Has anyone taken any of the 3D assets from the game and turned them into STL files so we can 3D Print them, make little dioramas and such? If so please share them, would love to do some models.
-
We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
As someone who does 3D CAD modelling for a living, leaving aside the legal issues with doing that(of which there is a significant amount, but I will leave other to explain that), I can assure you that on a practical level, that would require a vast amount of work. You cannot simply port a file in and have it be 3D printable.
Models for computer games and models for 3D printing are related but only distantly, as they work to entire different criterion. I would not be able to make you a CAD model for a game, for example (I do not know thing about rigging or texturing and probably some other things), and I suspect that the other direction might be easier, than the graphics design at PDX might have some difficulty making a 3D printed model without a good deal of trail and error, because unless they happen to be doing it on the side as a hobby, they won't know the magic numbers). They are, ultimately, different disiplines.
Game models do not have to be water-tight, for a kick-off. Polygon limits tend to be more of a thing on game models, and while the issue is likely less than is was with today's graphics, you would be surprised, for example, how if you look closely, you will see how many round shapes aren't, especially in older games, but actually hexagonal or something, with the illusion of a round surface coming from the texturing.
(Texturing is even a whole seperate thing I wouldn't even know where to begin one.)
Secondly, and vastly more pertiently, 3D printing has the limitations of minimum wall thickness. This has to be absolute and if you're using commerical print-houses, usually has to be on the order of 1mm; home printing can get away with less, but there is also the issues of support (extrusion printers like home printers cannot print over open space) and the removal of such, and the fact that anything less than 1mm can be extremely fragile.
Some printers CAN print to VERY fine levels - my Dad bought an N-scale railway man who was carrying a hoop about 0.5-0.5mm thick or something - but those printers are sufficiently small and expensive that you couldn't even use them for my 144th scale tanks or my starships.
Now, I HAVE actually done some CAD modelling from game files - I really wanted some of the ships from TIE Fighter to play with my old micro-machines SW game that I wrote. (And I did this purely for my own game, I would never even consider selling them.) What that actually entailed was essentially having to re-build the models from scratch, with the old .dxf file being principally only there for a guideline, as evern something as simple as TIE Fighter's meshes didn't work. (Heck, I actually used a lot of fanart, in the end, and did a lot of it by eye, using the same skillset as I do when doing my modern armour models!)
So you would very likely have to essentially re-build the model from scratch - after first deciding exactly what size you wanted it all at, since that determines the absolute minimum thickness of any small parts. Because you cannot properly even scale a 3D print model simply by just selecting the model and going "scale by n" - not if you want to garentee it will work and/or not look pants.
You can scale a model up (e.g. 144th-1/72nd), relatively easily (you ate least know that all the parts will meet minimum thickness, if they look a bit pants if you don't reduced the relatively oversized parts to the absolute mnimum thickness), but scaling down (e.g. 144th to 288th) requires essentially rebuilding the model, since aside from stuff now being below the minimum thickness, you have to start making decisions about parts you have to make abstractions on. Case in point: at 144th, I can model a tank's machine guns, a little bit oversized, admittedly, but as seperate parts you can turn. At 288th, those same guns would have to be the exact same size to meet the absolute minimum printable wall thickness (and are thus double the relative size they were before) or you would have to model them basically attached to the hull and as surface details, they couldn't be the right height, either.
In addition, if you are doing a model that is expected to be printed on a commerical print house, it has to be hollowed out to reduce the volume cost, and if it is to be printed on a home extrusion printer, it has to be solid (since the software itself fills a solid shape with a honeycomb). Hollowing can be a long and difficult job unless you have stupid expensive industry-standard software like Rhino (which my Dad has on his lap top only because he once worked at Rolls Royce), especially for complex shapes.
Hopefully, that will give everyone some idea. (You'll forgive me if I lecture, but it is, as they say, my particular area of expertise!)
Models for computer games and models for 3D printing are related but only distantly, as they work to entire different criterion. I would not be able to make you a CAD model for a game, for example (I do not know thing about rigging or texturing and probably some other things), and I suspect that the other direction might be easier, than the graphics design at PDX might have some difficulty making a 3D printed model without a good deal of trail and error, because unless they happen to be doing it on the side as a hobby, they won't know the magic numbers). They are, ultimately, different disiplines.
