A quick behind the scenes and demo of the 3d work I produced as underlay and volume for Bastien to be able to project his texture and add visual artefacts. This was the very first sequence in the shortfilm and one of the early sequences in production so I didn’t have eye rigs yet.
When working on the bar sequence aka “GG Bar ANIM 01” on Bastien Dubois’ “Souvenir souvenir” shortfilm, Bastien gave me a rough outline of what he wanted to see in the sequence and a great audio to interpret and cut.
Having two characters sat at a table and exchanging lines is the most boring kind of performance you can create so I always try to make the action more specific and find some additional business, some secondary action the characters can do, just so they belong to that situation and to that shot specifically, and not to any other part of that story, or to any other storyline.
I remember several great clips from Billy Wilder’s “The apartment” we were shown in the acting class at Animation Mentor and those are great examples I always keep in mind when thinking about “secondary action”.
This shot eventually got cut in the cutting room so I can only show you the 3d animation pass and few screenshots from the final treatment. I hope you won’t get distracted by the rig glitches.
For that sequence, Bastien only gave me an outline of what he wanted to see, an audio, a camera and 2 character rigs. With a careful editing, a touch of basic modeling and rigging, it was up to me to sell him the sequence and Bastien really like what I gave him until he had to make dramatic cuts few month later, to reduce the length of the short film.
As you might have guessed from my previous posts, a big chunk of the short film was actually animated in 3d and like a surgeon, Bastien carefully removed some frames, then added a magical compositing and 2d animation/texture pass to make the 3d animation look more 2d so there was no need for a high polish of the action in 3d. We were more interested in the pacing of the cuts, the composition, the energy of the shot, which a lot came from the secondary action, aka, the business the characters were doing.
I hope you will find this interesting and keep an eye on the Annie Awards next week as we are competing in the short film category!
I had never seen that video before. Some interesting informations here that weren’t publicly available before.
I like how they split head rotations in orbit/side/nod and twist instead of xyz. This makes communication easier.
I was already using the bend/side/twist terminology myself as xyz might not always be relevant: in some rigs xyz might represent xzy on others, so discussing rotation order with junior staff usually comes as a challenge and the information doesn’t come across clearly if you don’t use a more abstract or vernacular terminology.
My guess would be that the order in Premo and Apollo, is the rotations order like in Maya, with twist as the main driver, and orbit as an additional global rotation to feather in some additional layered rotation maybe?
I will make the most of the fresh Movies Insider video article to finally plug Chris Brejon’s almost biblical lighting article which I started reading this summer while I was rendering some shots. The 10 chapter series is so in-depth that I never found the time to finish it. One day!
Now if you just want an introduction to lighting, here is good appetizer for lighting from Insider again, with some little behind the scenes from Netflix’s “Over the moon” and other studio’s movies. Like other commentators I will mention that Lighters do the lighting work, not the Animators though, it would takes decades for a movie to come out otherwise ;-)
Insider has some really talented journalists I must say. I should repost their cinematography documentaries one day.
Even if Epic came out in 2013 and most of those screenshots might be outdated, I still feel they would be of interest for people interested in facial topology and a good resource for a future overall facial topology article and facial rig calisthenics.
Blue Sky topologies are very rare online so I used ChopRig system lead developer Daniel Lima’s showreel for reference. You will see meshes and wireframes from 2019 “Spies in Disguise” and 2013 “Epic”.
“If you don’t have time to listen to podcasts, maybe you have time to read them!”
Let me introduce to you, a little project I setup few week ago, a blog where I am sharing transcripts of my favourite animation podcasts.
Podcasts have grown massively in the past 10 years and if you are not part of a high-end or animated feature online school, podcasts would be the next best thing to learn what goes on at top Feature Animation studios and hear about the workflow and techniques used in those studios.
Sadly, if you are new to the animation industry, there is no way you will be able to catch up with all the podcasts and fabulous interviews that have been recorded until now. The iAnimate podcast alone, has featured more than 75 guests accounting to more than 100 hours of episodes!!!
On top of this, most of the information shared got lost over the time as it is not possible to share specific parts of an interview, neither can you search or cross examine podcasts.
Transcripts on an other hand can be easily be indexed, searched, quoted. Wouldn’t you be interested to quickly Google search the thoughts of all the animators featured on podcasts and who worked on Hotel Transylvania or Toy Story 4 or How to train your dragon?
If you are into Speed reading, having transcripts of podcasts would allow you to quickly swallow up the whole internet!
Most importantly, if English is not your first language or you are just not fluent enough to understand podcasts, services like Google Translate would allow you to finally access that incredible wealth of information.
The Animation Transcripts Community is a free non-profit educational website aiming at making great animation material more accessible.
At the start of this endeavour, I was willing to pay for the services of outsource Transcript companies to write transcripts of only my most favourites interviews but then, I came across Otter.ai, an incredible app/website that automatically generates transcripts using Artificial Intelligence.
