'Do it in camera' will be the new 'do it in post'

[David Peterson]

I guess no one better to answer these questions. We paid all our actors since they were Union. Getting non union actors was never an option.

As for the rest of the $$ My wife and I left our (vfx) jobs and worked on this for nearly 3 years EVERY DAY, obviously we didn't get paid anything.

WHOA! But how on earth did you survive if you were doing this full time every day for THREE YEARS? :-o

 

[Karim D. Ghantous]

Hi Karim mate. I'm a bit lost. The title says to do it in camera, then you talk about physical effects Infront of camera. But I get your drift, that you are talking about what can be done in future.

I think I sort of mis-titled this thread. What I wanted to say was that physical models are easier to make now, or will be in the future, thanks to new manufacturing techniques. Lower budget productions will not have to turn to CGI as much. So the industry could very well go full-circle. Computers are the best way to composite, but they will be used less and less to create models.

 

[Wayne Morellini]

Karim, did you read the post about doing physical models? It is partly as you say, but eventually an AI could get used to modify and do things according to what you direct, and it will become cheaper again. My proposed magnetic computing block has way more power than a server farm of computers, and given the compact size can make better use f that power through closer communications links. It's more compact than a mini itx box, so could be transported from set to set. But to figure out just the AI a d software, could take decades. Eventually once an AI is figured out, you could tell it to make such software, and AU, and not be sure of what it wrote.

 

[Wayne Morellini]

I've added some more details to my previous post last week, on the future of computing.

 

[Björn Benckert]

I think I sort of mis-titled this thread. What I wanted to say was that physical models are easier to make now, or will be in the future, thanks to new manufacturing techniques. Lower budget productions will not have to turn to CGI as much. So the industry could very well go full-circle. Computers are the best way to composite, but they will be used less and less to create models.

Sure there will be real props and models made in the future for different reason. But today as just about anything is made in a computer before its made in real its vey much so that if even if you now adays have rotoprototyping, 3D printing etc. Also the digital domain has developed. What was 5minutes per frame 5 years ago is realtime today in terms of CG reneders. Unreal engine and others are making it not only real time but also much easier to build sets and integrate AI models into those sets.

If you look at NINGYO in this thread it´s easy to see that if they would have tried to create it all in camera, or even half of it, then the film would be less. So the trend is that it´s not only the monster models that will be 3D but also the sets and the extras and soon also main characters.



Here is one example where we did a 3D robot for a commercial. The cost of designing, modeling, texturing, rigging, animating and rendering, basically creating what you see in the shots was actually less than what the model maker charged to create the same figure in real for JURA to have on events. Just standing still, not moving or doing anything. So I think CG is already leaps ahead and real prop / models are not really catching up even though it has become easer to make animatronics and such with todays motion capture and brushless motors etc.

Before after: https://vimeo.com/224059202
pass: jura

The you have the aspect of correcting whats created. In CG you have the option of altering and improving the shot until delivery date, when done in camera... things start to get costly quite quickly. In the case above. You would need quite a budget to have Roger Federer waiting around while some animatronic guys fiddle around with the animation of their robot... In this case the camera team got two half days to shoot Rogers action which was quite tight.

 

[C. Burkhart]

It has been my experience in the observation of other DP's (not me) that the more a DP chooses to control things in camera (to the point of deleting/not keeping anything extraneous to the precise outcome he/she was looking for) the higher up the food chain they are. They work consistently because they deliver and they do not give ANYTHING outside of the promises they have made and the exact brief that has been agreed upon. I think this attests to the level of their concentration, experience and precision in their choices.

Something to reflect on.

 

[David Peterson]

On a related point, this video is quite impressive:

https://www.facebook.com/groups/OctaneRender/permalink/923248807852198/

 

[MiguelOrtega]

WHOA! But how on earth did you survive if you were doing this full time every day for THREE YEARS? :-o

we barely did. the plan was to not leave our jobs until we had $100,000 in the bank saved to not work.
we blew through all of that in those 3 years just living. Now were broke as a joke.

 

[Phil Holland]

we barely did. the plan was to not leave our jobs until we had $100,000 in the bank saved to not work.
we blew through all of that in those 3 years just living. Now were broke as a joke.

The gamble is real and in my opinion towards your's and your wife's efforts I think worth it, though not a clear result yet.

So is how this whole thing really works on that level. At some point you need to take the big risk if you want to explore that unknown. It's not a pretty experience, but one you must take if you are looking to dive deep into that zone.

 

[Zack Birlew]

I think innovations like the Lytro Cinema camera showed with its light field image capture for real time compositing will really come into play and change the way VFX are done, especially for in-camera work. One idea I wondered about was the live chromakey units you see at trade shows like NAB in that we can load LUTS and, like Magic Lantern, programs and information to our cameras, so why hasn’t anyone incorporated a live chromakey feature into a video camera? Even loading a still would be helpful for VFX artists and potentially adding gyro sensors like motion capture stuff does, you could have a full 3D set plate displaying all in-camera. Of course, the plate would just be for metadata to sort out and not embedded in the footage directly, so that would already be a lifesaver for a large chunk of VFX tasks.

 

[Karim D. Ghantous]

Zack, I'm pretty sure all DSMC bodies have a gyroscope in them. I don't hear much talk about it though.

 

[Björn Benckert]

Yes, its nuts that RED does not make something like SteadXP out of their Gyro. I guess they though the third party market would make a module or such. But I think the rapid pace Red cameras have evolved / changed shape etc has held people back from developing. DSMC1 modules where not all delivering before the DSMC2 came out. People where sitting waiting for the Metzler module and all of a sudden there was a new line of cameras. Think such things push developers away.

But the new go pro has it built in so think it´s high time for RED to pick up the ball, if not to implement it in their phones...