I never really thought about the back of the Excelsior neck being curved, but it seems to be concentric with the saucer.
Hm, interesting. I seem to remember it being straight on the model kit I put together back in the 90s. But then, those kits weren't always the most accurate. It's straight on the Eaglemoss version too, I have the XL edition of the Enterprise-B. You may have found one of those little known tibits of knowledge about the ship.
Nice! The Excelsior is by far one of the best designs in Star Trek, in my opinion.
Starfleet keeping her in service for 100 plus years would agree!
New TV show, we need ships. Of course, it would have been preferred to not use any of the older ship models in TNG, but there wasn't enough room in the budget to build all new ship models, that's why the Excelsior, Oberth, Miranda, Klingon Bird of Prey and Merchantman ship models make numerous appearances. (though the Merchantman was constantly being modified.) That means many of those ships were in service for that long. Though, the real money saver was the K't'inga. Between TMP and Star Trek VI, no new footage was shot of that model. Star Trek II and TNG used stock footage from TMP.
Fun little fact, as a result of this, and the use of stock footage from EAF all the way through TNG, the most frequent version of the Excelsior seen in trek before CG was the search for spock version. The Excelsior model didn't undergo it's refit into the TUC version until season 7 of TNG, but TNG stopped producing new footage of the Excelsior by season 4, so all the stock shots were the TSFS version with the different bridge and shuttle modules.
Indeed, the producers had ILM do a bunch of stock shots of the Excelsior when Encounter at Farpoint was in pre-production and they recycled them throughout the series. I think they may have done a few shots at Image G, but not many. Mostly when you see the Excelsior class in TNG, it's from ILM.
I never really thought about the back of the Excelsior neck being curved, but it seems to be concentric with the saucer.
That’s where a tiny docking port needs to be really…not fouling a TMP-era bridges gentle teardrop that should terminate with a wide upward firing escape pod mirroring the TOS turbolift.
So help me…but the heels of some shoes look like that.
Here is one approach to the ribbing—-explosive supports:
I never really thought about the back of the Excelsior neck being curved, but it seems to be concentric with the saucer.
Hm, interesting. I seem to remember it being straight on the model kit I put together back in the 90s. But then, those kits weren't always the most accurate. It's straight on the Eaglemoss version too, I have the XL edition of the Enterprise-B. You may have found one of those little known tibits of knowledge about the ship.
Here's a great shot from during the Generations remodel
I've been wanting to do an Excelsior class, but one that has been refit w/ TNG technologies. It's not like there isn't historical precedent where the hull configuration remains mostly unchanged for decades on the surface, but there are actually major changes due to new technologies.
So I see a retcon where all of the Excelsiors that appear in TNG have the same general hull configuration, but with collimated phaser arrays, TNG life pod hatches, color schemes similar to the Ambassador class, and with nacelles maybe more like the USS Balmung, instead of the Linear Warp Nacelles.
But for now, the project I'm working on is keeping me plenty busy.
I never really thought about the back of the Excelsior neck being curved, but it seems to be concentric with the saucer.
Hm, interesting. I seem to remember it being straight on the model kit I put together back in the 90s. But then, those kits weren't always the most accurate. It's straight on the Eaglemoss version too, I have the XL edition of the Enterprise-B. You may have found one of those little known tibits of knowledge about the ship.
Here's a great shot from during the Generations remodel
That's really cool. Also, your model is looking great so far.
I've run into the roadblock of too many tris, and as of now, I find myself having to redo the saucer I was working on, because it was taking up too much memory! The maximum I could go was 14,700,000 (roughly), before the program became unresponsive.
I have decided to try redoing the who thing over with less vertices, and trying to add more loop cuts to get closer to the shape as I go, even though it will take longer before it will look presentable like yours.
For reference, I have 16 MB. I have also posed a similar question of the Blender forum on Reddit, and have gotten mentions of optimization.
What do you use to get your models to look so smooth and detailed? I am considering seeing if I can take my computer in to get extra memory? Is this necessary? Or should look into Voxel or Retopro (these were mentioned by someone on Reddit).
I haven't made a note of my recent models, but I think my Constitution is a couple million.
I usually get the general shape using as few polys as possible and add a subd set at 2 viewport, 3 render. Bevels I don't often do more than 4 segments unless I'm making a larger curve with it. I try to stick with only quads but ngons and tris are sometimes unavoidable.
I find it mindblowing that all of that, you did with only a few million polys. I have so much to learn.
The biggest thing is figuring out how to budget your polygons. IE: What absolutely needs to be as smooth as possible and where can you maybe shave off some polgyons? So, if you have a straight cylinder and you're set to smooth, how many polygons does it really need to be? If you aren't going to see the ends, because of how it connects to other pieces, it doesn't need to have 96 sides to appear smooth. It can probably get by with closer to 24, or possibly even 12, depending on the circumference and how close you plan to get to it. Doing bevels and edge rounding is another place where you don't necessarily need to go with gobs of polygons. It just really all depends on how close you plan to get to the model and how visible the part will be in most circumstances. For close shots, you can always do a section at higher polygons just for that shot, but not necessarily have the whole model be that high.
Like backstept said, a million used to be unimaginable. Rendering over a decade ago with older processors, particularly with less cores, was a chore and a half, so we learned how to skimp in polygons wherever we could. With modern computers and things like GPU compute to render, it's easy to just add polygons without thinking about it, but some programs do have limits to how many polygons can be in a scene, as you noticed. I actually tend to go on the low side on some objects, sometimes too low. I was building a part recently where I had gone too low, so I hit it with another round of Sub D. But that's easier to do than taking out polygons when you build with too many. Some software has tools for this, but I find they sometimes make a mess of a mesh.
I find it mindblowing that all of that, you did with only a few million polys. I have so much to learn.
Just got around to checking, my USS Lexington is 6,935,299 faces. Though most of that (4,190,259) is in the shrinkwrapped markings like the saucer registry, which I probably have subdivided far too much! I just wanted the registry to lay on the hull smoothly and conform to the panel lines. I think my next task will be to heavily optimize the registry.
The ship itself is 2,745,040 faces.
Posts
Hm, interesting. I seem to remember it being straight on the model kit I put together back in the 90s. But then, those kits weren't always the most accurate. It's straight on the Eaglemoss version too, I have the XL edition of the Enterprise-B. You may have found one of those little known tibits of knowledge about the ship.
Indeed, the producers had ILM do a bunch of stock shots of the Excelsior when Encounter at Farpoint was in pre-production and they recycled them throughout the series. I think they may have done a few shots at Image G, but not many. Mostly when you see the Excelsior class in TNG, it's from ILM.
That’s where a tiny docking port needs to be really…not fouling a TMP-era bridges gentle teardrop that should terminate with a wide upward firing escape pod mirroring the TOS turbolift.
So help me…but the heels of some shoes look like that.
Here is one approach to the ribbing—-explosive supports:
http://red-shoulders.blogspot.com/2019/01/redesigning-excelsior_29.html?m=1
Similar to Jayru’s
https://www.trekbbs.com/threads/jayru-jsnaiths-3d-trek.311502/page-12
Here's a great shot from during the Generations remodel
Got a little bit done today.
So I see a retcon where all of the Excelsiors that appear in TNG have the same general hull configuration, but with collimated phaser arrays, TNG life pod hatches, color schemes similar to the Ambassador class, and with nacelles maybe more like the USS Balmung, instead of the Linear Warp Nacelles.
But for now, the project I'm working on is keeping me plenty busy.
That's really cool. Also, your model is looking great so far.
And to me as well!
I've run into the roadblock of too many tris, and as of now, I find myself having to redo the saucer I was working on, because it was taking up too much memory! The maximum I could go was 14,700,000 (roughly), before the program became unresponsive.
I have decided to try redoing the who thing over with less vertices, and trying to add more loop cuts to get closer to the shape as I go, even though it will take longer before it will look presentable like yours.
For reference, I have 16 MB. I have also posed a similar question of the Blender forum on Reddit, and have gotten mentions of optimization.
What do you use to get your models to look so smooth and detailed? I am considering seeing if I can take my computer in to get extra memory? Is this necessary? Or should look into Voxel or Retopro (these were mentioned by someone on Reddit).
I usually get the general shape using as few polys as possible and add a subd set at 2 viewport, 3 render. Bevels I don't often do more than 4 segments unless I'm making a larger curve with it. I try to stick with only quads but ngons and tris are sometimes unavoidable.
Like I remember Battlestar Galactica and hearing how many polygons the ship Zoic made had and I thought "I'll never make something that big!"
It's easy to get high poly counts with subd. More important is being able to figure out how to build your model to use as few polygons as possible.
The biggest thing is figuring out how to budget your polygons. IE: What absolutely needs to be as smooth as possible and where can you maybe shave off some polgyons? So, if you have a straight cylinder and you're set to smooth, how many polygons does it really need to be? If you aren't going to see the ends, because of how it connects to other pieces, it doesn't need to have 96 sides to appear smooth. It can probably get by with closer to 24, or possibly even 12, depending on the circumference and how close you plan to get to it. Doing bevels and edge rounding is another place where you don't necessarily need to go with gobs of polygons. It just really all depends on how close you plan to get to the model and how visible the part will be in most circumstances. For close shots, you can always do a section at higher polygons just for that shot, but not necessarily have the whole model be that high.
Like backstept said, a million used to be unimaginable. Rendering over a decade ago with older processors, particularly with less cores, was a chore and a half, so we learned how to skimp in polygons wherever we could. With modern computers and things like GPU compute to render, it's easy to just add polygons without thinking about it, but some programs do have limits to how many polygons can be in a scene, as you noticed. I actually tend to go on the low side on some objects, sometimes too low. I was building a part recently where I had gone too low, so I hit it with another round of Sub D. But that's easier to do than taking out polygons when you build with too many. Some software has tools for this, but I find they sometimes make a mess of a mesh.
Aye. And if my grandmother had wheels, she'd be a wagon.
(it's a tragically underrated movie)
II, III and IV make for a nearly perfect TOS film trilogy. They really do.
Ok, that's funny.
Various Work: U.S.S. Constellation - Matt Jefferies Concept Shuttle
Just got around to checking, my USS Lexington is 6,935,299 faces. Though most of that (4,190,259) is in the shrinkwrapped markings like the saucer registry, which I probably have subdivided far too much! I just wanted the registry to lay on the hull smoothly and conform to the panel lines. I think my next task will be to heavily optimize the registry.
The ship itself is 2,745,040 faces.
Agreed.