It's been too long since my last update. I've been working on it steadily but life has been 'interesting' in the meantime. Finally done with the nacelles. I say 'done' even though there are still a few things to do (e.g. rebuild one piece from scratch and drill out the spotlights from the front). I am almost 99% certain that before long I'll be rebuilding the entire model so a huge grain of salt should be taken with just about any part being described as 'done' or 'completed'
On with the show!
I knew the nacelles would be a challenge; I've tinkered with building them in Max before. They are definitely a vexing collection of interesting geometries. Quite a learning experience, no doubt about that!
As I approach the 'finish line' of the model itself, I've started planning out some things that I would like to do once she is fully realized. Some of these will understandably require additional ships to be modeled (darn! ) so they will have to wait but many will require only simple props and creative cycles application. I can't wait!
Fore view
Aft View
Ventral View
Additionally I've already set out to willfully deviate from the studio models in a few key areas (the horror). It's already been commented on that the access panels that I have modeled on the pylons were never anything more than paint/decals on the studio models. Not modeling those was simply not an option. Adding to this list is my treatment of the RCS thrusters on the nacelles. Everywhere else on the ship the RCS pods are visibly separated by a seam, implying modularity and easy of maintenance. I have decided to continue this trend with the nacelle RCS. You can see it here in the pic below.
Nacelle RCS detail:
Purists may cringe but I have to stay true to the spirit of the model and I feel that may occasionally require that I deviate from a 'true' reproduction. Additionally and amusingly still on the subject of the RCS thrusters, I have every intention of lighting them properly. The studio model lit the damn things like they were windows. I feel that they should be lit only when the ship is adjusting attitude. We'll see how it looks when the time comes...
And now the evolution shots. This was the first time that I put some real work into the animated progress shots. From the onset I created three dedicated cameras to record the action. While it did slow the work a bit, it also forced me to evaluate the work from different angels throughout the build.
Fore
Aft
Ventral
Now to finish engineering (including more windows than I care to count), build out the shuttle bay doors, the shuttle bay (maybe... probably ), and finally the VIP lounge. Then to begin texturing!
Wow. Amazing work there. :thumb: I just love those evolution shots of yours. I fully agree with your wont to "deviate" from the model. As you describe it I believe it will more aptly depict the true intentions of the modelers.
Personally, I always like to put a little personal touch on every canon model I do. So, I understand your desire to change a few things here and there.
You ever blow an entire day on a section of a mesh that you know, objectively, should be scrapped? You keep fooling yourself into thinking "no, just one more tweak and all those flaws will disappear. It'll be 'perfect'." Or, worse yet, you find yourself thinking "I can barely see that flaw and I build the damn thing! Who will even know that it's an abominable mash-up of quads and tris that only the gods of sub-surf could love?".
Today has been exactly like that. And I think anyone who has modeled details into that damn fantail, no matter the generation of the Star Trek ship, will know that pain all too well.
I've just wasted 12 hours on a faintail that I'm abandoning and rebuilding. I really did a number on myself with the subsurf tool and the end result just isn't keeping up with the rest of the model. I'm going to try to cut off the fat, extrude anew from a cleaned up, unified edge, bevel it and call it a night.
I'm sick of this part and ready for better things :rolleyes:
I recently finished Dishonored, playing with as low a kill score as I could manage. I'm sorely tempted to try it again with a radically different MO...
Just discarded my latest attempt. Turns out the flaws were caused in part by a concavity I hadn't noticed on the fantail. Between that and the non-quad faces that I had ignored, the result is too much of a mess.
I've reverted to the mesh as of Friday <wimper /><sob /> and I'm cleaning it up properly this time.
Fortunately I still have all of my stencils for the grid and windows that I spent yesterday installing :shiner: so I can reapply them once I finish teasing the loop-cuts back into the familiar tapered cylinder we all know and love.
In the end I lose a few days work but the result will still be worth it. And a lesson learned about ignoring best practice for subsurf!
I learned some valuable lessons on this phase of the project.
1.) Always proof your mesh before applying the subsurface modifier. Quads or tris but never both. See my earlier posts for additional details and some overly dramatic self-pity.
