Jump to content
Sign in to follow this  
Zmotive

R16 Workflow: Being More Efficient

Recommended Posts

There are many talented people in here who have been at this longer than I, so I'm curious: once you have a sense for exactly what you're trying to create (done with any 2D sketching or Photoshop storyboards for example), which components of the workflow do you tackle first in C4D and why?

 

I know for movie production there's a pre-arranged pipeline where one group is working on the models, another the rigs, and so on. But when you're a one man team (or one woman team) and you have to produce a fairly complex scene / animation, how do you order things to avoid headaches and get through it as quickly as possible while maintaining quality?

 

Do you first handle the background and lighting, using pre-existing models and materials close to what you will be using, in order to "set the stage"... then start customizing your models, textures and any particle systems... then set up your keyframes and camera moves, then start tinkering with your renderer for final output?

 

Or do you model first under generic lighting, then set up the keyframes and camera moves, then work on the textures and lighting together, seeing how things play off one another, before finally working on the render settings?

 

Part of this question has to do with the new Reflectance setup in R16... after tinkering with it the last month and still a lot ot learn, I imagine the way it's set up might cause some changes in people's workflow / how they approach the process of getting from A to B.

Edited by Zmotive

Share this post


Link to post
Share on other sites

Every project is different. But usually we try and roughly block in the animation first. Lighting and texturing is last.

 

The reason for this is not to get bogged down in unnecessary detail, once you see what will be important in a scene. How close the various elements get to camera etc. You can make strategic decisions about what really needs TLC in terms of lighting texturing (sometimes even modelling) versus which elements will work if done more roughly (this also helps to keep render times down).

Share this post


Link to post
Share on other sites

I never do much in the way of look dev, lighting or animation before I have my geometry ready. C4D is far too screwy and hyper-sensitive in that department and fixing a temporary proxy animation or rig can be more work than just starting fresh. It also stands to note that things like Ambient Occlusion and Reflections can look completely different without sufficient details, so you may end up doing the work twice.

 

If a client wants an intermediate file for review and tests, I splice it completely from the main branch and only do the things that are critical for a given session. I may use some lighting setups and materials established in such sessions and simply copy&paste it over to the main scene when the time comes, but pretty much everything else is re-created, with the approved files merely being a guide. There's always some detail you still can tweak and in my experience it's not all too helpful if you show a too much refined file that the client may fall in love with even if it's not meant for final use.

 

Also in my world there is usually no such thing as modeling or texturing to meet a specific lighting. As an oldschool poly mangler who has been doing 3D for 20 years I just hate sloppily done meshes and at least for my visualization stuff models need to look exact and reasonably good from different sides, anyway. And naturally bad geometry can cause all sorts of rendering errors as well.

 

Mylenium

Share this post


Link to post
Share on other sites

camera and blocked out animation first. the most simple previz you can manage. edit it with audio. ( make the audio yourself if necessary, record just your voice making sound effects like a child).

 

After i have the story there. Basically its the whole project, just in terrible quality... Now it becomes pretty easy to see what you need.

 

I'll probably do some look development at this point. pick some key frame ( your favorite/the most important) and make them, don't worry about the animation, just push it, so it looks like a final frame of your animation. you, can use photoshop to fix stuff, whatever, just get to a point you are happy with.

 

Then just build up in order of importance. Its situation dependant from there.

 

id probably do the character animation first. so model the characters and rig them up. ( dont forget to uv unwrap em too).

 

texture painting can be done before animation, i'd probably just block in some basic colors though, and move onto animation.

 

then do the important props and visual effects.

 

it really depends on the project from there.

 

I think as a general rule of thumb work from Big to small and work away from the camera. So deal with the stuff close to the camera/focal point first) the rest of the stuff might be blurred out with DOF...

Edited by vozzz

Share this post


Link to post
Share on other sites

Great responses so far... thank you.

 

anothername / Vozzz:

 

So when you guys block out the animation, do you set up the geometry pretty close to what it's going to be (like Mylenium) and then set up the animation and camera moves under generic light and textures? Or are you basically using placeholder geometry?

 

I do find that if I try to make the scene look right first with generic models, I get bogged down in the details -- if not the light, then with the look of the texture under the light. I can see this happening a lot more with Reflectance. Endless tinkering with layers, etc.

