Jump to content
Sign in to follow this  
unfrail

3d Modeling Question

Recommended Posts

So with the latest update on the new website about the SAW development process, I am wondering whether that is the ideal method for developing weapons models.

I own a hex reciever Mosin and I was going to model that in 3ds max but I wasnt sure how to handle assemblies and animations.

Do i model each individual component of the bolt? or only parts that visibly move independently?

How do I assure that my model is useable for animations?

Are all animations handled within UE4 or whe I develop a weapon model do I need to develop animation maps as well?

No need for an immediate response, I am admittedly still at the "talking about it" phase, but I am binging on my 3dCAD apps at the moment for school and I am trying to find workflow similarities so I know where to start in 3dsMax.

much thanks.

Share this post


Link to post
Share on other sites

Do i model each individual component of the bolt? or only parts that visibly move independently? - You have to think ahead of every possible animation and angle you may see the object from and then you can decide what you need to do and what you don't need to do. In general you would only need to model visibly independent parts. I personally like to model almost everything, that way I don't have to re-do a mountain of work if we ever want something done differently or if we want an asset to be able to do a new thing.

How do I assure that my model is useable for animations? - Same answer as above, think ahead. Look at videos. what parts are moving when you reload? Does it need to have adjustable sights? What are you looking at when you are reloading?

Are all animations handled within UE4 or whe I develop a weapon model do I need to develop animation maps as well? The animations are always created in 3ds max, maya ect and then exported to UE4, only a few things like procedural animations are handled outside of a 3d modeling environment.

 

Added some quick notes after your questions.

Share this post


Link to post
Share on other sites

If you are new to Max, I would recomned doing some tutorials and learning the basics of the program and tools and not to do a weapon for your first go, and like Z-trooper has said, for animations, it will depend on what is moving and what you're planing on doing with the model, if you want to animate you weapon fully stripped to down to it's core, then yes, you will make everything, but if you want a weapon that will fuction in a game, then only the outside moving parts. And remember, depedning on the detail and where it shows for the player in first person, you may want to use a normal map or texture detail (things like serial numbers, small screws, names etc.)

 

This is a WIP that I am doing for PR:WW2 (Turbo smooth is off):

https://sketchfab.com/models/cdead088d73b460492db4ef648ce0620

 

Model is almost done, with minor things that I will I still have to clean and modify. The grip detail on the knobs will be baked in to a normal map, and applied. The screws on the mount will be also be normal mapped and textured to redeuce the tri count, while the screws on the scope it self will be kept as they will be close enough to the player for this detail to be nessacry (at least how I see it)

Share this post


Link to post
Share on other sites

Thanks for the replies gentlemen.

I am thinking I am going to model the gun, but I intend to start with the bolt-assembly, which I can break down into its individual parts. I realise that this might be a bit overkill in terms of modeling, but if I endup making components that are un-necessary or always hidden they can be simply omitted from the final assembly. Moreover these late 19th century parts are rather crudely hewn individually, though as an assembly they work fantastic. Modeling the individual pieces will be essentially my learning tool, starting with the threaded firing pin, and then a spring, and then the extractor etc.

In terms of planning bump-maps for optimization, do I model all the features and then select certain elements to be converted to bump maps, or do I need to omit features and reserve them for bump maps?

Also, I intend to build the model and then save texturing for the end. Do I need to take into consideration the texturing when I assemble my geometries?
 

This is a WIP that I am doing for PR:WW2 (Turbo smooth is off):

https://sketchfab.com/models/cdead088d73b460492db4ef648ce0620

 

Thats a nice looking gun component my friend. Should I draft my circles into x-number of faces as I construct or is that something I can define/edit later on? Thanks again for all the info. Look forward to posting some progress.

Share this post


Link to post
Share on other sites

In terms of planning bump-maps for optimization, do I model all the features and then select certain elements to be converted to bump maps, or do I need to omit features and reserve them for bump maps?

I'm not a highly-skilled weapons/character artist like others here, but I can still answer:

 

You actually do both. You make a highpoly model, as well as a lowpoly model. The lowpoly model is the one that the game uses, and it bears a normal map that gives it the illusion of having more detail than it really does. The highpoly model is used only to generate that normal map.

Share this post


Link to post
Share on other sites

Ok, so just make the model high-fidelity, and then some of the non-functional geometry can be converted directly into the bump map.

 

Good to know. Next thing I post will be progress. Less talk more rock.

Share this post


Link to post
Share on other sites

First thing u need to know is what will you use your model for.

 

If real time rendering is not required (its not a game) then you can go ahead and and produce every litle piece of geometry to exquisite detail. Go for the many milhons of polygons. Make the most detailed model possible. Create every little piece. This is specially true if you just want to have the model for yourself.