2.) If a knife projection isn't working properly, try sliding some of your edges. Previously I had been subdividing faces to get greater resolution. This time I found that sliding only two edges got all of the detail I needed without upping my face count unnecessarily.
3.) Always, always, always inspect the results of bevels on a mesh of any significant detail. I've known that bevels can do weird stuff but this took the cake. I had vertexes thrown miles off course in a few cases. Some I could hand-tease back into shape, others I had to back out entirely and then re-apply in smaller batches.
I still have to add top details, the shuttle bay, and some accents on the bottom before the hull will be feature complete. I'll be adding the central shield grid line on the ventral and fantail grids once I apply the mirror modifier. It cost me nearly a week of evenings to undo my earlier mistakes but the result is better than before so I'm really happy with the outcome.
Very happy with how she's looking at the moment.
Fore view
Port view
Aft view
And the progress gifs. Not as impressive this time since the bulk of the geometry is either already complete or invisible (when done right) but I still like looking at how it flows from start to finish.
1.) Always proof your mesh before applying the subsurface modifier. Quads or tris but never both. See my earlier posts for additional details and some overly dramatic self-pity.
This isn't strictly be true. Quads, tris, 5-gons, n-gons are all fine for subsurf, but you have to be aware of the impact non-quads have on topology. Non-quads create poles, which behave in predictable, but not necessarily smooth, ways when subdivided.
2.) If a knife projection isn't working properly, try sliding some of your edges. Previously I had been subdividing faces to get greater resolution. This time I found that sliding only two edges got all of the detail I needed without upping my face count unnecessarily.
This is a neat tip! I've had this bite me more than a few times.
3.) Always, always, always inspect the results of bevels on a mesh of any significant detail. I've known that bevels can do weird stuff but this took the cake. I had vertexes thrown miles off course in a few cases. Some I could hand-tease back into shape, others I had to back out entirely and then re-apply in smaller batches.
Ugh, yes. Usually, this stems from some kind of face/edge/vertex normal irregularity and can be cleaned up, but good grief can it be awful if you don't catch it beforehand.
It cost me nearly a week of evenings to undo my earlier mistakes but the result is better than before so I'm really happy with the outcome.
3.) Always, always, always inspect the results of bevels on a mesh of any significant detail. I've known that bevels can do weird stuff but this took the cake. I had vertexes thrown miles off course in a few cases. Some I could hand-tease back into shape, others I had to back out entirely and then re-apply in smaller batches.
Something I found out (by accident, I think) is that selecting the entire mesh and recalculating normals seems to eliminate any problems from beveling. Most of the bevel errors I came across before were edges crossing over one another due to normals that got turned around at some point, possibly from mirroring. That's pretty much why any time I'm getting ready to bevel, I save, select the entire mesh, and hit "Recalculate". I haven't had any bevel problems since I started doing that regularly.
As any Enterprise veteran knows, the shuttle bay is a collection of challenges and this time it was no different. My sincere regret is that I failed to rough in the shuttle bay doors back when I was modeling the engineering hull. I am not 100% happy with the curve from the fantail into the shuttle bay doors but I am content to leave it be for now rather than to attempt to rebuild a new part and merge it with the rest of the hull. I'm extremely happy with the look of the finished product. I've done my best to blend the styles used for several of the studio models and the result turned out far more convincingly than I had anticipated.
Engineering hull is all but done. After some additional inspections I'll apply the mirror modifier and add the central shield grid along the ventral.
I'm planning on fleshing out the shuttle bay interior (enough for some exterior shots with the doors open at the very least) but that will wait for now.
With these updates the ship is complete for a first pass. I will probably spend the next week or two revisiting sections of the hull that need tweaking or outright replacing (still not happy with the panels on the pylons as well as one part of the warp nacelles).
Next major stop will be applying textures. This is more new territory for me so I'll be doing some research first.
Fore view
Port view
Aft view
Shuttle bay exterior
Ventral view
Progress views below. Please excuse the chaos of parts coming into and out of reality. I've moved items to layers for organization but frequently did not remember to reactivate the layers for test renders.
That whole rear end was the hardest part of the model when I made it, but you have made it look easy. You've done a bang up job on it making it look very smooth and clean.