No, the results are not perfect, especially when dealing with animation specific terms. It is however a really great start for indexing podcasts and with the help of a community, we could eventually verify and correct and huge amount of A.I. transcripts. This is why I created, the Animation Transcripts Community or ATC, a community website where I would feature my favourite podcasts as transcripts and sometimes with show notes if I feel I need to highlight specific parts of the transcripts.
As explained, direct A.I. transcripts are great if you are in a hurry to share a specific part of an interview but not so great if you want to completely understand an interview so this is why I separated the transcripts in 3 levels of accuracy.
A one mic level of accuracy is a direct A.I. transcript.
A two mics level of accuracy is a first human review of the direct A.I. transcript. Some possible errors might have slipped and some confusing areas will be marked with several stars (***) but the transcript is already very good.
A three mics level of accuracy is the holy Grail. Several reviewers would have gone through and thumbed up the transcript. you don’t have to wait for a three mics transcript though, a two mics is already 95 percent accurate.
Head over to ATC to enjoy some great interviews and maybe become part of the team!
If like me you have been focusing on the artistic side of animation, focusing on life or just hiding under a rock for the past few years, you might now come across several new words that seem to have come out of nowhere and making you feel stupid. Don’t worry, I got you covered!
I will update this CG Lexicon as I come across or remember new ones.
ACES: To paraphrase Chris Brejon, “ACES is a colour management system developed by dozens of professional under the auspices of the Academy of Motion Picture Art and Science.”. I won’t spend time explaining it, Chris did an outstanding work on his website so go read it
Albedo: This is the new fancy word for diffuse when talking about materials.
Alembic caches: Alembic is an open computer graphics interchange framework to cache animated or non animated assets. Nowadays in most production, instead of receiving animated rigs from the animators, lighters would receive alembics in the form of .abc files. Alembics is the record of each vertex position in time.
AOV: This is the new fancy name for render passes.
BRDF: This is related to materials and how they react to light. Here is the Wikipedia short description: ” The bidirectional reflectance distribution function is a function of four real variables that defines how light is reflected at an opaque surface. It is employed in the optics of real-world light, in computer graphics algorithms, and in computer vision algorithms “
Cached Playback: Animators jumping onto Maya 2019 got particularly irritated with the introduction of the Dynamic Cache playback as Autodesk removed the timeline’s Play button and set that preference to on by default. Cache playback is a cool feature but it swallows up all your memory and not all animators are ready to switch workflow yet. It can be really useful when working on a sequence with several characters if you don’t have a GPU toggle script so I am a lot more excited about fast interaction with rigs and parallel evaluation for the most part.
DCC: (Digital Command Control) This is simply an acronym for the word software/package. Since when is it uncool to say “Software”? Let me know in the comments.
Nuke: Before Nuke, Shake was the predominant compositing software but Apple bought it, then killed it. Good bye Shake, hello Nuke. Nuke is the current almighty compositing software used at top companies.
Parallel evaluation: Until Maya 2016, no matter how many processors or cores your machine possessed, only one core would be used by your 3d animation software (DCC ;-) ). Parallel evaluation is a game changer, with Parallel Evaluation, all the cores available on your machine are now used when interacting with your rig, giving you substantial speed gain, even on rigs deformed by numerous deformers or corrective shapes. GPU acceleration is also available with Parallel but I haven’t work in a production where the GPU acceleration was stable yet, hopefully it will happen eventually.
DG evaluation: well there is nothing new with DG as this is the way Maya has always linearly evaluated mesh deformations but the term might come more often in conversations since Maya might revert to DG in situations Parallel fails.
UDIM: Traditionally when creating an asset’s texture, you would unfold all the parts of the asset tightly into a single square texture. This technique was very memory efficient but had the drawback of complexifying resolution changes on specific elements of the asset . Instead, using UDIMs, you gain a greater amount of flexibility as you are not bound to that single texture anymore as you can now separate the different parts of the asset into separate UV geographic locations (UDIMs). Making a higher res or lower res of a specific part of the asset doesn’t require modifying the entire layout of all the UVs anymore.
Deep compositing: People might compare Deep to Zdepth but Deep is a bit more complex. I will just say that it is a compositing term and working in Deep allows fancy compositing tricks in a package like Nuke. Instead of addressing 2d pixels, Nuke can now sees the scene in 3d and interact with it in 3d. This way you could modify the 3d lighting of a scene without having to re render it. Unlike Zdepth, Deep works fine with motion blur, depth of field or semi transparent object. Obviously this requires a ton of hard drive space so it is not suitable for every production.
Cryptomatte: This is an other compositing term where mattes are automatically generated by the DCC so mattes can be extracted through directly picking objects or entering object names.
Deltamush : Originally a plugin for several DCCs, Deltamush has been integrated in Maya 2016. Deltamush helps to smooth out the skinning of a character.
