More holes punched in. No window boxes yet so they just look to the interior of the hull.
After a long break I might be breaking again to model up a new ship, if STOL 2025 happens I do not think I should rely on old assets again. We will see how things work out. AND yes there is a mesh error from more pesky hidden edges autoturning the wrong way.
I can imagine... But all those windows gives her a proper sense of scale.
Just a question about the plating: there is a triangle which the inset looks a bit out of the line (excuse me the bad choice of words, english isn´t my native language), and you can see a tiny green light there too. Was that intentional, or it´s cause the mesh is still being refined? It is located near the end of the first row of windows (counting from top), at the right side.
I'm starting to learn all about mesh errors (except for how to get rid of them). The only problems I see is edges failing to subdivide and curve properly, which I've been struggling with. Seems to happen whenever I try to make one complicated shape. At least with yours, it's only on small, itsy-bitsy details like the sensor emplacements and phaser warning strips!
How many tris does it take to get that much detail? (If it's over 13 million, I'm probably not even going to try to replicate it, because that would crash Blender again, although I do like the separated-hull-plating look.)
polies is 3.5m verts 2.7 so maybe 6m? But that is also everything in scene. SO that includes a galaxy class and a mess of shuttles and 20 or so Xinyan's (genshin player model) This ship is like my Galaxy a bit large cause of the complex curves chew up polys.
Max least the newer ones tend to get a lot of strange surface aberrations. A good example is that one triangular shadow in the top panel near that "grill" there is a dark shadow on the mesh. There was absolutely nothing wrong there but it still had that shadow. Only way to get rid of it was to physically delete all the faces and rebuild it. Thankfully that area is fairly flat and restoring the curve was a simple task of bridging edges laterally. Most of my common mesh errors are what max calls hidden edges. Control wires that shape the poly "skinning" Max always chooses the absolute wrong ways to have these edges flowing on the mesh. SO I have to manually go in and turn them. What is worse newer ver of max will sometimes just RESET them all. So bits of model I had them perfect on will suddenly become a mess.
Anyhow best way I could describe this is these hidden edges and even the edges are the poles in a tent and the polies are the skin of the tent. So if the structure under does not support it right itll wrinkle, fold etc.
Oh and I often call these mesh errors when I guess they are more uh I dunno. . . they are not non planar things or folded up edges.
Ok 2.5m tris 1.5m polys for just the ship and any figures that might be trapped inside.
oh nevermind the galaxy it is just a bit of the onimaru for I dunno what reason now. hahah.
Love seeing updates here... As beautiful as your work is I really enjoy it when you post wireframes and things like that... Its a huge help if you're like me and struggle with modeling..
edit some window boxes, not focusing too well today lol. Sad I am sitting thinking a 24 thread pc is slow now. . .
I’m going to echo @markmassey. Posting the wireframes provides beginner modelers like myself a wealth of information, and gifts me with alternative ideas of how to model in a way that will provide for a detailed finished product like yours.
A quick glance reveals that you’ve separated even the main the hull into different objects, with each color code denoting a different object. Looks like the hull and hoop are different objects, rather than the same.
I’ll have to analyze it more later after work. I’m starting to think that I should redo my Sovereign primary hull with the raised back section as a separate object.
Thanks everyone. BTW Atolm deserves the cred for the overall form and shape of this.
The hull and hoop were one piece. I break it up to help in managing all the parts and details. I also break it up to aid in mapping as well, if you will notice the hull grids are broken up in a checker pattern this is so I can fool the eye a bit more by having the mapping pattern change between the panels. In the past breaking panels off was mostly to cheat the smoothing but in max that isnt that important with "smoothing groups"
however this model is probably not the best to judge how to break up parts in sets as it is mostly one large subdiv object due to the organic nature and unibody like design. But from what I saw on your model I would def break parts up as separate units along hard edges then if you really want a chamfer or rounded edge model that in outside of subdiv. Kinda like how I will merge objects into the bridge BC decks on TOS saucers. I also did a lot of this on the Balmung, merging in all the superstructure decks mostly post the sundiv work. It also does depend on the client more these days too as I see alot of things blender can one click that max needs manual work. But then you lack per poly smoothing groups. (least last time I looked)
Here are some caps of the base control mesh. It might have some older shapes in it as this is just some mesh in the junk folder in the scene.
just in case they are not clicky just drag the image to a new tab or in chrome open image in new tab to see full size. Keep in mind the loops on corners remain tight cause they have a separate smoothing group from the main body. In apps not able to smooth this way youll need multiple loops or others effects I think I have heard blender has some other modifier you can apply to corners to control the bias of a corner.
ah a lot of what I said is irrelevant as backstept and a few others mention these tools.