That whole rear end was the hardest part of the model when I made it, but you have made it look easy. You've done a bang up job on it making it look very smooth and clean.
Please excuse the very low-quality test-render. I wanted to test the linked rotations for the doors. I am not 100% happy with it yet but I don't want to devote too much time to animation while there is still detailing and ultimately texturing to be done.
I'm about 6 weeks behind on some work related training so updates might be a bit sparse for a bit but I can assure you that progress will continue.
Everything's looking fantastic so far. I'd suggest not having the bay door rotation linear, having it slow in/out with a smooth curve. Aside from that, I don't see a problem.
Everything's looking fantastic so far. I'd suggest not having the bay door rotation linear, having it slow in/out with a smooth curve. Aside from that, I don't see a problem.
That's actually what I ultimately want to achieve.
Right now I'm experimenting with constraints in Blender so I've got all of the leaves tied into one 'master' leaf. While this made animating the opening and closing a breeze, I'm not entirely sure that this is the solution I will use for a production render.
I ultimately intend each leaf to have it's own flow, easing in and out of motion. That will definitely give it a greater sense of size and mass. To do this I think I'll need to animate half of the leaves individually, using copy rotation constraints on the mirror half. That way I can tweak the lead in and lead out for the rotational speed to give them the appropriate 'feel' for inertia weighing them down as they open and close.
Great idea. Having them move individually will make it more visually interesting, especially if they're moving a bit slower to give them that sense of size/weight.
Great idea. Having them move individually will make it more visually interesting, especially if they're moving a bit slower to give them that sense of size/weight.
That's definitely in the works. The current speed is way too fast but it is effective for testing the overall motion. Once the textures are done and the bay fleshed out, I'll redo the sequence for a bee or two to land in the bay with some supplies.
Thank you everyone! Sorry for the lack of updates. As mentioned previously, there's a lot on my plate at the moment so work has slowed but certainly not stopped.
Adding to the delays, I discovered some serious problems with my saucer. Apparently I let the z-axis drift ever so slightly before I spun the segments WAY back at the early weeks of the project. This resulted in artifacts appearing in every radial run of the shield grid and this necessitated cleaning each one by hand. A real mess but it's all better now.
It took me a bit to start to get a hang of the nodes/cycles material process but it's starting to click. I've got images that are further along but those are using an Aztec I downloaded months ago as a placeholder. I don't want to share those images since I can't provide proper citation of the artist contribution. I will post the latest once I've got my own Aztec built.
Can I just say that I have a real hate for the current incarnation of Adobe Illustrator?
I've never been an Illustrator expert but I was able to achieve decent results using versions prior to CS5. The latest Creative Cloud edition has an interface that isn't working well for me at all. I'm starting to get a handle on it so I hope to have the new Aztec finished by next weekend. Fortunately after the saucer, the remainder of the texturing should be a LOT easier to complete.
Regarding Blender materials, I'm using Blender Master Class and Blender Cycles: Lighting and Rendering Cookbook as my primary references. Can anyone recommend any online resources with advice for getting up to speed on Blender materials?
I'm finding that I am becoming distracted by the allure of OSL features. At my current skill level I think it's unlikely to bear fruit for this project.
Oh, and whoever is making the multitude of Star Trek TMP hull fonts has clearly never looked closely at the hyphen used on the hull stencils. The fonts all have a hyphen barely longer than the period but TMP hyphen is three times that length. Stretching that in Illustrator was quite a chore!
Posts
Books: [ Ashes of Alour-Tan | Embers of Alour-Tan ] | Blender Tutorials | Blog
On with the show!
I knew the nacelles would be a challenge; I've tinkered with building them in Max before. They are definitely a vexing collection of interesting geometries. Quite a learning experience, no doubt about that!
As I approach the 'finish line' of the model itself, I've started planning out some things that I would like to do once she is fully realized. Some of these will understandably require additional ships to be modeled (darn!
Fore view
Aft View
Ventral View
Additionally I've already set out to willfully deviate from the studio models in a few key areas (the horror). It's already been commented on that the access panels that I have modeled on the pylons were never anything more than paint/decals on the studio models. Not modeling those was simply not an option. Adding to this list is my treatment of the RCS thrusters on the nacelles. Everywhere else on the ship the RCS pods are visibly separated by a seam, implying modularity and easy of maintenance. I have decided to continue this trend with the nacelle RCS. You can see it here in the pic below.