 

People who work with 3D animations and special effects will only produce this detail for those pieces geometry that will be clearly visible in the sciene. They dont enact that level of labor on things that will be far away, or moving fast, or shown for less than a second (Theese are surprisingly crude). Parts of the geometry that wont be visible are not produced at all. But if u are modeling for the fun of it, u can do every little bit.

 

If u r going for real time rendering (like a game or other interactive stuff) then rendering performance becomes an issue.

 

For rendering performance:

        - Poly count (triangle count really, there are no polys, only triangles) should go to a minimum.

        - You should use only one material per mesh, if u can (meshes will be rendered at least one time per material).

        - If u can, use the same material in various/all meshes (less draw calls).

        - If you are going to use normal maps (probably will), then whatever can be normal mapped instead of geometry should.

 

For memory performance:

        - Few materials (preferably one)

        - Few maps (defuse color, normal, specular. Some engines will mix specular and defuse into one to save memory)

        - Small maps/low res (512x512px better than 1024x1024px)

 

Its up to you how u balance theese things, because they are incompatible. For example. If lots of details will be normal mapped instead of modeled, you gain in redering performance, but u need a higher res normal map, and u loose memory performance.

 

Modelling for games is like designing a gun. You want a gun that shoots a fast, big and heavy bullet, very far, very acuretly, hi ROF, lots of ammo, no recoil, very light, silent, and small enough that it fits in your back pocket. You cant have it. You need to find a balance.

 

A good aproach is to make a very detailed model, save it, and then go stripping, like they did with that SAW. The detailed model can later be used to project normal maps, as people have pointed out. This is a task done automaticaly by most 3D packages.

 

PS. Im not a skilled artist either. I write code. But I know some of the technical issues. And I do some modeling for fun. Thats my 2 cents. Hope I help some.

Share this post


Link to post
Share on other sites

Bull made the recommendation for 3DS max a few months back. I have been working on learning it in my spare time. It is NOTHING like cad apps, which are dimensionally driven, whereas this is polygonally driven. So that has taken some getting used to. Also, trying to get polys to 4sides, especially after a lathe operation. I was going to model the bolt of my Mosin for starters, piece by piece, but its taking me some time to get familiar with start-to-finish operations in Max.

Why Maya over Max?

Share this post


Link to post
Share on other sites

How much different is current 3d MAX compared to technical 3D softwares like Solidworks / Catia etc. in creating basic objects? Have someone some knowledge of both? Is there total philosophical difference how the objects are created and modified.

 

I have used 3D MAX (3D ViZ) last time in Windows 98 :D and what I can faintly remember it were back then pretty similar in basic operation what current SolidWorks is.

Share this post


Link to post
Share on other sites

3ds Max and Maya are both equally good tools, for the purpose of creating content for video games. There's not really any reason to prefer one over the other. You should use whatever you are most comfortable with.

 

I use Max and I would recommend it to people who are new to modding/game-making, because it's one of the easier ones to get into but it's also very well supported by the UnrealEngine devs and documentation.

Share this post


Link to post
Share on other sites

Yes there is. Maya is industry standard. Do yourself a favour and learn Maya. If Autodesk killed off a software, which they do, they'd kill of 3DS before Maya.

Share this post


Link to post
Share on other sites

How much different is current 3d MAX compared to technical 3D softwares like Solidworks / Catia etc. in creating basic objects? Have someone some knowledge of both? Is there total philosophical difference how the objects are created and modified.

 

I have used 3D MAX (3D ViZ) last time in Windows 98 :D and what I can faintly remember it were back then pretty similar in basic operation what current SolidWorks is.

I'm studying Mech Eng, so I have some backround in various drafting utilities, SolidWorks/SolidEdge/Inventor. I find 3DS Max a frustratingly different approach than what I am used to in 3d CAD programs, and I think the end-function is the reason for this difference in operations. In CAD programs, you dont really care about polycount, because polys are a means to an end in CAD applicaitons. That is to say, if your cylinder in a CAD file is a series of 20 faces or 200 faces, it doesnt matter because your CNC path or whatever you send to a drawing render, its going to be shown as a circle, and calculated as a circle.

In 3DS Max, the number of faces is the end. Its ALL that matters. So instead of dimensionally driven design, its poly-count and poly-form driven design. This creates a very plastic working environement, which comes with its advantages. Granted you *can* use dimensional structure in 3DS Max, but it just seems that its more "unitless". Also, as I mentioned above, in CAD apps, you are building representations of real-world objects. In 3ds max, you are building polygonal structures. These have their own set of ideals. In CAD apps, you need to consider what you design with respect to what you can feasibly produce (machinability, precision, 3dprintable etc). In 3ds you want to make things that are easily rendered and caclulated, thus you want 4-sided polygons (though I am a bit confused as I thought it all came down to triangles). Basically you want to minimise your polygons, and you want very finite, and spatially unique polygons. This makes skinning and rendering much more feasible for engines.