 

Tangent to that, I use the Physical renderer most of the time but considering V-Ray for C4D 2.0 when it arrives (the RT stuff looks pretty promising and not limited to only Nvidia setups AFAICT), and wondering if the Reflectance side of things will change how you guys texture when working with V-Ray or other dedicated renderers like Octane. My undertstaning is Reflectance takes a page from V-Rays material playbook and handles certain things similarly.

Edited by Zmotive

Share this post


Link to post
Share on other sites
So when you guys block out the animation, do you set up the geometry pretty close to what it's going to be (like Mylenium) and then set up the animation and camera moves under generic light and textures? Or are you basically using placeholder geometry?

 

 

Really depends on the scene and how complex the models are. Most of the time we do projects with pretty simple models and/or slightly altered stock models so in that case the models in the rough animation would be more or less the finals. On some projects where it's more complex and we're a small team we'll rough things in with stock stuff or even primitives, or placeholder models, and then swap in the different versions of the models as they're ready.

Share this post


Link to post
Share on other sites

Years ago, I heard someone, somewhere say something along the lines of: "You know the saying 'lights, camera, action' in the world of production? In our world, its best to flip the order to 'action, camera, lights'". Action being models, the environment, particles, objects, etc. That really stuck with me, so much so that my C4D projects upon opening them only have three collapsed master nulls named 'action', 'cameras', and 'lights'.

 

I generally try to do the heavy lifting on materials after lighting, and render settings after that, but thats just my personal preference. Like anothername mentioned, its easy to get caught up doing potentially pointless detail work in materials and render settings.

 

As others have already said, every project is different so there is no single answer. Its a bit of a mix of the deadline you are dealing with, the review process of various stages (mixed with your client's ability to understand rough stages such as software preview renders), and most importantly, what the primary focus of the piece is. Typically, if the main focus needs to be a logo, product, etc I try to fully get the geometry in order before proceeding to cameras, lighting and texturing.

 

Again, there is no single answer that works best for every scenario. Maybe the client needs to see still frames of the final concept before starting animation. Maybe you are working with editorial and need to time things out first. You will generally be fine if you keep workflow in mind to keep things flowing nicely. Since you are asking the question, it sounds like your head is already in the right place. You just have to evaluate the job at hand and decide on a course of action that initially seems the most streamlined and roll with it.

Share this post


Link to post
Share on other sites

Depends on the job, for more complex stuff i've done, i model first, sign that off. Then uv and rigging. You don't want to go back to modelling after uving if you can help it. Then do textures, shaders/look dev under a test lighting setup. When thats done you have your asset. Then goes to animation, dynamics (&sim) then lighting then comp - its pretty standard for film/vfx pipeline work too. But with quick mograph stuff this isn't always possible to do things this way.

 

If i have the time thats the way i do it. Otherwise for simpler stuff, animated temp geo under null objects normally - so that you can replace later but keep the animation. Xrefs can be twitchy with animation, i've lost whole animations before. Which is a shame because in theory Xrefs are great workflow. Ideally the best way would be to animate null objects which are linked to your xref by expression, that way your animation is kept in the scene but it drives the xref. Then if anything does go wrong you can reimport and just reconnect the animation controls - but its a lot of work if its only for a short project. So it really depends on the time you have to setup your project pipeline

Share this post


Link to post
Share on other sites

for me, if im working from scratch, then nothing that takes more than 5 mins to model. so stick some boxes together, get overall size/shape and that's it. Because in the end its all about what the camera see's. some models you'll see a lot less of than you thought. so you won't need to put as much detail into them.

Share this post


Link to post
Share on other sites

Thank you all for chipping in your thoughts on this. I like the "action, camera, lights" mantra. It does make sense when you think about it. Also an easy way to think about structuring projects when the structure isn't being provided for you by XP or similar plugins (...who am I kidding? ...there are no "similar plugins.").

 

Sounds like unless a special case requires a different approach, most of the time people get away with roughing out a model (and perhaps some background if required) first, then diving right in to the animation and dynamics, later finalizing the geometry details, textures, and lights in whatever order makes sense to the individual. This process should work work especially well for the abstract particle animation / dynamics projects I'll cook up, since that type of geometry rarely requires the level of planning and sketching that a replica of a manufactured machine or a human / creature face would require, before being rigged and keyframed. In fact the "roughing out stage" would probably end up being the "final geometry" in many cases since the intermediary geometry is driven by mograph / XP / whatever.