Nacelle RCS detail:
Purists may cringe but I have to stay true to the spirit of the model and I feel that may occasionally require that I deviate from a 'true' reproduction. Additionally and amusingly still on the subject of the RCS thrusters, I have every intention of lighting them properly. The studio model lit the damn things like they were windows. I feel that they should be lit only when the ship is adjusting attitude. We'll see how it looks when the time comes...
And now the evolution shots. This was the first time that I put some real work into the animated progress shots. From the onset I created three dedicated cameras to record the action. While it did slow the work a bit, it also forced me to evaluate the work from different angels throughout the build.
Fore
Aft
Ventral
Now to finish engineering (including more windows than I care to count), build out the shuttle bay doors, the shuttle bay (maybe... probably
Gigabyte RTX 3080 Gaming OC 12GB
1TB NVMe SSD, 2 x 1GB SATA SSD, 4TB external HDD
32 GB RAM
Windows 11 Pro
Personally, I always like to put a little personal touch on every canon model I do. So, I understand your desire to change a few things here and there.
And having made similar changes to my Ambassador build, I 100% agree with changing the "real" design to fit the canon intent of the design.
Books: [ Ashes of Alour-Tan | Embers of Alour-Tan ] | Blender Tutorials | Blog
Today has been exactly like that. And I think anyone who has modeled details into that damn fantail, no matter the generation of the Star Trek ship, will know that pain all too well.
I've just wasted 12 hours on a faintail that I'm abandoning and rebuilding. I really did a number on myself with the subsurf tool and the end result just isn't keeping up with the rest of the model. I'm going to try to cut off the fat, extrude anew from a cleaned up, unified edge, bevel it and call it a night.
I'm sick of this part and ready for better things :rolleyes:
Fantails and clamshell/overlapping doors are a massive pain.
Books: [ Ashes of Alour-Tan | Embers of Alour-Tan ] | Blender Tutorials | Blog
Just discarded my latest attempt. Turns out the flaws were caused in part by a concavity I hadn't noticed on the fantail. Between that and the non-quad faces that I had ignored, the result is too much of a mess.
I've reverted to the mesh as of Friday <wimper /><sob /> and I'm cleaning it up properly this time.
Fortunately I still have all of my stencils for the grid and windows that I spent yesterday installing :shiner: so I can reapply them once I finish teasing the loop-cuts back into the familiar tapered cylinder we all know and love.
In the end I lose a few days work but the result will still be worth it. And a lesson learned about ignoring best practice for subsurf!
I learned some valuable lessons on this phase of the project.
1.) Always proof your mesh before applying the subsurface modifier. Quads or tris but never both. See my earlier posts for additional details and some overly dramatic self-pity.
2.) If a knife projection isn't working properly, try sliding some of your edges. Previously I had been subdividing faces to get greater resolution. This time I found that sliding only two edges got all of the detail I needed without upping my face count unnecessarily.
3.) Always, always, always inspect the results of bevels on a mesh of any significant detail. I've known that bevels can do weird stuff but this took the cake. I had vertexes thrown miles off course in a few cases. Some I could hand-tease back into shape, others I had to back out entirely and then re-apply in smaller batches.
I still have to add top details, the shuttle bay, and some accents on the bottom before the hull will be feature complete. I'll be adding the central shield grid line on the ventral and fantail grids once I apply the mirror modifier. It cost me nearly a week of evenings to undo my earlier mistakes but the result is better than before so I'm really happy with the outcome.
Very happy with how she's looking at the moment.
Fore view
Port view
Aft view
And the progress gifs. Not as impressive this time since the bulk of the geometry is either already complete or invisible (when done right) but I still like looking at how it flows from start to finish.
Fore
Port
Aft
This isn't strictly be true. Quads, tris, 5-gons, n-gons are all fine for subsurf, but you have to be aware of the impact non-quads have on topology. Non-quads create poles, which behave in predictable, but not necessarily smooth, ways when subdivided.
This is a neat tip! I've had this bite me more than a few times.
