Meet the Artist: Sharon Calahan

#6

Hello Ravindra,

I’ll first start with the disclaimer that I am not a technical person. At Pixar, our lighting, shading and rendering tools are all currently in-house proprietary software designed by our amazing software developers. For clear reasons we use RenderMan as our renderer. We do however use some non-proprietary software packages, for instance Maya and Nuke.

Our philosophy at Pixar has never been to try to create photorealistic imagery, but rather looks that are more stylized while still being believable. I think that this is more of an art direction issue rather than a tools limitation. Ultimately the look of a film is determined by the director’s vision for the story. Creating something that looks more real definitely requires the artist to pay much more attention to subtle details in how light and surfaces interact no matter how sophisticated the illumination model. I will try to answer more specific questions if you feel that this answer is too vague.

Thanks!

–Sharon

#7

Regarding the lighting aspects in Ratatouille, how far do you go when trying to push the envelope to get the right mood, and by this I mean when you have a clear idea of how the lighting should be done, is there any limitation imposed by the software or you’ve always achieved the look you have in mind and a second more specific question, you mentioned that the way light surfaces interact creates something more real, do you have any example in mind that you can share with us where lights can rescue the looks of a scene and a model?

Thank you for your answer :slight_smile:

#8

Hello Eduardo,

There are always limitations when creating a movie, which can be both a blessing as well as a curse. The most common limitations are schedule and budget. There are definitely times where I wish we could spend more time on something and make it better, but in addition to the look of the film, I am also responsible for delivering shots on time and without killing the crew. Often also render times or memory footprint is a limitation. There have been many instances where I would like to use something, for example more volumetric lights or more raytracing, but have to make compromises to fit within a rendering budget. It is all about making smart choices at every stage to make sure that we get the biggest bang for the buck and that it shows on the screen. And of course, there are always some limitations with the tools, as good as CG has become, it is still only an approximation of the physical properties of light and surfaces.

In answer to your second question, here is one very small example. Say you have a white porcelain mug. If you simply throw on a reflection without tweaking it, it doesn’t look quite right. If you really look at a white mug, you might notice that it doesn’t tend to reflect a lot of color and it doesn’t tend to reflect anything except the brightest highlights, and even these are fairly dim. If the reflection isn’t dialed in right, it can start looking too metallic and not like porcelain. Porcelain also has some subtle subsurface effects that scatters the light. This is a small example, but it took us several months to tweak the surfaces (after shading) of the hundreds of the objects in the kitchen in Ratatouille to get them to respond perfectly to the lights and reflections in the scene. We were spending a great deal of time in the kitchen in that movie so the investment was worth it. For a movie like Cars 2 where we spent comparatively little time in any location, it didn’t make sense to invest as much time refining light responses per object.

Thanks!

–Sharon

#9

Hi Sharon,

Thank you for taking the time to do this. It’s much appreciated :). I was wondering on some of the productions you have worked on how often the shading team resort to multi-purpose shaders (sometimes called “Über” shaders ) as opposed to writing / constructing (with SLIM) etc every shader from scratch?
I imagine ‘hero’ character/items obviously warrant more attention/time/effort but even for those mid distance items of which there could be many, especially with different material properties; what policy (if any) do you determine when a stock shader you guys have developed is what should be used?

Also I would love to know your thoughts on compositing for animated features. I find it a similar relationship between principle photography for live action and the editing process, giving the director a second pass at constructing the film. Having different passes for all lights/shadows/objects/diffuse/spec. etc. how much of a role does compositing play in the final look of the film, specifically to lighting and colour correction.

Again thank you for taking the time to do this. :slight_smile:

-andrew

#10

Hi Sharon, thanks for sharing your thoughts with us! :slight_smile:

How much artistic and how much technical is your profession? I’m asking because i see that your education is fully artistic, and lighting feels that it’s mostly technical.

Your paintings are lovely, by the way!

