Week #19

Advanced, Experimental VFX Animation and Techniques.
Blog WK #19 (June 8), Houdini Fluids

Intro to Houdini

This week found us receiving a special introduction to the topic of Houdini Fluids from a new instructor, Kelvin Simmonds. While attending his tutorial, the obvious relevance of the fluids topic to my (general, on the back-burner) interest in trying to incorporate a 'diffusion tensor imaging' (DTI) component to my music-video-brain-film project was obvious. This was most apparent when he serendipitously stopped the animation of fluid flow during the moment when the fluid path lines were apparent (IMG below). I subsequently presented a few brief inquiries during his talk which were oriented toward my curiosity about use of fluids to mimic DTI tracts . And, I emailed Kelvin w further inquiry re modifying various details of the 'fluids algorithm' to more specifically achieve the effect I am interested in.

Fluid Path Lines Approximate
Neuronal Projections

As is apparent in the IMG on the right, retaining the 'string-paths' of individual fluid 'molecules' (in an animation designed to simulate fluid flow down a shelf w staggered shelves or dividers) actually mimics a neuronal projection (Yamamoto et al 2007, Soares et al, 2013).

Diffusion Tensor Imaging Brain Art

The aesthetic relevance of diffusion tensor imaging (DTI) to my music-video-brain-film project has been obvious to me for some time, but I have not encountered a technique that would allow me to animate the projection of a fiber from a starting to an ending point. Comparing the following IMG to the screen shot from Kelvin's lecture, the rationale of approaching this thru the window of "fluids" is obvious, and indeed, involves the capacity for an animation of fluids to start at some specific point (e.g. in the brain) and target some predefined endpoint in an animated manner, i.e. mimic a live path that a neuronal fiber path might follow.

Alternative (Non-Fluids?) Approach to DTI Brain Art

Following my email, Kelvin was helpful with offering a different strategy for approximating a DTI effect by means of Houdini. The following approach is very good. It eliminates the liquid quality of fluids and readily allows for visualization of individual fibers. I look forward to learning more about this strategy.

Houdini Applied Skills

In addition to attending the class session for the Houdini Intro, I did watch the entire recording a couple times. My first effort w generating some sort of fluids clip is on the right. My main interest w this theme is as described above, i.e. modifications of it that allow for an approximation of animated brain DTI visuals, which I will revisit during our extended summer break. In general, I didn't find the Houdini interface very troublesome, i.e. it was perhaps more intuitive to me than Maya.

Again, my main goal is to move onto the applications of relevance for generating brain DTI visuals. Kelvin was very helpful w sending a tutorial dedicated to this issue, which I have watched in part (screen shot is on the R). To delve into this topic w greater attention won't be realistic until the summer, e.g. the term is nearing its end and I don't have anything ready for the showreel!

Bibliography

Soares JM, Marques P (2013). A hitchiker's guide to diffusion tensor imaging, in Frontiers in Neuroscience. Available at doi: 10.3389/fnins.2013.00031

Yamamoto A, Miki Y, Urayama S, Fushimi Y, Okada T, Hanakawa T, Fukuyama H, Toagshi K (2007). Diffusion Tensor Fiber Tractography of the Optic Radiation, in AJNR.