Game models do not have to be water-tight, for a kick-off. Polygon limits tend to be more of a thing on game models, and while the issue is likely less than is was with today's graphics, you would be surprised, for example, how if you look closely, you will see how many round shapes aren't, especially in older games, but actually hexagonal or something, with the illusion of a round surface coming from the texturing.
(Texturing is even a whole seperate thing I wouldn't even know where to begin one.)
Secondly, and vastly more pertiently, 3D printing has the limitations of minimum wall thickness. This has to be absolute and if you're using commerical print-houses, usually has to be on the order of 1mm; home printing can get away with less, but there is also the issues of support (extrusion printers like home printers cannot print over open space) and the removal of such, and the fact that anything less than 1mm can be extremely fragile.
Some printers CAN print to VERY fine levels - my Dad bought an N-scale railway man who was carrying a hoop about 0.5-0.5mm thick or something - but those printers are sufficiently small and expensive that you couldn't even use them for my 144th scale tanks or my starships.
Now, I HAVE actually done some CAD modelling from game files - I really wanted some of the ships from TIE Fighter to play with my old micro-machines SW game that I wrote. (And I did this purely for my own game, I would never even consider selling them.) What that actually entailed was essentially having to re-build the models from scratch, with the old .dxf file being principally only there for a guideline, as evern something as simple as TIE Fighter's meshes didn't work. (Heck, I actually used a lot of fanart, in the end, and did a lot of it by eye, using the same skillset as I do when doing my modern armour models!)
So you would very likely have to essentially re-build the model from scratch - after first deciding exactly what size you wanted it all at, since that determines the absolute minimum thickness of any small parts. Because you cannot properly even scale a 3D print model simply by just selecting the model and going "scale by n" - not if you want to garentee it will work and/or not look pants.
You can scale a model up (e.g. 144th-1/72nd), relatively easily (you ate least know that all the parts will meet minimum thickness, if they look a bit pants if you don't reduced the relatively oversized parts to the absolute mnimum thickness), but scaling down (e.g. 144th to 288th) requires essentially rebuilding the model, since aside from stuff now being below the minimum thickness, you have to start making decisions about parts you have to make abstractions on. Case in point: at 144th, I can model a tank's machine guns, a little bit oversized, admittedly, but as seperate parts you can turn. At 288th, those same guns would have to be the exact same size to meet the absolute minimum printable wall thickness (and are thus double the relative size they were before) or you would have to model them basically attached to the hull and as surface details, they couldn't be the right height, either.
In addition, if you are doing a model that is expected to be printed on a commerical print house, it has to be hollowed out to reduce the volume cost, and if it is to be printed on a home extrusion printer, it has to be solid (since the software itself fills a solid shape with a honeycomb). Hollowing can be a long and difficult job unless you have stupid expensive industry-standard software like Rhino (which my Dad has on his lap top only because he once worked at Rolls Royce), especially for complex shapes.
Hopefully, that will give everyone some idea. (You'll forgive me if I lecture, but it is, as they say, my particular area of expertise!)
Not talking about selling, I am talking about doing something like Thingiverse for fans to make models for themselves.
Still a bit dubious, frankly.
Also doesn't change the fact you would have to find someone VERY keen to spend the tens of hours (maybe plus) to do it for the sake of doing it. (I certainly wouldn't sink that kind of time for nowt; hell, I couldn't AFFORD to.) I mean, you're looking at, I imagine, the sort of hours it takes for making a full conversion mod, only for an even smaller audience.
(I mean, crap, I 'm pretty sure that I know of people that write fanfic that are on Pateron that make more than I do...!)
I suspect that you'll find that there simply isn't likely to be anyone around who is a) that keen, b) has the requisite skills and c) the time to do it, since it is, as I am trying to emphasise a major and difficult undertaking. (Even aside the other issues.)
And yet we see tons of 3D STL models for various games come out all the time. We see a ton for Battletech and Star Trek, tons of SpaceX models.
Both of which have well-established fanbases, one of which is literally centred around a tabletop wargame and Star Trek has a fair fanbase of starship gamers for decades (because of Star Fleet Battles, fo one) and had a thriving unoffical model market long before 3D printing (and there are a number of places where people sell models of that stuff, too, for that matter). So it's not surprising, really, that you can find models for them, new BattleStar Galactica and Star Wars relatively easily, because they are big and well-known franchises with brand recognition.
(To the point that most of the forums about starshop gaming are almost entirely dominated by Star Trek, Star Wars and BSG.)
Surviving Mars just isn't that big, nor a wargame, so the chances of finding that one guy (let alone more than one) that keen in the fanbase with the right skills is pretty small. I didn't say it was impossible, just very unlikely. As I say, it's non-trivial job.
(To the point that most of the forums about starshop gaming are almost entirely dominated by Star Trek, Star Wars and BSG.)
Surviving Mars just isn't that big, nor a wargame, so the chances of finding that one guy (let alone more than one) that keen in the fanbase with the right skills is pretty small. I didn't say it was impossible, just very unlikely. As I say, it's non-trivial job.
You can generate STL files your self. It takes less than 5 min to convert game model to STL in blender/max/maya. Just add supports in printing software and you are done. Ofcourse for better quality and easier print you should split the model to separate parts and orient them properly, this can avoid supports and make nicer surfaces.
No offense to lecture above, but you can easily convert any game model without extensive re-modeling.
No offense to lecture above, but you can easily convert any game model without extensive re-modeling.
Converting is not necessarily the same as getting a model that will print (aforementioned wall thicknesses being the main issue, as you might be able to get around with a mesh fixer). (Double-especially if you have to use a print-house f you don't have a printer yourself.)
But if you've actually done it, eithy, I would be very interested (and I don't mean this sarcastically or faceiously anything, but genuinely; as we have established, this stuff is of professional and personal interest to me) to see how it came out and what sort of settings you used on the print (and what sort of printer).
(I use a Replicator 2, now armed with Simplify 3D, which was worth the money to buy as it was like buying a new printer.)
So first try, still hot from printer. I did not touch the model, straight from game, no editing.
That is "TOTALLY COOL"... Wish I had the software or understood how to do this. I can do basic work in a slicer with an STL but the rest escapes me.
Last edited:
Huh. Okay, I'm impressed.
So, my burning questions must be what printer/slicer you're using, and how big did you make the model?
(I work right down at 144th scale and starship scale, where you are close to the borders of wall thickness.)
I guess you must be able to unpack with some, like modding-related software to get the model files (I'm curious now) and then convert to .stl?
It's 7 cm high and 6 cm diameter. Printed with Prusa MK3, it can handle up to 0.5 mm wall thickness and 0.05 mm layers.
They use propriertary binary model format, so I just grabed it with RenderDoc instead.
They use propriertary binary model format, so I just grabed it with RenderDoc instead.
Interesting. Bit better than ours, but the sort of stuff we have looked at recently has been the bottom end of the market. The Rep 2 will do stuff like 0.5mm thin in places, but usually the problem it PLA is too fragile (I sometimes have a bad day with some of the 1mm stuff); commercial places tend to have fairly beefy requirements.
Attachments
You should capure vertices on vertex shader output and unproject them back to world space. This way it will include bone transformations to exact pose you see when you capture frame. As you can see, that drone is in T-pose and the generator i printed would be closed.
Ah, thanks was wondering about that. How do I unproject them back into world space?
I am just wanting to print some stuff to make a Diorama for my sons space club at high school 
Well, download and install Blender and RenderDoc
install this plugin in Blender (save user settings to keep it active next time you start blender)
https://github.com/sbobovyc/GameTools/blob/master/Blender/import_pix.py
You'll need to start SM from renderdoc, so you can capture with it (I don't think it'll work with a steam copy since it does that steamurl starting stuff, gog or paradox launcher)
Check out that video for what to look for in the capture
Then we'll both see about non-t posing
install this plugin in Blender (save user settings to keep it active next time you start blender)
https://github.com/sbobovyc/GameTools/blob/master/Blender/import_pix.py
You'll need to start SM from renderdoc, so you can capture with it (I don't think it'll work with a steam copy since it does that steamurl starting stuff, gog or paradox launcher)
Check out that video for what to look for in the capture
Then we'll both see about non-t posing