BakeDeformerTool: This is a tool that can be used in conjunction with Deltamush to bake the deltamush result into a new skin binding, without the Deltamush overhead.
Denoiser: Traditionally when rendering with radiosity, a lot of noise can appear at lower settings and when trying to speed up rendering time. Using a Denoise filter, you can automatically process the picture and reduce the noise to simulate higher settings. It is a cheat but the results are very impressive especially with the Nvidia Optix A.I. accelerated denoiser
animBot: well … who doesn’t know animBot? ;-) animBot is the non-free replacement of aTools. Some animators can’t live without aTools or animBot. Personally I only use aTools’ arcTracker and sometimes the aTools Tween machine if I don’t already have Justin Barrets Tween Machine already open, for the rest I already have my own scripts or techniques.
IK/FK match: I worked on some production where people didn’t understand what IK/FK match meant. IK/FK match is a rig feature or a separate script given by the riggers or TD to allow the animators to seamlessly switch from IK to FK or FK to IK without having the arm or leg going back to an original pose. This is a feature heavily used when posing characters as animators would pose an arm in IK for ease of use for example, then do an IK/FK match to start animating in FK.
A.I.: as far as I know, A.I. can’t do quality keyframe animation so we will ignore it for now!
Viewport 2.0: For years it was just a joke, as soon as you started at a company, you would be told to disable it or Maya would crash. It got stable on my last gig and Maya relies on it a lot nowadays so give it a try again. My friend Mariano is the one who reintroduced me to VP2 as he uses it to make most of his renders
That’s it for now, let me know in the comments if you think I should add more terms to this lexicon.
While watching that fun video, I thought about the “Breaking the fourth wall” concept.
As noted on the AnimC Pro Tips few month ago, you always want to avoid the character looking straight at the camera.
To me the most believable characters in that video are the ones that just, “live their life”.
The “hot grip” was okay as it was just the introduction to the skit but as soon as the “Crafty” looked at the camera, this took me away from my “voyeuristic” experience and I immediately became aware I was watching an actor performing for the camera rather than actual characters evolving in their natural habitat.
Related post :
Being my professional workhose since 2007, I have been naturally promoting Autodesk Maya on this blog, with some occasional foreays into Blender when enraged at the price of Maya’s license/subscriptions, then back to Maya when Autodesk introduced the free “Maya and Autodesk softwares for Displaced professionals” back in 2012. to support professionals during challenging times.
That scheme gradually disappeared without much noise and recently Autodesk introduced various options on top of the classic educational subscriptions.
Currently, you can get Maya LT, a 300 dollars a year cut down version of Maya that unfortunately doesn’t allow you to run most of the free animation rigs available on the market so for animators who want to focus on character performance this is a no-go. Python is not available either when everybody in the industry uses Python and scripts like AnimBot or any Phython scripts won’t run on it. Rendering is not possible, dynamics are not available, referencing is missing….
Then in some selected countries, you might have access to Maya Indie which is the full feature Maya, for “artists earning under $100,000/year in revenue.” This would perfectly cover hobbyist living in those 5 countries if only there wasn’t a catch! So what is the catch? The catch is, that this scheme is only valid for … one year .. then your subscriptions automatically gets renewed at 250 a month ;-)
With my inclination for education and support for struggling populations in the light of the murder of George Floyd, I just can’t continue predominantly posting articles related to Maya.
“Learning 3d at home is a way for modest populations and kids, to potentially gain access to employment and revenue streams”
Learning 3d at home is a way for modest populations and kids, to potentially gain access to employment and revenue streams, outside of expensive institutions and without incurring lifetime debts or adopting piracy habits.
Sorry, that’s a long winded introduction to finally explain that I will start covering Blender more and more on this blog ;-)
This first Blender post is a two years old video but a major one as Blender founder Ton Roosendaal, will cover the history of Blender, its philosophy, and its challenges with some interesting comparison with Autodesk and Zbrush.
I never dug too deep into Blender in the past as the interface and the Right Mouse Button (RMB) slection really annoyed me. Unlike some other software creators (TV Paint, Zbrush…), they realised that they shouldn’t force users to their philosophy and accept that some user interaction have now become standard, the left mouse click (LMB) for selection for example.
There are plenty of interesting bits in that video and 1 hour into it, Ton reckons there are just below 20/35k paying Maya customers and less that 20 developers working on it. This could explain why we get so many half baked features every releases and makes me wonder how many companies might be using pirate copies.
If you are in a hurry, Andrew Price was nice enough to include timecodes in the video description.
Being mostly interested in Animated Features and Animated TV series, I don’t visit FX Guide much but I just found a series of interesting interview I will be posting in the next few days.
Here is a very interesting one with Disney “Feast” director Patrick Osborne where he goes into a bit more details about the unique visuals of his short film.
Via : FX Guide