 

This is my current understanding of the mindset of 3d modeling. Experts please correct where I am amiss.

Share this post


Link to post
Share on other sites

Yes there is. Maya is industry standard. Do yourself a favour and learn Maya. If Autodesk killed off a software, which they do, they'd kill of 3DS before Maya.

 

I dont mind  a steeper learning curve if the ROI is greater.

Why Maya over 3DS or vise-versa.

 

Share this post


Link to post
Share on other sites

To answer your question about quads. Quads are good for your high-poly, the object from which you get your normals and possibly low-res, and your low-res, game-ready model, will be in triangles. The low-res model is usually converted into triangles by game engines. Quads are good for smoothing when you subdivide your model. Look up a tutorial to get edge flows. There's also sites like military meshes and smcars.net for good hard-surface models. Look at them and play around with similar shapes, say, in another monitor and get used to how they smooth out. Most programs are catmul-clark for subdividing.

 

As for why Maya over 3DS. I prefer it and it's used for animation in games and film. A lot of studios don't want to be switching from different software and want to just ship work off to the next department without a change in software as problems can arise. Soooo when the animation team is all using Maya, the studio will want modelers to use Maya. Maya also usually comes with mudbox too, or for cheap, so studios will usually pick Maya over Max and the reason a software like softimage being discontinued is because studios stop using it. No revenue. No software.

Share this post


Link to post
Share on other sites

Is Maya a comparatively more fluid production environment than 3DS Max? Or is it really just about the animation portions?

If I am modeling guns, that will have recoil, cycling, reload, and perhaps other animations, is 3DS Max going to handicap that option?

Share this post


Link to post
Share on other sites

3DS max has a worse interface and is better only if you're going to be using shaders because of its superior shader network, which you won't since this is all for a game. Both have the same tools, Max has one tool that would be beneficial. It creates evenly sided objects (not describing it well) but you can get around it by doing a boolean using a newly-created primitive in maya. Apart from that inconsequential thing, Maya is better for modeling. Interface. Same, or better tools. Just use Maya.

Share this post


Link to post
Share on other sites

I would say "better" and "worse" can hardly apply as it is a preference thing.

 

I can't stand maya's interface, hence why I stick with 3ds max.

 

Besides most studios employ either both of them or as an artist you have to be flexible and pick up what modeling package is used the place you work anyway, so no point in arguing about it :)

 

Be open, be flexible.

Share this post


Link to post
Share on other sites

The general impression I am getting is that as a hobbyist, it doesnt really matter.

I would like to make useable assets, and would love to contribute even just soup-cans or dumptsers to Squad, but in the event Im doing mods/reskins for individuals to use, I am not going to find a handicapped finish product because of either work environment?
 

Having worked in various DAW environments, finding an environment that is conducive to creative workflow is key to getting things done. There's always a learning curve, but if the potential output is equivalent, might as well use the one that im able to work in more intuitively yeah?

Hate to keep beating about this, I am trying to get stuff done. My first lathed, skinned and rendered model is rather crude but its proportionally accurate.

http://i.imgur.com/ygOOUeI.png

Skins are an art, and I'm not trying to detour down that path just yet. Mostly just trying to get quadrilateral junctions on my lathe-endings and figure out a good work flow/tools that facilitate the most rapid and flexible composition of geometries...

This is scale 7.62x54r. Obviously trivial, but its something.

Share this post


Link to post
Share on other sites

I my self got into 3D by starting with google sketchup and then moved on to Maya 3D (I found Maya to be somewhat easier to use)
And I used http://www.3dbuzz.com/training/view/maya-advanced-modeling/the-socom-pistol to get into Maya.

I made this one about a year after I started modeling https://www.dropbox.com/s/x16fhw9mjlmqgbr/maya%202014-01-14%2022-00-47-10.jpg
Now this is not perfect... I was only doing a model and not for game or anything. So there are no high poly version of it. Im hoping that one day I will do it.

Share this post


Link to post
Share on other sites

maya lt preview looks pretty cool. integrated sculpting, probably just basic, but better than nothing. Mudbox should just be in maya so that you don't need to transfer, but it's a good step. We'll see what the full Maya release brings. Maya looks really game friendly with this release. It looks easier to retopo too now. New UV features too.

 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×