#11

Hello Sharon,

I am very happy to see you’re participating in this post and I have just one question for you. I am interested in pursuing a career in animation for films or video games, but I don’t much know what the REAL WORLD does in that field. What are the divisions of labor for a film like Ratatouille and how does each group work with the other in order to create such a beautiful finished project? I understand this might be a big questions and would be happy to be more specific if needed.

Thank you,

Phillip

#12

Hi Sharon,

A warm welcome to you on cgtalk and thank you for taking the time to answer our questions.

It would be great if you could tell us a bit about the lighting process in your films. For e.g. how early on do you actually start lookdev for a set and how closely do the lighting and shading team work with each other? Are issues like those related to fidelity of surface detail, hold out matte AOVs and geometry complexity resolved at this stage before shot production begins?

Also, how do you usually light your fur/hair/foliage assets? Do they use the same kind of toolsets that are used to light regular geometry or are they treated and rendered separately and integrated back in compositing?

And how much of a role do you play in taking stock of your existing toolsets and then coming up with your requirements for further development on a show? Because any kind of R&D would probably require a lot of time and of late we’ve been seeing an amazing amount of development happening in prman in terms of raytracing. I guess a lot of it has to do with your “demands” for such lighting solutions?

Apologies if any of my questions are vague and I’d be glad to expand on them. Your work with Pixar has been inspiring and all props to you and your team for continuing to inspire us all.

#13

Hi Andrew.

We definitely use Uber shaders for creating classes of materials that share similar visual characteristics. We also create custom special-purpose shaders when necessary. Especially for vast sets, it can be a big time saver to use an uber shader. It can also allow somebody who has great visual skills, but more limited technical expertise, to easily create beautiful surfaces. The shading lead(s) is who determines which approach is best for any given surface.

At Pixar, we do not do a heavy amount of compositing. I have definitely worked that way in the past, but generally I prefer to get as much of the look as possible “in camera” with minimal compositing. So we don’t render out each light on a layer and take another pass at tweaking the film in compositing. I find that whatever tweaks I would like to do post-lighting I can generally accomplish in the color correction suite. I’ve definitely worked in a more compositing heavy pipeline in the past, but I very much prefer to get the look I want “in camera” for the most part. I find it easier to manage continuity and I think it looks better and less “processed”. I also feel that the lighters become more skilled at lighting by working with the actual lights. Both approaches are certainly valid however. When I was doing commercials and effects, I preferred a compositing approach. Now that I’m doing feature work, I prefer an in-camera approach.

Thanks!

–Sharon

#14

Hi Dionysis,

Thank you for your kind words!

When I first got into CG many years ago, it was much more technical since tools interfaces were scarce. My formal education is in art, but I also took several programming and related classes to be able to work in this world. Over the years the tools have become easier for non-technical people to use, and over the years my role has become less technically focused, although my early experience writing code still serves me well when I need to conceptually understand how something works under the hood. But it is a bit rusty, the other day I was struggling to remember how to write a simple shell macro and had to look up the syntax for some sed stuff. Fortunately my vi muscle memory was still functioning! And back in the day, I remember writing a particle system in C++ for a commercial I was working on, and writing enough shader code to understand the fundamentals of an illumination model. The lighting team itself is widely varied in peoples’ artistic and technical educations and abilities. It may seem too obvious to state, but the people who tend to be the strongest lighters tend to be both artistically and technically strong.

Thanks!

