project[3]

lim jia sheng,
0344034.

BDCM
.Video & Sound Production
::project[3]





project[3]: Stop Motion Video

todo:

  • Narrow down 3 cool stop motion shorts
  • Create a proposal
  • Execute

research:

Finding nice stop motion shorts was harder than expected. With the child appeal, it has become a term that's super exploited & watered down. I did, however, find a few older ones that really defines this genre of art.

This first one is definitely a classic. All the animation & the surreal ambiance just makes it such a satisfying watch. It doesn't use any super advanced techniques either, just tons of patience & some clay.

For the second one, I found an animation that's as eccentric as things go.

You know it's good when it's only on Vimeo. The art style & concept of it is not only extremely creative, it is also superbly well done. My eyes were definitely blessed.

As my final entry, I had to include Ryan Higa.

This last one isn't precisely a short, or a film, but it technically has a concept & is technically stop motion xd. I just really enjoy the tongue & cheekness of it & how even "amateurs" can produce something extremely well made given enough time.

process:

As with any video related project, the first step is to plan, more specifically plan with a storyboard & proposal. We had the choice to do either something to do with "Societal Issues", or basically anything else. I chose to go with the theme, because I kinda found an issue that might be really interested to tackle via stop motion — Non-violent domestic abuse.

Proposal

It would be a short film about two figures, one which seems to be verbally degrading the other.

This is the initial proposal. Things will change.

Between now & the shooting, there were updates & such to it, but nothing particularly documented. Most notably the cutting of both front & back parts, just to focus on telling the whole story in 1 scene. I feel like the core of the message is still able to be delivered without tacking on the "nice-to-haves". Besides that, there was also the key addition of claymation to visualize the hurt done to the parties.

The reasons for these changes are however, documented at feedback.

Preparation

For the actual execution, I 3D printed my own figurines, which were a pain. This was because I have some (still) undiagnosed issue with my printer, as it would under extrude at random intervals & cause the prints to be weak.

I tried everything under the sun to fix, including:

  1. Realigning the print bed
  2. Using hotter temperatures
  3. Increasing/decreasing flow
  4. Increasing/decreasing speed
  5. Cleaning the nozzle by cold-pulling
  6. Replacing the feed tube
  7. Replacing the nozzle

3D print mess, 18/10/2021

Figure 1.1.1, 3D print mess, 18/10/2021

Even in the final result you can still see some jank, but I kinda grew fond of that aesthetic, as I'll take whatever analogy I can get to represent the figures as "broken people".

Shooting

I did my shooting across two weekends at a design studio on campus, & thus didn't document much.

The shooting set, 14/11/2021

Figure 1.1.2, The shooting set, 14/11/2021

Lighting was achieved using a top-left point, rim, & crinkly reflector, while gear was simply my good old 700D & kit zoom lens.

In order to ease the process, I developed a small workflow, as well as wrote a mini script to help things. Here's the gist of it.

  1. The camera connects to my phone via a mini-USB to USB-C link.
  2. The camera mirrors its screen, controls, & JPG versions of the taken shots to the phone via DSLR Controller.
  3. The phone connects to my laptop via wireless ADB debugging.
  4. The phone mirrors its screen to the laptop via scrcpy.
  5. The phone screen is captured into a view via OBS as a live feed.
  6. The last 3 pictures taken by DSLR Controller are piped onto my laptop via the small script I wrote.
  7. The pictures are overlaid on top of the live feed to create an onion skin.

Very long winded & probably unneeded process, but imo it made the experience of using a DSLR so much more pleasant by having the onion skin. It still didn't end me up with perfectly smooth shots, but I honestly wouldn't know what would happen without it.

Compositing/Editing

After the shooting, I went to compose it all together in After Effects. The heads were tracked on manually since AE didn't enjoy the low frame rate (10) for tracking. Then another big process (generating echos & messing with mattes to not break gradients) was done to get procedural smearing onto said heads. I have no regrets for the result of the smearing, as I think it really sells the stop motion look & serves as a point of movement when everything else chops up.

Besides that, time remapping was also a trick I used heavily to get movement to look right. The key after some proding around was to have a higher frame rate encapsulating composition, with the proxies/precomps at native frame rates (10). This made it so that the smooth parts felt a lot more natural & offset the choppiness it many others.

Layer stack in the main composition, 24/11/2021

Figure 1.1.3, Layer stack in the main composition, 24/11/2021

Sound

For sound, I did all of it in FL Studio. I just needed its versatility, plus I was simply more comfortable in it. I took lines a friend & I (cheesily) voice acted, & ran it through some trippy-ing VSTs. I also added many many atomospheric sound effects & rises/impacts to emphasize the tension building up.

The project within FL, 24/11/2021

Figure 1.1.4, The project within FL, 24/11/2021

final:

feedback:

  • 27/10/2021
    • The concept is pretty difficult to understand; it's too complicated.
    • The dynamic of them needs to be setup quickly in the beginning; the montage is probably not a good idea.
  • 3/11/2021
    • Will need to see the final product to know if the story was successfully told.
  • 17/11/2021
    • Might need subtitles to understand.

reflection:

This was a an experience threading the line of fun & frustrating. There were many moments where it really sucked, like how when the figures wouldn't print properly & numerous times the workflow would crap out due to wireless debugging being, wireless. However, the moments it shined was great, & I'm still proud of what came out at the end of the whole process.

I learnt a lot on how to execute stop motion. The main few were the compositing tricks, the workflow itself, & frame-by-frame animation basics. Whilst those were really important, I felt like the gold's in the crumbs — securing the camera & figure, what to move where & when, pacing between frames.

Even with all that though, I know there's still miles for me to crawl until I can walk or run. Stop motion is a rabbit hole to hell & back, but damn its results can be stunning if one makes it out. Will I ever do stop motion again? Probably not for fun & hopefully never again. Will I be able to do it better than before if I had to? Certainly.

Comments

Post a Comment