of the aft shuttlebay. I'll spare everyone the gory details; suffice it to say, I tried many, many denoising approaches, got a lot of great advice from the braintrust on SFM's Discord, manually compiled a few 3rd-party denoising plugins (the D-NOISE plugin linked above, Intel's OpenImage Denoiser) and tried those, until finally deciding to just dive into a brute-force analysis of the different permutations of settings for Blender's internal denoiser. I planned to try a full range of denoising radii from 2px to 16px, strength settings from 0 to 1, and feature strengths from 0 to 1, rendering out images for all combinations. It was a daunting number of snapshots to capture, but I was at wit's end! I started comparing the various initial results using strength and feature strength settings of 0.5 and noticed that one of them among the rest looked pretty good...

    ...oh, wouldn't you know it: the default denoiser settings!



    Here's the bay interior finished (for now; I'm going to come back and add some floor guide lighting once I put some textures down to indicate landing areas):

    coro_2019-06-27-1736.jpg&size=320 coro_2019-06-29-1703.jpg&size=320

    Ready to move on, I headed up to add some tiny, low-poly furniture to the officer's lounge aft of the bridge. The layout and shape of this furniture follows a 2D plan view done by one of the Coronado players back when the game was still running.


    Odds of any particular render actually showing this furniture is pretty small, but the room and its windows are large enough that I didn't think relying on the apparent parralax of a textured box alone would do the trick.

    Next, onto the upper shuttlebay! :grimace:
  • BrandenbergBrandenberg1557 CaliforniaPosts: 1,872Member
    I just have to say this is fabulous work.
  • evil_genius_180evil_genius_1801854 Posts: 9,999Member
    It looks great. I hear you on the sampling noise, that bugs me in my stuff too.
  • McCMcC347 Posts: 697Member
    If you're just joining this thread, hi! Please read the first post for some background info. Many questions have come up over and over again that are answered there. Thanks!

    Thanks, Brandenberg and evil_genius_180!

    Managed to knock out the modeling of the dorsal shuttlebay over the course of the day today. Hooray for largely-reusable elements!

    coro_2019-06-30-1109.jpg&size=320 coro_2019-06-30-1333.jpg&size=320 coro_2019-06-30-1531.jpg&size=320

    Terrible denoising artifacts on that one, but meh.

    coro_2019-06-30-2137.jpg&size=320 coro_2019-06-30-2155.jpg&size=320

    Noticed some pretty nasty old topology in that last render there around the panels just below the launchways, so I next set to work rebuilding those panels from denser, more evenly distributed curves. Got through 3 of the 5 before realizing it was probably time for bed! :grin:
  • evil_genius_180evil_genius_1801854 Posts: 9,999Member
    Great work on the shuttlebay. :)
  • ashleytingerashleytinger807 Central OhioPosts: 764Member
    Great work on this. Looks fantastic
  • McCMcC347 Posts: 697Member
    If you're just joining this thread, hi! Please read the first post for some background info. Many questions have come up over and over again that are answered there. Thanks!

    Thanks, @evil_genius_180 and @ashleytinger!

    Mentioned rebuilding the panels around the launch platform in the last post, so here's the before/after of that in-process and then completed:

    coro_2019-07-02-2124_ogl.jpg&size=320 coro_2019-07-02-2125_ogl.jpg&size=320 coro_2019-07-02-2126_ogl.jpg&size=320

    I then spent a bunch of time cleaning up the geometry on the launch bay interior to get a clean rendered plate to use as a paintover backdrop to plan out the next stuff for this final interior.


    Those arms about halfway down are inspired by the runabout landing platforms on DS9. Of course, those are pretty clearly Cardassian in designg, so I'll need to retool them to be more aligned with Starfleet aesthetics!

    While working on the paintover, I got to chatting with the old Coro gamers about the size changes in the various bays. The upper bay is about the same, though significantly wider than it was on the "old" blueprints we made back in the day. The launch bay is much longer, while the main (aft) shuttlebay is significantly shorter, but also significantly taller. This in turn led to us talking about the fighters attached to Coro, which we called Specters. They were pretty much straight-up Bearcats from the Wing Commander series, with impulse engines and phasers tacked on. Back in 2001, I made a model of one in Inspire3D (LightWave's kid brother) that we used to make renders for the game's website. I decided to dive back into my old archives to pull out the LWO file and see how hard it'd be to make it look not-awful in Blender. This included hacking the Blender LWO importer from v. 2.79 to work in 2.80! I then spent the next ten hours working through the night on fixing polygons, cleaning up the remnant of boolean operations I'd never cleaned up 18 years ago, re-UVing, and surfacing the model. I decided to stop when the sun started coming up and I could hear birds chirping... :sweat_smile: ...then woke up five hours later to keep going for another three hours until I was happy with the result.

