Around March - April 2020, I was involved in a really exciting production based around the science of creating packaging from proteins.
I wanted to be part of this, this is really exciting...
Initial steps were to produce ideas as a scamp, the initial draft was formed along with a quick explanation as to how the animation would unfold. Unfortunately, this wasn't quite enough for my client, my rough scamp, my thought pattern didn't translate so well.
With that I had to resort to my basic and very rough drawings. Probably the last time I drew anything decent was 35 years ago! 35 years ago GULP.
The second phase was forged (pre-lockdown) whilst lounging around in the local coffee shop waiting for my 6 year old daughter to do her thing in the dance studio. The amazing aroma from the coffee and cake helped give birth to a storyboard.
As I write this blog I reflect upon "How times have changed...", that might be another story.
We arrghhhh go, budgets are confirmed...
The hard work is yet to start. An animatic, a spider, a plant, rigging, IK chains, how do you portray protein? How do you portray amino acids?
You get the picture. A lot of hard work, and using my new array of tools, Cinema4D (C4D) along with Redshift, and a little Octane Render.
The pressure was seriously on...
This is where C4D stepped up and took the crown, the procedural nature, the artist friendly approach meant rapid iteration and that is exactly what I needed for a certain part of the animation.
Fortunately, I had a year working at an agency where my MODO designs would take form in C4D, so I knew my pipeline inside out, and if I got a little unstuck, I had a back-up plan.
One of the more complicated things to realise, this is what this blog will focus on, was the structuring and unfolding of the companies IP. How do you realise this science so it can appeal and make the science a little more simplified for a wider audience?
I had to take a punt on how to visualise and unfold a Polypeptide chain (obviously a little reference was supplied).
That didn't quite work out, in reality the chain(s) are very unstructured, very random, differing lengths and this had to be portrayed to be credible.
2nd attempt - not bad, I am close, but...
So I need to step up and create chains with altering length, back to the drawing board. A bit later in the day, 3rd attempt is formed with timed offsets in each chain.
Still not right, a few scribbles from one of the scientists showing each chain is unique. So I need to take a step back, a rethink is needed...
During this phase a conference call was setup, verbal description is required, and this leads to getting closer to the scientists vision for this 16 second part of video.
This time, independent length, more erratic movement with slowing down when amino acids try to form a bond.
Very nearly there
I have now created several splines, morph targets (to unfold), a cloner (amino acid) along with secondary animation (erratic / undulation movement) and it takes 1 more iteration to get it right.
The final result as seen below :-
This whole process for creating and rendering using C4D / Redshift was quite a joy. Redshift was always predictable, it is very fast when couple with 3 GPU's (although it is never fast enough).
Geek fact: I have resorted to building an eGPU - an external frame which can house an array of Graphics cards.
I have used Octane extensively with previous renders, but sometimes, when dealing with specular materials, it can take a long time to resolve those parts during rendering. Redshift was much more optimised for this problem, however, there are other gotchas, but they are minor in comparison :)
Having options, knowing the limitations of render engines forges a smoother path.
What I have learnt from this particular gig is that C4D allowed for random setups, rapid prototyping which meant animation was fun to produce, also and probably more important, it was predictable.
Thanks for reading, and what you have read about here was worked 16 second sequence taken from a 1 minute 30 second production.