This helps a lot, and confirms what I was thinking of trying! The back raised section of the Sovereign will be better modeled split as a separate object, just as you have done with the Atheling class's bow and center decks. This will allow me to cut down on the number of subdivisions, and keep the shape with less tris (which I'm going to need for later details). Seems that you make each level of the "command section" (the raised decks on the primary hull) also as separate objects.
It might take a bit of experimentation, especially since I'll be activating and using a new tool for the corners, but I have high hopes that it get me much closer to the results that I'm aiming for!
Thanks Madkoifish (for being an excellent modeler, and a huge help!)!
Def take a look at early wip wires too, and keep in mind some decks look seprate but were just detached in the cleanup and detailing phase. Sub-div just takes a bit of time to see what can or aught be part of one mesh or another. Some of the more complex shapes I will actually overlap meshes wi5th the intent of deleting those overlaps later on. A good example of this is my Mughi cargo ship. The nacelles would be impossible without this style of modeling. Well not impossible but a PITA. Best thing to do is keep the base mesh for the subdiv as low poly as possible, just nave enough to get the form to work at say 2x or 3X division.
As you might have noticed, the saucer section is a mess of subdivisions. After considering maybe removing subdivisions, which would theoretically make separating the raised aft end, the forward saucer, the stacked ventral sections/trench, and impulse drives into different objects easier (although I do kind of like how the shapes flow between the impulse drives and the saucer), I decided it would just be easier to just start over, using the current saucer section as a reference for the magnet tool.
It's going to be at least some trial and error from here on, as I figure out the best configuration (or at least what works adequately).
yup, trial and error. As you see what works you'll pick up how things need to flow. That saucer is def a mass of converging shapes.
ok the update, that was. . . well, was suppose to be posted at 7 last night. oops
Inner hull windows laid out. I will need to so a final alignment check pass on them as the hull here even with as much as I did to keep the curvatures consistent still has a taper making for the need to manually place the windows and a altering curvature long the horizontal. Some will never match op as I do not want the interior window shapes to "slope" Think of windows angles to go up a staircase.
and yeah, a bit dark I need to adjust my HDRI again I suppose. I have been considering just using a super lit indoor like scene for wips.
Posts
yet more holes. . . .
After a long break I might be breaking again to model up a new ship, if STOL 2025 happens I do not think I should rely on old assets again. We will see how things work out. AND yes there is a mesh error from more pesky hidden edges autoturning the wrong way.
I can imagine... But all those windows gives her a proper sense of scale.
Just a question about the plating: there is a triangle which the inset looks a bit out of the line (excuse me the bad choice of words, english isn´t my native language), and you can see a tiny green light there too. Was that intentional, or it´s cause the mesh is still being refined? It is located near the end of the first row of windows (counting from top), at the right side.
Interesting addition!
How many tris does it take to get that much detail? (If it's over 13 million, I'm probably not even going to try to replicate it, because that would crash Blender again, although I do like the separated-hull-plating look.)
Max least the newer ones tend to get a lot of strange surface aberrations. A good example is that one triangular shadow in the top panel near that "grill" there is a dark shadow on the mesh. There was absolutely nothing wrong there but it still had that shadow. Only way to get rid of it was to physically delete all the faces and rebuild it. Thankfully that area is fairly flat and restoring the curve was a simple task of bridging edges laterally. Most of my common mesh errors are what max calls hidden edges. Control wires that shape the poly "skinning" Max always chooses the absolute wrong ways to have these edges flowing on the mesh. SO I have to manually go in and turn them. What is worse newer ver of max will sometimes just RESET them all. So bits of model I had them perfect on will suddenly become a mess.
Anyhow best way I could describe this is these hidden edges and even the edges are the poles in a tent and the polies are the skin of the tent. So if the structure under does not support it right itll wrinkle, fold etc.
