So there are basically zero resources for Cinema 4D within a game Art / Development / Realtime workflow, it took me a long time figuring things out and I feel like I need to put out some resources for people scrambling for information like I did.
This is tailored towards beginners or curious people
1 - Yes, Cinema 4D works very well for a game workflow if you know your Software
2 - No, you 99% won't get a job as a C4D Modeler in a AAA Studio, but small indie studios generally use what they want
3 - The main advantage of C4D over all others is the amazing 100% customizable UI, best hierarchy/scene handling, and ease of use
4 - The main advantage of C4D over Blender, Maya, Max is that you can work a lot smarter / parametric and non-destructive, and more accurate
5 - The main advantage of C4D in a game workflow is the new amazing R20 Volume Builder and making smart automatic setups
6 - The main disadvantage of C4D is the lack of Multi UV editing
, (Face weight) Normal tools and lack of community
/ plugins for realtime workflows
7 - 2023 - A big new advantage is parametric vertex color based on fields. Fields in general are amazing
gamedev tools.C4D KILLER FEATURES:Volume builder.
Its easily the best highpoly modeling tool in the industry integrated in a mainline 3D package. The options are endless and its fully parametric and plays amazingly in other c4d features like fields or cloners and randomziers. You can throw lowpolies in volume builder and have highpolies in seconds. You can apply complex noises to create procedural wood grain per example with real 3D depth. You can do so many complex things. Blender also has a voxel feature but its very far away still. If you like baking highpolies, its worth to get a cinema version for your studio for volume builder alone easily, its a complete secret tip. It is also easy to decimate things for lowpoly as there are 3 types of polygon reduction and it bakes really well without any retopo work even for complex stones or such. You can make complete parametric models with arrays and cloners without any editable polygons if you want. You can voxel splines, or even particles. Over time, by sound, whatever you can think of.
Also by fields. You can make a rock where you just put in a lowpoly box and it will make a highpoly rock out of it, complete with water erosion on bottom and horizontal lines and everything if you set it up that way. You can shoot particles on a wall and volume them for bullet holes.
You can make splines into slimy tentacles in seconds or make extremely technical models with infinite stacking of overlapping additive or subtractive booleans which just has hard limits on normal polygons. (Houdini can do the same things technically, but the workflow is completely different, here you can control every centimeter and model completely normally but its not that effective to make big batches in comparison and it takes seconds to set up what you want)Effectors, Parametric Workflows & Hierarchy
Historically the main power of cinema is the special hierarchy and the way it is set up. The way cinema works is that all objects on the same level of hierarchy are affected by effectors. This means you can use a bevel deformer on many unconnected meshes at the same time per example.
You can stack and connect things and reference them in other places and forming deep interactions that way which are just not possible in other softwares in such a form. This is where the power for the motion graphics comes from which C4D is known for. You use simple parametric setups all the time and work very little in editable poly or merged meshes. You can keep things fully editable for very long times if you really want to and change their subdivisions, bevels and keep instances and update them across your thing in seconds. In most other softwares, the first thing you must do before being able to even start is make your model editable, so this is quite different.
The hierarchy is also very well developed and could be greatly used for scene handling for your game or use as editor tool (if you are on feature poor engines such as Unity). Cinema exports are also extremely reliable and never have bad applied scales, rotations and the sort compared to blender.Fields.
Fields are a collider type structure, like box volume, which intertwine really well with C4D features and volume builder.
Now in 2023 you can even do parametric generated vertex colors. Example C4D setup:
I use a cloner with 4 rock meshes. I use a cloner to scatter 200 rocks based on these 4, still changeable any time.
So I can scatter rocks on the ground. Use my rocks in a deformer field to move down the ground where the rocks touch the ground to give the rock a "bed" and make them tie in nicely to the ground. I use a polygon reduction to optimize it and keep only a polygon density below the rocks so they have enough vertexes for a good vertex AO but the rest stays lowpoly. Then I use the same field to vertex color the bottom of the rocks so I can have a vertex color based occlusion there. Then I vertex color the ground around the rocks the same. I add a random field to give my ground a blue and green random vertex color for automatic color variation. Then I use a box field to clip all and make the colors not go on the plane border, so it can tile perfectly. This way I have ambient occlusion on an untextured ground plane for tiny pebbles and vertex painting for variation built in.
