In these examples I hope to see how suitable FDM based facial substitution might be for production.
These heads were printed at 0.07 layer height, a smidge above the minimum layer height of 0.05 on the Prusa Mini+. It is also using a standard nozzle diameter of 0.04. This can be swapped for a 0.025 or 0.02 nozzle, which technically would allow greater detail at a small scale, but at the cost of longer printing times.
The heads are about 25mm in height, scaling for a relative full body height of 115mm. This would be to a 1/12 scale which is still quite small. At 1/6 scale, the typical sizes for puppets used in large scale stop-motion productions are between 180-300mm. This means that I could print it twice as big for a 50mm head and 230mm overall height to create a character of about 4’ 7”/138cm in terms of real world height.
Printing larger would give you greater relative detail and less obvious layer lines along with more flexibility to include eye mechanics or other mechanisms inside the head. It would also make it easier to do any smaller movements in camera since everything is larger and easier to handle. The main draw back is requiring a larger shooting space with bigger sets to match.
In terms of post-processing, this batch has had a minimal amount of sanding. You can see in the below gif how the micro-particles of dark sand paper fibres have stained somewhat the PLA.
With some plastic primer and additional sanding, I feel like this level of fidelity may be tolerable for some use cases.
One benefit of utilising FDM over SLA 3D printing is the environmental and workshop safety, and given that this PLA is a bio-plastic made using corn oil it has a much shorter lifespan than resin or other rapid prototyping materials.
The limits of its use come down to factors of scale and detail. It does not hold up as well when thin shells are a requirement, as well as when you need to print smaller or more detailed surfaces, compared to SLA printing.
I will next explore printing at 200% scale (matching 1/12 scale production), then test fidelity at 0.02 nozzle diameter and 0.05 layer height to see if there is a marked improvement, especially given the draw backs of speed of production.
In future I wold also like to explore using these relatively low-fidelity prints as a ‘mold’ for thin papier-mâché shells that could allow some level of light transmission to mimic sub-surface scattering in skin.
In terms of scaling this for a workshop, I am considering using a negative mold of these heads to create plasticine casts of the faces. This would allow the participants to sculpt additional details and even manipulate the 3 relatively limited forms to create more range of expression in their in-camera animation.
Tying this back to my LII methodology (Limitation, Improvisation, Iteration): Firstly, there is a Limitation of scale using just 3 Expressions. Through the usage of the material (plasticine) there is an improvisation of additional expressions. And thirdly, the participants would have the chance for iteration of numerous versions of these 3 expressions to create a performance.
This is a similar method used by the Aardman team in order to maintain character consistency in the forms for mouth and face shapes in dialogue while enabling the animator to push expressions to emphasise and achieve the desired performance.