Oh and I often call these mesh errors when I guess they are more uh I dunno. . . they are not non planar things or folded up edges.
Ok 2.5m tris 1.5m polys for just the ship and any figures that might be trapped inside.
oh nevermind the galaxy it is just a bit of the onimaru for I dunno what reason now. hahah.
this is that shadow whatever thing I mentioned.
edit some window boxes, not focusing too well today lol. Sad I am sitting thinking a 24 thread pc is slow now. . .
That's a sexy angle...
I’m going to echo @markmassey. Posting the wireframes provides beginner modelers like myself a wealth of information, and gifts me with alternative ideas of how to model in a way that will provide for a detailed finished product like yours.
A quick glance reveals that you’ve separated even the main the hull into different objects, with each color code denoting a different object. Looks like the hull and hoop are different objects, rather than the same.
I’ll have to analyze it more later after work. I’m starting to think that I should redo my Sovereign primary hull with the raised back section as a separate object.
The hull and hoop were one piece. I break it up to help in managing all the parts and details. I also break it up to aid in mapping as well, if you will notice the hull grids are broken up in a checker pattern this is so I can fool the eye a bit more by having the mapping pattern change between the panels. In the past breaking panels off was mostly to cheat the smoothing but in max that isnt that important with "smoothing groups"
however this model is probably not the best to judge how to break up parts in sets as it is mostly one large subdiv object due to the organic nature and unibody like design. But from what I saw on your model I would def break parts up as separate units along hard edges then if you really want a chamfer or rounded edge model that in outside of subdiv. Kinda like how I will merge objects into the bridge BC decks on TOS saucers. I also did a lot of this on the Balmung, merging in all the superstructure decks mostly post the sundiv work. It also does depend on the client more these days too as I see alot of things blender can one click that max needs manual work. But then you lack per poly smoothing groups. (least last time I looked)
Here are some caps of the base control mesh. It might have some older shapes in it as this is just some mesh in the junk folder in the scene.
just in case they are not clicky just drag the image to a new tab or in chrome open image in new tab to see full size. Keep in mind the loops on corners remain tight cause they have a separate smoothing group from the main body. In apps not able to smooth this way youll need multiple loops or others effects I think I have heard blender has some other modifier you can apply to corners to control the bias of a corner.
ah a lot of what I said is irrelevant as backstept and a few others mention these tools.
https://bolidecascade.files.wordpress.com/2019/08/tsukuru-nx-0571.png
https://bolidecascade.files.wordpress.com/2019/08/tsukuru-nx-0572.png
https://bolidecascade.files.wordpress.com/2019/08/tsukuru-nx-0569.png
https://bolidecascade.files.wordpress.com/2019/08/tsukuru-nx-0570.png
https://bolidecascade.files.wordpress.com/2019/08/tsukuru-nx-0573.png
https://bolidecascade.files.wordpress.com/2019/08/tsukuru-nx-0574.png
I do need to update the wip blog with a new wires post and all the stuff done in the last few yrs.
This helps a lot, and confirms what I was thinking of trying! The back raised section of the Sovereign will be better modeled split as a separate object, just as you have done with the Atheling class's bow and center decks. This will allow me to cut down on the number of subdivisions, and keep the shape with less tris (which I'm going to need for later details). Seems that you make each level of the "command section" (the raised decks on the primary hull) also as separate objects.
It might take a bit of experimentation, especially since I'll be activating and using a new tool for the corners, but I have high hopes that it get me much closer to the results that I'm aiming for!
Thanks Madkoifish (for being an excellent modeler, and a huge help!)!
It's going to be at least some trial and error from here on, as I figure out the best configuration (or at least what works adequately).
ok the update, that was. . . well, was suppose to be posted at 7 last night. oops
Inner hull windows laid out. I will need to so a final alignment check pass on them as the hull here even with as much as I did to keep the curvatures consistent still has a taper making for the need to manually place the windows and a altering curvature long the horizontal. Some will never match op as I do not want the interior window shapes to "slope" Think of windows angles to go up a staircase.
and yeah, a bit dark I need to adjust my HDRI again I suppose. I have been considering just using a super lit indoor like scene for wips.