    Original Specter render from 2001:

    Updated Specter in Blender:
    specter_2019-07-06-1325.jpg&size=320 specter_2019-07-06-1319.jpg&size=320

    And only after doing the side-by-side comparison of those renders did I realize that there's a typo in that insignia texture -- that's been there for 18 years! :lol:

    I might go in later and try to further Trekkify the model, but for now, it was a fun diversion and felt nice to finish something in such a comparatively short amount of time.

    With the Specter pulled out of mothballs, it was back to the paintover plan. Drawing heavily from this John Eaves concept of the Enterprise-E main shuttlebay, I used a combination of Blender (to warp the image via planes and UVs) and GIMP (to warp the image in other ways and do the actual paintover) to end up with this:


    Fast-forward a bit more and the bay begins to take shape.


    A lot of that floor detail from the paintover is going to be pure texture, so the bay is going to look pretty barren even after all of the modeled detail is in place, but so it goes.

    For fun, I decided to down a render from long range looking down the launch axis of one of the bays...and immediately noticed a problem: the arboretum was dark. I hopped into LookDev mode and drove the camera inside...where it was fully lit. WTF? A bit more digging and I came to realize that somewhere along the way, the normals of my glass window panes had gotten messed up not just for the arboretum, but in numerous places all around the ship! Cue the arduous process of going through window-by-window and flipping normals until everything was pointing the right way -- the outer pane facing out, the inner pane facing in. The problem seemed worst on the engineering section, with most of the saucer windows still being correct. I bet I was a lot more careful doing the saucer windows--which I did well before the engineering section ones--and then got more cavalier about doing them when I got further along. Sigh.

    Anyway, here's that render with the windows cleaned up, but revealing some more hull paneling surface issues that I'll need to dig into!

  • evil_genius_180evil_genius_1801854 Posts: 9,999Member
    edited July 2019 #149
    I guess a Cornado is what you get when a tornado blows through a con field. ;) I like that Specter model, that's really cool. Like you said, Wing Commander meets Star Trek. I'm jealous you were able to get anything .lwo to import into Blender. I tried a while back, but it was a no go. It might have worked if I'd broken the ship down into smaller parts, but I didn't have access to Lightwave at the time.

    Ah, Sub D and its WTF wireframe that looks OK when rendered. ;) Those new panels definitely look much better. I like your idea for the shuttlebay.
    Post edited by evil_genius_180 on
  • psCargilepsCargile355 Posts: 601Member
    Denoising... Yeah, I've only ever been able to make renders look worse fooling around with the parameters, always going back to the default. I've noticed the worse artifacts on mesh lights and illuminated volumetric, to the point of preferring noise. I also think it's the nature of space based art with black backgrounds and few light sources contributing to noise. Bright, well light scenes tend to render cleaner.

  • McCMcC347 Posts: 697Member
    edited July 2019 #151
    If you're just joining this thread, hi! Please read the first post for some background info. Many questions have come up over and over again that are answered there. Thanks!

    Thanks, @evil_genius_180 !

    Haven't been idle on this, just haven't taken the time to put together a post!

    Continue to add ceiling detail and added big ol' status displays to the main columns. They're about 3 meters tall, so nobody'll miss w hat they have to say! For now, I just slapped the old MSD image that Jester made onto it, but the idea is that it'll display ship status, launch status, and Specter readiness level. That'll be a fun vector graphic to put together... :grimace:

    Next thing to do was the aft end of the bay, which I had been largely ignoring. The main issue is that the wall panel detailing I've been using for all the bays didn't quite coincide nicely with the deck layout for this bay in particular. After some hemming and hawing, I ended up putting a mid-deck observation/control room there, reasoning that people needing to walk up or down a half-deck of stairs was probably fine.

    Originally, I had envisioned that the Specters would be retrieved similar to the way runabouts landed on DS9's landing platforms. Here's what a mockup of that looks like, with mostly-finished wall detailing:

    However, the more I thought about it, the less I liked the idea of the Specters having to gently come in for a landing on a landing pad, and then be slowly lowered back down into the bay. Actuated, moving parts are neat and all, but does it even make sense here? After conferring with folks in the SFM Discord as well as some of the old Coronado players with a vested interest in this particular topic, I decided to ditch the platform and instead go with the idea that tractor emitters placed around the retrieval hatch, as well as in the floor, would be sufficient to pull the Specters back in.

    Regardless, I still needed a hatch through which the Specters could return, which meant making one! Proof of concept, the timing of which almost certainly needs revisiting:
    Post edited by McC on
Sign In or Register to comment.