You can also use a sphere field to damage areas on a tank, crumble the surface, shoot holes in it, bend it the same time and vertex color the borders on the lowpoly at the same time as doing the highpoly and all sorts of things. Or you use a gradient field to vertex color all bottom areas of tons of meshes to create an AO effect. These can also be animated or you can expand them, search for nearest and many things. Use random noises, color by formula, audio, its very deep. You can also transfer the vertex colors even if the meshes don't match!
So you can have a highpoly dragon and have a sphere boolean cut its head, and the lowpoly can get the vertex colors at the same time, or you copy it to the lowpoly, or even a completely different mesh, as they are positional and not ID based.
It opens a new dimension for vertex color based workflows.
Cinema has a fully customizable UI with every component being separated. You can set up a completely tailored UI for your teammates which has just what they need as example. You can also set your hotkeys and camera exactly like in Max or Blender or your game engine.
WHERE DOES CINEMA STAND?
Cinema 4D should be seen as an in between
of a 3DS Max / Blender (Brute force modeling tools) and Houdini (Parametric Tools)
So cinema allows you do to smart parametric setups and combine it with manual work very well. The famed motion graphics tools allow you to work much smarter and you can go very far non destructive with full manual control. However it is not made at all to create automatic pipelines like Houdini offers to do and not to the same horizon.
GENERAL INFOS FOR NEWCOMERS
(BASICS) VERTEXES AND SPLITS - MUST KNOW
1 - Vertex count is what matters most, more than Triangle count, and look up how draw calls work
2 - Each UV Split doubles Vertices at Split
3 - Each Overlapping Material doubles Vertices
4 - Check object information of your models to show vertx count
5 - Check object information does not show split extra vertices (expect 3x the vertice count in engine than what you see)
6 - Ignore phong breaks (smoothing splits) until after you have UV mapped (unless you don't bake from a highpoly)
7 - Keep in mind Phong angle overwrites manual set phong breaks (smoothing splits)
(BASICS) UV MAPPING MUST KNOW
1 - Don't try to UV map in C4D it is unusable, use 3D Coat or other (Update: With S22 UV tools are much better, but still lack multi UV edit)
2 - Smoothing splits need to be where UV splits are, UV so you get as little splits as possible to save performance
3 - If splits do not line up, you get seams in your normal map generally, but we have a script to remove the issue
4 - After you have UV mapped, align smoothing groups/phong breaks to UV with the Script https://hostr.co/gdz7YlMDNL7X
(Make sure to unbreak all phong breaks on the lowpoly before using the script)
5 - Since 26 or so there is a new feature to select UV islands, this is perfect to align UV islands to UV splits
(BASICS) NORMAL MAP BAKING MUST KNOW
1 - Really use Marmoset or Knald to bake with a realtime preview it saves unending hassle
2 - Most errors are due to cage distance or false UVs / UV Splits
3 - There is a ton of baking resources, check Polycount wiki
4 - Export your meshes with FBX, seems to be the most consistent. Obj is like a feature stripped FBX.
5 - To preview baked Textures in C4D, turn up resolution in the Material > Editor tab
6 - C4D dosnt display normals as they are in-game, flaws are usually exaggerated, renderer also displays differently than preview
7 - Minor flaws are ok and nobody will notice, its not exact science
8 - You can export single objects with the command "export selected object as.." - get it from (Shift+F12)
9 - Null object folders are not exported, you can use meshes as groups however (delete all polygons of it)
10 - If you import smoothing splits from other software, use the "apply phong breaks" option on import, not normal tags
C4D VS BLENDER VS MAX VS MAYA
(I mostly know C4D, then Max and Blender, Maya I just know from the team)
First of all there is no best software. In our Team we have different users and the clearest thing I know is that all have significant
weaknesses and advantages so knowing multiple is always great. Knowing Blender and C4D can be very practical as both can do things
the other can't. Blender Max and Maya seem rather similar in terms of feature sets, they all have a brute force approach, strong modeling
a lot of features, poor usability but you are not modeling in a smart way but in a normal "collapsed layers" style mostly destructive fashion.
