Cineversity

Hello,

I’m currently creating a 3D set in C4D that I’ll be compositing my actors performances from green screen in After Effects.  I was curious to find out what is the best way to do this.  Would it be better to build the entire 3D set in C4D and then composite the actors OR will it be best to take the individual model designs and repurpose them in a program like VideoCopilot’s Element 3D, and begin compositing that way.

This is new territory for me so I just want to make sure that my time is maximized in which ever route I take.  Or is there a better pipeline that I haven’t mentioned?

Thank you for your time and insight!

Hi Lurkey_JEL,

This is pretty much a typical VFX question, and not a question I think can be answered without any details and not in general.

The industry standard for this workflow is today certainly based on NUKE, or Fusion, yes After Effects will do as well.

To be as close as possible to get things “In Camera”: Get the 3D scene light set up already as close as possible. With dummies for the actors.
Also take chrome ball and gray ball shots from the green-screen set to have a reference back in 3D.
Have an eye on the color temperature and resolution. There is certainly more, but think of the interactions of light and you get most out of it.
Compare the 3D renderings with the light on the actors, do the light matches (color, shadow qualities) especially the key to shadow ratio.
To shoot a MacBeth chart is standard as well, or any comparable color support chart.

Besides that, lenses, make certain that you shoot a distortion grid. To use the best lenses goes without saying.

Markers are critical and not only on the green-screen, also on c-stands or as proxies for 3D elements.
If you can avoid it, use zoom lenses (i.e., zooming), rack focus and shallow depth of field, change of exposure during the shot, fog/dust on set for atmosphere.

With those renderings you can set the light for the actors.  The light is not a spot to actor set up only, sometimes it is good to have practical elements already on set. For shadows, caustic and bounce light.

The 3D set itself needs to be based the areas where you place the actors, either your render in foreground and background or use Object Buffer for the given scene is certainly a case by case basis.

My tip:
• For detailed lessons: check out FXPHD.com
• For a quick overview:  http://www.hollywoodcamerawork.com/visual-effects-for-directors.html
• Other sources are certainly easy to find via green-screen as keyword. Also, there are tons of books, with a wide variance in quality though.

If possible get someone on set who has done it many times, there are too many people with ideas that might ruin your shot otherwise (e.g., filling with magenta light to fight green spill [don’t!])

Again, this is certainly a question that must be scene in context, and not as a general “do this - do that”

All the best

Thank you for your incredible insight as to how to prepare/shoot and execute while shooting green screen on set.  While this information is truly helpful, I must have miscommunicated my question.  My fault.

I’m mainly focused on the creation and rendering of 3D sets in Cinema 4D, meaning is it better to composite your shots inside of C4D with the 3D set or is it better to do that inside of After Effects, with either the use of Element 3D or using TIFF Sequence renderings from C4D.  The problem I had using Maya before was I created my set extremely way too big and had way too many vertices that my computer took forever to render out just one second of a shot.  I’m utilizing C4D because of the great cross platforming capabilities it has with Adobe, which is my compositing software (don’t have enough money to afford Nuke yet).  So I’m trying to figure out what is the best pipeline to create a 3D set = just in C4D or build the set pieces and recreate them inside Element 3D which has faster render times.

Does that make sense?  Again, apologies for not being more specific/clearer.

Thanks for the additional details. Pipeline means to me from start to finish, not just authoring the elements. Hence my longer text and my caution to be not able to give anything but some general directions.

Well, as detailed in my “JET” series, where I combine practical shots and 3D, I did most of it in a 2D application, speaking of compositing.

On the end your knowledge and experience (training) will always count more than which app you use, etc. Focus on the final results!!!

There are certainly two main directions to organize it, you know what you want and those things are created. OR: everything needs to be as “last minute” in the decision pipeline as possible. This rough differentiation will guide you and determine your workflow.
There is no right or wrong, but there are of course consequences of each.

A good prepared shot will be cheaper and better than to keep things open until the end. On paper 1%, on set 10% in post 100% to express roughly a rule of thumb in cost and work. To keep things a live, e.g., 3D in during compositing slows most-likely things down close to the deadline. The more you can render already, the faster your compositing pipeline and feedback loop will be. If the green-screen key is pulled and the renderings are done, it is easy (eventually via proxies) to focus on color and details. If everything stays alive, not so much. I hope you get my drift. There are certainly other ideas about, but I hope it gives you something to consider and find your own sweet spot, as no one else can do that for you, without discussing the project like a VFX supervisor would do.

After Effects is great, but Node based compositing has so much more to offer, but the learning curve is steep compared to Ae, which is rather simple. Which brings up the point who much time you have to get this knowledge.

My advice, just take a shoot with a single actor and move it one time through the sketched out pipeline, then try another approach. You might have a personal preference and that counts as well. Or your team.

I guess these three main parameter should support your discussion process. Decision based pipeline, Knowledge based option, personal preferences.
years ago I would have added hard ware and software, but with Fusion now for free, and CPU’s fast enough and cheap, that has a minor influence, but it has one.

On the end nothing maters more than the quality you get out of it, anything else is forgotten when people become your audience, just what is really visible.

You have obviously Cinema 4D, build your set there, keep it flexible with Layers and Object Buffers.
With the sets already in 3D, you can create a Story Board with dummies of your actors in Photoshop.
While doing so, you might see already problems.
If you use Ae or Fusion is a call only you can make. Yes, learning an application is the biggest investments, not the price of the package. But free and good as in Fusion can’t be beat very easily.

All the best

Perhaps have a look at this presentation:
https://www.cineversity.com/vidplaytut/nab_2015_rewind_barton_damer

P.S.: To know the core challenges in compositing of practical and 3d render/CG content, you might check out my little introduction.

https://www.cineversity.com/vidplaylist/integration1/1_integration_introduction_01

I called this series “Integration” as compositing is typically an integration of content into another.

I started with ten main areas which I think are a must to know basis. With knowledge you might be way more comfortable to answer what way you like to go. Example, practical footage has always more or less artifacts, e.g., noise—which can’t be fixed in Cinema 4D. Based on what ever render method you chose, you might get noise as well in the rendering (or not at all), to match this, a compositing app is required, not a pure 3D animation software.

In the practical part I go deeper into the typical problems. More practical in the JET series or on my YouTube channel [see below].

My best wishes.