Editorial: A Solution on How To Continue Film Production Through An Actors Injury

I could see it working in some cases but if I'm not mistaken(I could be wrong) RDJ's injury wasn't gonna set them back to much so it may be to much trouble to go through that process instead of waiting it out.

Thomas Wayne raping a bat has made me think differently about the way comic book movies should be made. I think that's point I was trying to make. I think differently because of John Jones.

You're not fooling anyone Jacky/Comic Critic

Spend more money when the majority of the money your spending is going to someone who is now doing less "work" and not "eaning his pay" I don't think the studio execs would like that or see your point

I think you just solved that problem!!!!! I wonder why the huge money making studios never thought of that!! You are a brilliant gifted person!!

As a VFX artist myself, this solution would burn up far too much time and money than i think you perhaps realise. Yes this could be good if the actor was in a critical condition and they would take months out of the schedule for filming, but for a twisted Ankle and taking a few days out of the production, just pointless

It is very expensive to build a CGI double, ESPECIALLY if you are only creating one just for 1 or 2 shots, as there are so many people that need to be behind and involved in the creation, and so many tiny nuances that can make or break the CGI human if its off by a millimeter of a %.

There are several stages to building a digital double that include Modelling (either through Lidar scanning -> Point Cloud clean up, to taking photo's of the actor from every angle, modelling the head, and then creating separate models for each facial expression to be tied into the rig, aswell as full facial muscles to be built into the head to not only drive the animation, but deform nicely and not look paper thin, every single time a double is built, tons of muscle and anatomy research is done to determine how the muscles sit and fold depending on the expressions or actions required.

Then comes the texturing, the photos captured of the actor would then go through a High pass filter to get rid of any and all lighting information so you end up with a diffuse and this is then painted/projected onto the head from the multiple angles/photo's. And this is done for many layers due to Sub Surface Scattering to think of (how light is absorbed through the skin as it hits each layer of the skin, so you need to paint in the colours and depth for the skin underneath. Getting this to work in Maya is no walk in the park as there is no magic click method for this, sure there are many defaults but none of them physically accurate. This has to go through hundreds of calculations and be shot and scene dependent and change for each scenes lighting information.

Speaking of lighting, there is also the specular map which is delightfully easy to [frick] up considering it determines what sections of the skin skin when light hits them and by what intensity. The skin is, by its nature shiney, so to balance that and to make it look organic instead of plastic while playing off the SSS is a nightmare if you want it to be photoreal (which i'm assuming you do). Not to mention the HDR needs to be collected from the shot to get all of the sets lighting, which would then burn up days on the set and cost more money just to get the lighting information with no other scenes being shot that day, so you would then have to probably work off a still image (not even bracketed for exposures, so no way of determining the luminosity value of the light sources)

And not to mention the hair similations and the clothes, etc. Every single word of the steps above cost alot of render time and would then cost more money unless they just waited for the actor to recover for a few days

All of these steps contain tons of information to think about. Yes it works for digital doubles doing stunts of epic proportion, but you never see them close up, they are never at a stage where they have to perform and emote very close to the camera and be integrated and perform right next to co stars, so you automatically have a frame of reference the audience will subconciously match to. Corners are normally cut for the CG stunt doubles (not too much, just toned down to make them renderable/nice to the eye). So having to upgrade the Digital Doubles and make them Hero performing Digital Doubles just for the sake of a twisted ankle is nice in theory (barely), but just more time consuming and moronic and pointless when put into practice