Blender realtime advantages:
+ Face weighted normals as Stack, Easier stacking of Booleans, Noticeably Better Decimate, very strong community and resources
- Clunky UI, hard to get into, similar to the other softwares you don't work smart but destructive "just get it done", poor hierarchy, export and import (you could say the exact same for Max really)
We had compared all the plus sides for all important Realtime aspects (outside animation) and while cinema 4D was clearly having noticeably stronger upsides, it was then kinda crippled by the lack of multi UV and face weighted normals, forming a bit of roadblocks. Also the lack of community and plugins are a big downside. Blender also has just more tailored gamedev solutions, while the C4D userbase hardly cares for anything vertex or normal or these things at all. (There's also no Baker outside AO)
After some years of seeing Blender and C4D side by side, the conclusion is really that both have BIG upsides and BIG downsides each. I think C4D offers more unique tools and options for greatness if you like experimenting with parametric setups or generating things, but if you want to use houdini, on the side you might as well go with Blender to do classic destructive but efficient modeling and benefit from better built in rendering options and some other features.
I haven't use max for some time, I think max is a more extreme version of blender, very fast in destructive modeling but a patchwork monstrosity I couldn't wait to get away from personally. Maya I have little experience but from what I've seen it was neither fast nor very impressive but it surely must have its status for a reason.
Cinema modeling at a glance is similar but not really. Cinema is a little bit more in the "Blueprint" direction. You can work heavily with the hierarchy, non destructive primitives (like Cylinders, Bevelled Cubes etc) and change your mind about Details or Proportions later. In Cinema primitives stay editable all the way, while in other software they become Polygonal once you used them the first time. In Cinema you heavily model with editable primitives and seperate objects usually, you work a lot on an object and primitive property level while in 3DS or Blender you would usually work on merged meshes and edit a lot more on a vertex level. Cinema is very strong on modifiers and automatic setups and using these within and cross hierarchy. You can make effectors or bevels apply an entire group, link, mix and combine group and do all kinds of crazy combinations over multiple levels. You can even model without actually modeling with the correction modifier. Arrays, Cloners, Random effectors, Instances, Modifiers, these are your big advantages. Other softwares do have these as well but it does not work on multiple levels / cross hierarchy and are pushed at lot less so they are often not incorporated into the core workflow in any meaningful way and often exist more for completion sake. You can model similarly fast as in blender and also equally destructive as in the others if you want from my experience, but boolean and bevel stacking on the same level works noticeably better in Blender. Max seems to have superior modeling tools however to both but not to a degree where it really matters. They all have upsides and downsides depending on the tools and the modifier. C4D does have some modifiers which are surely inferior by themselves to other equivalents but it is the structure that is the advantage, which does allow the wild multi dimensional usage and combinations across multiple objects and folder depths, while you generally can apply one modifier to one mesh in the other softwares. With expresso (like Unreal blueprints) its also just seconds away to link properties and more making this even more extendable with simple visual nodes.
Edit: Accuracy - Compared to Blender or Maya (3DS its been too long) Cinema is also much less prone to buggy exports and poor measurements, as working accurately is much easier in Cinema, while I see my artists struggle to have perfect (not almost perfect) measurements and vertex positions in maya or blender, these others have very poor with transform workflows, mess of "applied scales" and all the sorts, which is basically never an issue with cinema as everything is always applied when scaling, empty objects actually work and the grid, snapping and quantizing is working much better than in blender per example, not to forget exports are rarely broken. Blender is extremely prone to broken exports even if you have saved settings.
We have objects with factor 100 scale all the time or other things.
Here is per example a setup you can do easily in cinema. I instanced a wall piece 8 times. Then I set up curve splines for different corners and circular diameters all feeding from the same mesh. But the best thing is that these are all primitives. These planes are still uneditable and you can change the subdivision at any time. Even this rounded corner is a primitive, I added a correction modifier which actually allows remodelling the vertices to make a corner without being edtiable (!), then I applied a bevel modifier to the corner for a smooth curve. Now I would instance that primitive multiple times and add different bevel modifiers to get different curves, having a fully real time editable setup with zero editable poly required.
In Max or Blender you realistically
would just brute force model these walls and then replace pieces usually and you very rarely work with the hierarchy. Edit: Working a lot with parametrics and hierarchy can lead to "never wrap it up" syndrome a bit however ...
Here is another setup we struggled with heavily in Blender, where we ended up making all the pieces by hand, copy pasting the screws 100 times and more. But such things are extremely easy in cinema and non destructive. All of these are still primitives, none is a editable mesh.