–Sharon

#15

Hi Phillip,

You might deduce the following with a close scrutiny of our credit crawl, but hopefully this is easier to parse:

Story - They draw storyboards and help the director develop the story
Editorial - A continuous process of editing the film from pre-production through post-production
Art - They design the sets, characters, etc.
Layout - They stage the action to the camera and move the camera
Sets - Modeling, shading, set dressing, matte paint for sets
Characters - Modeling, shading, hair/fur grooming, cloth tailoring, rigging for characters
Animation - For hero characters as well as other moving objects
Crowds - Set up and animation for non-hero crowd characters
Simulation - For cloth, hair, fur, etc.
Effects - For special effects such as explosions, water, etc.
Motion Graphics - For any 2D animation graphics
Lighting - Includes support team for optimization and lighting tech support
Rendering - Renders final images and manages the render farm

All of these groups work closely with many other groups, it is a very collaborative process. As you might expect, the character riggers work closely with the animators as an example. The various groups change or merge with other groups depending on the needs of the particular film and who the leads are.
I hope that this begins to answer your question.

Thanks!

–Sharon

#16

Hi Sachin,

Yes, the lighting and shading teams do work very closely together and this collaboration begins very early in pre-production as we start to develop looks. We do try to work out as many issues as possible as soon as possible, especially for anything that we don’t already know how to accomplish. In particular we want to make sure that the light/surface interaction is worked out and easy to light. We don’t tend to do a lot of hold out mattes, but for any special needs we talk about them as soon as possible. I am involved in the modeling and shading processes all the way through production and am able to flag any potential issues with geometric complexity, etc. as the work is created.

For hair/fur/foliage, they are not usually rendered separately, but it isn’t uncommon for them to require special handling in lighting, either with special lights or special shadowing.

I am usually on a production for a little over 3 years or so, and in the early stages of designing the look for the film, a big part of that is specifying any new tools or processes that will be needed to accomplish the look. It does take time to develop these tools, so yes, the sooner we can identify them, the better.

Thanks!

–Sharon

#17

Thank you for your time Sharon.

#18

Thank you so much for your time Sharon. All of these questions are being answered with some very awesome answers :slight_smile:
Now I was wondering, as far as an animation pipeline goes, what lighting methods are usually preferred? Usually for vfx intensive, there is a high use of HDR and indirect lighting methods that are used, and was wondering if this was also the case for an animation pipeline. Or do you stick primarily to using Area lights, spot lights, etc. to achieve the look you’re going for and have more control over the final look? Thanks.

#19

Hi Fausto,

The lighting approach for a project is very dependent on the preferred working style of each lighting DP and the look for the movie, although the trend lately is toward more use of HDR and indirect lighting. I prefer to have more control over the final look, but some DPs prefer a more streamlined process. Our tools seem to be settling into the ability to provide the flexibility necessary to achieve a variety of looks and working methods, which is the best one could hope for.

Thank you for the awesome question. http://forums.cgsociety.org/newreply.php?do=newreply&p=7301560#

–Sharon

#20

Hi Sharon, welcome to our cg-cave!

My question is a little different from what has been asked here. As many people here, I felt in love with cg after seen Toy Story and since that I knew I would want to work in something related with computers and movies. Sometime back I studied vfx and I remember being very active and inspired. But then the global crisis came shake everyone.

I live in an area where it already is difficult to get a job at this. I’ve seen some succeed and be able to maintain their positions although is being really difficult to find new positions in something cg related. Since some time back now, I feel that I have lost that sensation of loving what I wanted to be, that inspiration, that sensation of pursuing what you want…

Now I would want to ask you if you have ever felt like this or know someone who has been in the same situation, and what would you do to get back to the path.

–BTW, sorry for going so long digression!

#21

Thanks a lot to Sharon for participating to this and CG Society for organizing it. Good thing I received the CG Society newsletter today or I would have missed that amazing opportunity to learn about her thoughts and process.

Hi Sharon I am a CG animator and buddying film maker but I have also started learning oil painting so I love both your CG and traditionnal work.

I have few questions for you mainly related to the technical aspect for now :wink:

1. Since your job entails a fair amount of technical skills on proprietary softwares, how one could learn the job of a Lighting TD or create a portfolio/showreel in line with what you are looking for?