Ugh, yes. Usually, this stems from some kind of face/edge/vertex normal irregularity and can be cleaned up, but good grief can it be awful if you don't catch it beforehand.
I know that tune all too well!
Books: [ Ashes of Alour-Tan | Embers of Alour-Tan ] | Blender Tutorials | Blog
Something I found out (by accident, I think) is that selecting the entire mesh and recalculating normals seems to eliminate any problems from beveling. Most of the bevel errors I came across before were edges crossing over one another due to normals that got turned around at some point, possibly from mirroring. That's pretty much why any time I'm getting ready to bevel, I save, select the entire mesh, and hit "Recalculate". I haven't had any bevel problems since I started doing that regularly.
Engineering hull is all but done. After some additional inspections I'll apply the mirror modifier and add the central shield grid along the ventral.
I'm planning on fleshing out the shuttle bay interior (enough for some exterior shots with the doors open at the very least) but that will wait for now.
With these updates the ship is complete for a first pass. I will probably spend the next week or two revisiting sections of the hull that need tweaking or outright replacing (still not happy with the panels on the pylons as well as one part of the warp nacelles).
Next major stop will be applying textures. This is more new territory for me so I'll be doing some research first.
Fore view
Port view
Aft view
Shuttle bay exterior
Ventral view
Progress views below. Please excuse the chaos of parts coming into and out of reality. I've moved items to layers for organization but frequently did not remember to reactivate the layers for test renders.
Engineering dorsal details aft
Shuttle bay high view
Shuttle bay close up
Shuttle bay doors
Yeah, what he said.
Looks great!
Books: [ Ashes of Alour-Tan | Embers of Alour-Tan ] | Blender Tutorials | Blog
Per your request
Please excuse the very low-quality test-render. I wanted to test the linked rotations for the doors. I am not 100% happy with it yet but I don't want to devote too much time to animation while there is still detailing and ultimately texturing to be done.
I'm about 6 weeks behind on some work related training so updates might be a bit sparse for a bit but I can assure you that progress will continue.
That's actually what I ultimately want to achieve.
Right now I'm experimenting with constraints in Blender so I've got all of the leaves tied into one 'master' leaf. While this made animating the opening and closing a breeze, I'm not entirely sure that this is the solution I will use for a production render.
I ultimately intend each leaf to have it's own flow, easing in and out of motion. That will definitely give it a greater sense of size and mass. To do this I think I'll need to animate half of the leaves individually, using copy rotation constraints on the mirror half. That way I can tweak the lead in and lead out for the rotational speed to give them the appropriate 'feel' for inertia weighing them down as they open and close.
That's definitely in the works. The current speed is way too fast but it is effective for testing the overall motion. Once the textures are done and the bay fleshed out, I'll redo the sequence for a bee or two to land in the bay with some supplies.
:thumb:
Adding to the delays, I discovered some serious problems with my saucer. Apparently I let the z-axis drift ever so slightly before I spun the segments WAY back at the early weeks of the project. This resulted in artifacts appearing in every radial run of the shield grid and this necessitated cleaning each one by hand. A real mess but it's all better now.
A shot of some of my progress:
Constitution_Refit_649_001.png
It took me a bit to start to get a hang of the nodes/cycles material process but it's starting to click. I've got images that are further along but those are using an Aztec I downloaded months ago as a placeholder. I don't want to share those images since I can't provide proper citation of the artist contribution. I will post the latest once I've got my own Aztec built.
Can I just say that I have a real hate for the current incarnation of Adobe Illustrator?
I've never been an Illustrator expert but I was able to achieve decent results using versions prior to CS5. The latest Creative Cloud edition has an interface that isn't working well for me at all. I'm starting to get a handle on it so I hope to have the new Aztec finished by next weekend. Fortunately after the saucer, the remainder of the texturing should be a LOT easier to complete.
Regarding Blender materials, I'm using Blender Master Class and Blender Cycles: Lighting and Rendering Cookbook as my primary references. Can anyone recommend any online resources with advice for getting up to speed on Blender materials?
I'm finding that I am becoming distracted by the allure of OSL features. At my current skill level I think it's unlikely to bear fruit for this project.