To make a pipe set I would just make a spline with all possible curves in one go, place instances of the connectors by hand or by a cloner, and then finish the pieces, cut it at the end and export them. In Blender this process for a larger set took almost 2 weeks as iterations were required, requiring manual remaking and copy pasting of all the hard modelled pieces. (Not even a radial array that can be changed later is possible in Blender nor having an primitive Cylinder that stays primitive) Anyone can do such setups with zero code requirement and you can do even stronger things with the UE4 blueprint like Espresso node editor. Edit: Blender now also has Geometry nodes, however setting up such things in cinema is much faster and just a couple clicks away and is something you casually do while working. The same holds true for many of these type of simple setups.
Maxon recently added field structure which works almost throughout the entire engine, benefiting from the strong modular core architecture and is another very powerful way of achieving things you would otherwise require something like Houdini for.
Fields and modifiers allow for crazy combinations and all the things interact very well.
I could voxelize the mesh, then apply an acid rain emitter burning down holes into the mesh based on multiple sphere fields and then
use the same sphere fields with a cloner + randomize to place metal sprinkles to the surface of the object in the same location, apply a corroded metal surface noise actually changing the geometry in 3D (not the same as a displacer) and then save this out as a highpoly and paint the damaged area into a vertex map in the lowpoly at the same time. (The Gif is not what I described above exactly)
And of course everything is still editable - that really is the beauty and power behind it;
Another anecdote for cinema, I have 1 concept artist working with C4D and had made some props for our first game in C4D. These props were very rushed and poorly made, only for offline rendering, no acceptable topology, no UV nothing. But because these were modelled with C4D workflows, it was insanely easy to optimize the 80% of them for realtime. Turn down the array modifiers to have less clones, turn down the cylinder and pipe subdivisions, reduce the bevels, and I had dozens of props ready for realtime in minutes which would otherwise have taken days or remodelling if they were made with max or blender or saved as fbx. The concept artist can also just not give a shit about any geo or topology and I can just turn down the subdivisions, replace the instances and have 70% game ready models which would be totally unusable if he used a different software.
The Volume builder is extremely powerful and a secret tip for baking in game art atm, it can be very laggy on big meshes but you can do crazy powerful things with it like my AK here below. You see the super lowpoly this was entirely based on. In comparison with the simple Blender voxelization you can have way more complex interactions and you can apply surface deforming noises and much much more. It took me a long time to see the upsides but cinema 4D definitely offers some very powerful advantages over these others.
This mesh was entirely done with Volume builder, ZERO sculpting or texture work was used for these height information. All procedual and changeable. I could just replace the text with a different one, or apply a wood noise on the Stock. Laggy but very powerful and entirely non destructive.
I hope this gives a better impression about what Cinema offers and where the downfalls lie right now in a realtime workflow. We are using a mixture of Blender and C4D right now which seems to make sense with the current weaknesses of both, but I do not see a lot of value in mixing something like Blender and Max or Maya right now as these overlap more heavily (for game asset creation).
Here a video how to make welding seams with the Volume builder:https://www.youtube.com/watch?v=m-eXiFgz0ww&feature=youtu.beI don't expect many people interested here but hope this will show up on search results for People looking for C4D resources.
You can also get one month only for 100$ (for testing basically), or 60$ per month for a year.
Here is a good showcase of the volume builder as people seem to care about that:
I know much better what the advantages and disadvantages of C4D are now especially compared to Blender and Max (although its been some time I used max) To my surprise C4D fares a lot better for realtime than I had the impression and has some strong advantages, especially in recent versions.
Maxon if you read this, please bring multi UV editing and Face weighted normals, and allow splitting phong by UV split, these are the main crutches right now.
Even though i might not use it in the future, at least now i know what to expect
Edit: You still cant multi unwrap, but multi edit. So the issue is that you can not make your auto unwrap fit them ideally in scale to one UV. This is still a shame. The best workaround in my opinion is to do something like this:
- Crate a texture with a grid. Like 2x2 or 3x3 or such. Put it in the UV window so you see the grid.
- Now think about how you divide the pieces after your unwrap. Lets say you have 2 elements you want to unwrap but cant unwrap at the same time of course. So you unwrap, then shrink and place it to fill 2 of your grid cells.
- Switch to the other piece and do the same, and fit it to the remaining 2 grid cells. Now you have your UV filled and its all fine.
Edited with more info
The new 2023 parametric vertex color tools are amazing for gamedev
I can scatter rocks, vertex color their intersections for occlusion, color the rocks by height, add random noises on top for variation, blur them, make them expand and much more. Cut off the borders so its a seamless pattern.
You can also transfer vertex colors even if the match dosn't match at all.
Or you can shoot bullet into a tank and vertex color the intersections. Really good stuff.