 

This also gets me thinking, when you guys work with texture looks, do you (typically) find you go back and forth between those and the light parameters / positioning, in order to get the desired effect? Or more often something more like "I have to mimic the position and type of real world lights here, so I'm going to use an IES light setup (or a canned rig I like to use for this type of project), in a more or less fixed position and orientation, then do whatwever I have to do to make the textures pop"? I imagine the latter case only for architectural type work or automotive ads, etc? Is Reflectance changing the way you handle this part of the process so far?

Edited by Zmotive

Share this post


Link to post
Share on other sites

Actually just the other way around. A good texture/ material should look pretty much good under any reasonable lighting conditions (assuming the combined "energy" of all lights is the same per sample point, regardless of the actual placement, that is), though compared to other renderers that's 100 times more difficult in C4D. Or in other words: A material that doesn't look right under a neutral studio setup is unlikely to look much better with a stylized setup with colored lighting. You can of course crank up values, but it may eat up texture details.

 

I'm not sure if there's any artist out there who can match a real world light setup with the same number of lights and without extra tricks. Even Global illumination does not automatically give you that. You usually need a lot more lights and will need to work with inclusions/ exclusions and falloffs to control where light goes. In my visualizations it has also been necessary to create separate lights for specular and diffuse shading many times since large even areas look flat when catching too much diffuse, but you still want them to have a nice specular gradient. IES lights are pretty much only relevant for arch viz or some types of product viz like cars. Many IES files simply look like light gels, anyway.

 

With regards to light placement one shouldn't get locked into fixed setups. I usually make sure that there is a baseline setup that "just works", but from there I turn off lights that I don't need in a given shot or add additional contrast lights. They even do that in VFX, sacrificing continuity and consistency across shots in favor of keeping things looking cool, so it's not uncommon and for motion graphics pieces should be even less of an issue. as with all these things - it needs to feel real more than actually being real.

 

Mylenium

Share this post


Link to post
Share on other sites

I always try to block out the animation and camera path before any detail work like most of you said. I try to carry over a workflow from painting, which is start roughly blocking in the shapes, composition, and slowly build up finer and finer detail.

 

But here's a question for you guys, when do you show the client and get some kind of approval along the way? How rough do you keep things before you get something signed off on?

 

Our clients can barely make heads or tails out of storyboards we send them and I know they won't understand what they're looking at if we send them a flat shaded playbast. It's incredibly frustrating to have to go through the whole process of animation, lighting, rendering and comping, just for them to say the camera path should be different, which forces me to redo everything. We've tried on a few occastions on keeping them in the loop early on and getting approval of layout and rough animation, but it doesn't matter, they seem to always change their mind.

Share this post


Link to post
Share on other sites

Our clients can barely make heads or tails out of storyboards we send them and I know they won't understand what they're looking at if we send them a flat shaded playbast.

Ultimately that's why I'm thankful that I mostly only did industry/ corporate porn. Three days before a tradeshow there's simply no time for all that nonsense with mood boards, storyboards and whatnot. Even on the few occasions we got handed elaborate design concepts from some agency they mostly got ditched/ scrapped because the clients didn't feel the agency actually understood that showing technical details is more important to sell heavy machinery than fancy particle swirls. Most of that concept stuff in my world is simply a waste of time. I also find it extremely frustrating - you sit in some conference for hours with some marketing head and a jerk AD/ external designer they hired and nothing ever comes of it while that valuable time is taken away from actually getting work done.

 

It's incredibly frustrating to have to go through the whole process of animation, lighting, rendering and comping, just for them to say the camera path should be different, which forces me to redo everything. We've tried on a few occastions on keeping them in the loop early on and getting approval of layout and rough animation, but it doesn't matter, they seem to always change their mind.

The smart thing to do is to not show anything/ as little as possible. ;-) The projects that worked best for me and the client were always the ones with the least interference. You have to keep people out of the loop, especially aforementioned crazed ADs and designers. Though redoing stuff is somehow inevitable. Most people outside the graphics industry don't make educated decisions about this stuff, but rather use their gut feeling. They can only judge it when they actually see it in a form they can understand. E.g. in case of my technical visualizations you could only got anything approved if it looked reasonably like a machine and the function was animated correctly, so the marketing head could talk it through with the engineers and ultimately they would give it a green light. That only happens rather late in production and that's something you have to live with and plan for it, like reserving extra time on the render farm or have people on standby that can help you 5 minutes before the deadline...

 

Mylenium

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this  

×
×
  • Create New...