  • Should someone hoping to pursue a career in Feature Animation take one of the very expensive Renderman courses or the more widely available Mental Ray/Vray are just as good to create a porfolio?

  • A friend of mine made a showreel a while back, is it the kind of reel you would be interested to see when hiring or you think an artist should also show traditional paintings (pastels, oils…) ?
    http://www.olivier-ladeuix.com/blog/2011/11/17/robert-craig-lighting-artist/

2. As an animator, I am very interested in characters.

  • On Ratatouille how many passes did you output for the Characters?

  • Do you render the eyes, skin and clothing on separate layers?

3. What do you think about Ambient Occlusion and Global illumination… well I guess you probably don’t use those as Renderman is based around Point Cloud but maybe you have some thoughts to share with us? I have been told you were reluctant to use Global Illumination on Cars. Why would that be?

Thanks again Sharon

#22

Hi Sharon,

All Pixar films recently have been in stereoscopic (as in 3D glasses) does this influence your decisions regarding layout, composition etc, for example, they say that “3D” films work better with static cameras so that the viewer can look around, also people say that travelling though tunnels works well in 3D, I particularly like deep scenery where the parallax really comes to life in 3D.
The polarised glasses affect the colour brightness of renders, do you have a system to counter act this?
Do you have a department at Pixar who think in stereoscopic and advise on what will work or what will not? or is do you just build a movie then render it out of two cameras?

in short how much does 3D influence the cinematography processes.

many thanks,

Will.

#23

Hello Sharon!

I’m so glad you are here to answer everyone’s questions and I want to thank you for giving your time to us for a bit.

My question is not particularly related to your field but rather the animation field itself. I was curious if you know how concept art within the industry (particularly Pixar) works and what kind of a profession is it like? For example, when a concept artist comes in to work on a new project with you, does that person start the project and then leave, or are they a consistent part of the team throughout the entire project? Are they more like contractors or full-time staff?

Thanks so much, and I’m just so happy to read your story (Ratatouille is my favorite movie, I think I’ve watched it over 100 times just for the art).

Warm regards,

Michelle

#24

Hi Markel,

I’m so sorry that you’ve had so much trouble pursuing your dream. I feel so fortunate to be in an area at the right time to have been able to make a go of it. I didn’t originally think I was going to be doing this, CG didn’t even exist when I when to school. I ended up here because I pursued my passion wherever it led, not knowing where it would go. I was unafraid to pack up and move great distances (even countries) to take a few chances and hope it worked out, most of the time it did. Even if CG isn’t where you end up, simply pursuing what interests you will lead you on an amazing journey. Where you end up may be even more exciting. For me, making images makes me happy. And I am lucky that there are people who will pay me to do that on the computer. But, if I did not have that, I would be making them another way (which is also why I paint). Ask yourself what it is about it that really makes you happy and try to find a way to do that, CG or non-CG. Good luck!

–Sharon

#25

Hi Olivier,

I’ll try to type with fewer typos in this post than I did in my last one!

I’m not sure if there are any schools in your area with a CG program, but a lot of people take advantage of course curriculums and rendering resources to do projects to create show reels. The software really doesn’t matter, what the result looks like is all that is important. That, and the ability to talk intelligently about the work and why you made the choices you did. Your friends reel is a great example of the types of reels that we typically receive. It is also great when candidates show personal art work; we are very interested, but it isn’t necessary.

On Ratatouille, we did not render the characters on multiple passes. We tend to render scenes more “in camera”. We have special lights to do special handling of some of the surfaces, but they are all rendered together.

On Cars, we did use some global illumination effects, at least whenever we could afford them. I had a pretty strict rendering budget and needed to choose carefully where we spent the render cycles. I like GI and occlusion in general, but I think a lot of people over use them. I don’t particularly like a look where somebody has cranked up irradiance for example and then slapped a bunch of occlusion on top. I think as long as somebody are using them tastefully and know why, I love them.

Thanks!

–Sharon