One spatial reasoning test I did not fail

I have tremendous respect for CGI. At this point, with enough money and digital artistry, any set of images which can be imagined can be put on screen. I also have deep psychic scars related to sub-par CGI work in projects I've shot.

In any case, I am a dedicated proponent of doing special effects in camera. If a particular shot can be accomplished by either method, I'd rather do it on set. On my most recent commercial, the director was of a similar mindset, only more so. We had several SFX shots and the post schedule was too tight for CGI.

For one shot in particular, I'd normally have expected to hear "We'll do it in CG." or at least as a composite. However, the brief was to do it in camera.


Dry noodles fall at an angle and land nicely in bowls, one after another. 
The director's main concern was creating a "flavor image" that nicely showed the texture and color of the noodles. This is no small task for dry noodles. The staples of food photography, translucence and specular highlights aren't going to be part of that picture. My main concern was how to get three (later four) packets of noodles to land nicely in their bowls on cue. 

Using plexiglass slides to drop the noodles in was suggested. I was concerned that the lighting would be constrained to avoiding reflections and shadows rather than making the noodles look nice. Also, dry noodles are brittle and break easily (in retrospect, cushioning the fall with broth might have been a visually interesting solution - I wish that idea had occurred to me in the preproduction, though to be fair the brief called for looking down into empty bowls).


More importantly, getting a thing to land is a specific way on camera just once is hard enough. We needed this four times in a single shot on a tight shooting schedule.

Which brings me once again to lessons learned from my teacher and mentor Chris Chomyn. He shot a couple films about a chrome sphere that flies around trying to kill people called Phantasm III & Phantasm IV. You can take a look here. All of the sphere effects were created in camera. Chris takes his students through the process he and his team took in achieving those shots. What he covers in that discussion has carried me through all the SFX shots I've encountered thus far.

What I find helpful is that it taught me to think backwards. One example is starting a shot at its end. Film the action in reverse. Or if you're shooting digital, reverse playback in the edit. That was the obvious solution to my noodle challenge.

After that little leap it became conceptually simple and physically achievable. Start with the noodles in place. Take it backwards. Since forward is now reverse in a shot controlled by gravity, up becomes down.


Rough reference frame taken on location. 


We took a double for our kitchen table and set the frame we wanted. We glued the bowls in place. We drilled holes in bottoms of the bowls.  We ran monofilament through to the noodles to hold them in place. Then we turned it all upside down and set a frame. 


Reference frame we took in the studio once we were ready to place camera. 

Since we wanted the noodles to enter frame at a dynamic angle relative to the plane of the table, we hung a plum line and angled the table until we had a vector we liked. At this point the disconnect between what was on the monitor and what we were physically doing started to get surreal. 

I had to match lighting from shots on location. This involved flipping a lighting setup upside down and getting the camera up pretty high. It's one thing to conceptualize and another to execute. Especially for a guy that struggled with those spatial reasoning tests you take in middle school.




I'd love to say that once we were set up, everything went off without a hitch. Unfortunately there were a couple bumps, mostly involving the stands holding up the table. Fortunately, a reset simply involved reeling in some fishing line. In the end we got our shot. 

There is far more to thinking backwards than reverse playback. But it's a good place to start.


If I had this shot to do over again, I'd try to find a method with simpler execution. Or I'd introduce a dynamic dolly move and pose myself a real mind-bender.

No comments: