Advanced, Experimental VFX Animation and Techniques.
Blog Wk#20 (June 15), Neuronal Surfacing.
(and Houdini LX#2 & VFX Festival Panel Discussion)
Blog Wk#20 (June 15), Neuronal Surfacing.
(and Houdini LX#2 & VFX Festival Panel Discussion)
Surface Texture: The Basic Problem
This week focused on developing a reasonable texture for the neuronal surface and solving the rendering problem from weeks prior. My effort for this involved reference to the LinkedIn Learning (LIL) and Pluralsight websites. The LIL site provided a decent overview of assigning materials to the asset, but the Pluralsight website provided a very nice summary of how to change from polygon mesh rendering to volume rendering, which notably enhanced the quality of the render. Moreover, the Pluralsight tutorial summarized how to change the surface texture of the asset globally, i.e. beyond merely assigning a material (Malik 2018). This is important because the task of assigning a texture to a structure as complicated as a neuron by means of UV mapping is not realistic, rather something more automated is necessary.
Surface Texture: Scientific Inquiry
Most neuroscience articles addressing the surface appearance of the hippocampal pyramidal cell are (understandably) focused on highlighting details regarding dendritic complexity (Piccione Eyal et al 2018). Nonetheless, an article from Kominami and coworkers (2013) has highlighted the lamellar and lacunar nature of the pyramidal cell surface. While the term 'lacunar' tends to refer to the histological surface of bone, it does also have a more general use in biology and medicine, e.g. it is used to describe empty spaces found in plant cells and it also references a nodular variant of Hodgkins lymphoma (Klatt 2018). In other words, more generally, it can refer to the visual effect of crevices and depression-type spaces. It is notable that the Maya attribute editor for the aiNoise surface texture includes a 'lacunarity' feature and a render implementing this feature does actually generate a surface similar to a scanning electron micrograph (SEM) of a pyramidal cell surface (Malik 2018). The GIF on the above right shows the similarity between a sphere rendered to include this texture and an SEM image of a pyramidal cell surface.
This week focused on developing a reasonable texture for the neuronal surface and solving the rendering problem from weeks prior. My effort for this involved reference to the LinkedIn Learning (LIL) and Pluralsight websites. The LIL site provided a decent overview of assigning materials to the asset, but the Pluralsight website provided a very nice summary of how to change from polygon mesh rendering to volume rendering, which notably enhanced the quality of the render. Moreover, the Pluralsight tutorial summarized how to change the surface texture of the asset globally, i.e. beyond merely assigning a material (Malik 2018). This is important because the task of assigning a texture to a structure as complicated as a neuron by means of UV mapping is not realistic, rather something more automated is necessary.
Surface Texture: Scientific Inquiry
Most neuroscience articles addressing the surface appearance of the hippocampal pyramidal cell are (understandably) focused on highlighting details regarding dendritic complexity (Piccione Eyal et al 2018). Nonetheless, an article from Kominami and coworkers (2013) has highlighted the lamellar and lacunar nature of the pyramidal cell surface. While the term 'lacunar' tends to refer to the histological surface of bone, it does also have a more general use in biology and medicine, e.g. it is used to describe empty spaces found in plant cells and it also references a nodular variant of Hodgkins lymphoma (Klatt 2018). In other words, more generally, it can refer to the visual effect of crevices and depression-type spaces. It is notable that the Maya attribute editor for the aiNoise surface texture includes a 'lacunarity' feature and a render implementing this feature does actually generate a surface similar to a scanning electron micrograph (SEM) of a pyramidal cell surface (Malik 2018). The GIF on the above right shows the similarity between a sphere rendered to include this texture and an SEM image of a pyramidal cell surface.
LinkedIn Learning Arnold Materials Tutorial
While UV mapping can be a very effective means of achieving the desired texture for an asset, it may not prove realistic for very complicated surfaces such as the hippocampal pyramidal cell. The LinkedIn Learning tutorial provided a good orientation to the many Arnold materials preset options that exist for surfacing. The GIF on the left walks through the 3 basic steps of selecting an AI surface material (Maestri, 2018).
While UV mapping can be a very effective means of achieving the desired texture for an asset, it may not prove realistic for very complicated surfaces such as the hippocampal pyramidal cell. The LinkedIn Learning tutorial provided a good orientation to the many Arnold materials preset options that exist for surfacing. The GIF on the left walks through the 3 basic steps of selecting an AI surface material (Maestri, 2018).
Pluralsight Arnold Volume Render Tutorials
Assigning a Volume Shader and Texture
As per the Pluralsight tutorial by Pankaj Malik (2018), the option to render an asset as a volume, instead of a polygon mesh, improves the visual quality of the final render. This requires one to enable a volume render (under the Volume Attributes for the asset, change the Step Size from '0' to '0.1'). After a volume render has been enabled, the volume requires a shader, which then needs to be textured (specifically, it is the Transparency Attribute of the Shaded Volume which is actually textured). Slide shows for these steps follow below.
Assigning a Volume Shader and Texture
As per the Pluralsight tutorial by Pankaj Malik (2018), the option to render an asset as a volume, instead of a polygon mesh, improves the visual quality of the final render. This requires one to enable a volume render (under the Volume Attributes for the asset, change the Step Size from '0' to '0.1'). After a volume render has been enabled, the volume requires a shader, which then needs to be textured (specifically, it is the Transparency Attribute of the Shaded Volume which is actually textured). Slide shows for these steps follow below.
Revisiting the (Prior) Rendering Problem
As per Blog wk# 13, my efforts to render initial experimental efforts along with the pull focus skill found me unable to render, even though I had been able to render the wk prior! Upon further review of the Maya Autodesk website (Autodesk 2018), I opted to intensify my lighting effort, and "overlighting" the scene did solve the issue. As summarized below, the lighting necessary to view the scene render is quite substantial, e.g. I needed both a foreground and background light, both set to maximum intensity! Review of the Rendering Volumes tutorial by pluralsight ("Shading Volumes w Arnold"), also highlighted the option to increase the Arnold Lighting exposure via the attribute editor [note the Lighting section of the GIF further below, i.e. the 'exposure' entry field (Malik 2018)].
Rendering: Basics … and Volume Shading
Because my wk#13 effort w rendering was so problematic, I opted to both review the rendering topic more generally and further highlight an additional (potential) manner in which specifically volume rendering might be compromised. The top slide show (below) reviews the basics. The 2nd slide show below reviews the need to apply a shader to a 'volume render' effort, -and further assure that the shader is not fully transparent!
Because my wk#13 effort w rendering was so problematic, I opted to both review the rendering topic more generally and further highlight an additional (potential) manner in which specifically volume rendering might be compromised. The top slide show (below) reviews the basics. The 2nd slide show below reviews the need to apply a shader to a 'volume render' effort, -and further assure that the shader is not fully transparent!
Houdini LX #2
This wk found Kelvin presenting a summary of smoke effects in Houdini. As a brief overview, this lecture included basic components of: 1. the sphere source or core effect, 2. the noise attribute, 3. the disturb micro-solver, 4. visualizer tools, and 5. additional micro-solvers (dissipate, density, turbulence, and wind). At this point in the term, I need to focus on generating a showreel, but I am interested in learning Houdini and I will revisit this topic over the summer break.
VFX Festival Panel Discussion, 2020: The Next Ten Years
Continuing w the ongoing supplemental topics for the term, this wk involved the option to watch an online panel discussion addressing the changes anticipated over the next 10 years within the VFX industry. Personally, I am very curious what direction AR & VR might take and thus opted to attend the session.
This presentation from Escape Studio included the panelists Andrew Schlussel, Andrew Oliver, Will Cohen, Asha Easton, and Nicole Yershon; the panel was coordinated by St John Walker. The first theme of the session involved the relevance of developments in gaming, -which doesn't interest me. The second theme involved the nature of creativity, whether the relevance and concept has changed, and the benefit of collaboration across depts. In this regard, Will made an interesting point about how our cultural indoctrination process can be stifling, e.g. it is perhaps based on the process of persons being pigeon-holed into specific roles and then being stuck there. Persons not extending beyond that role otherwise is probably determined in great part by managers. The next theme addressed the topic of globalization, e.g. outsourcing of various scene requirements. Andrew Oliver commented on how much this has blossomed and noted that the issue of developing or acquiring components abroad has become notably less expensive.
The next theme involved the issue of whether the process of getting a college degree is still relevant, e.g. the internet has made much quite accessible. All the panelists agreed that there are advantages to getting a degree, e.g. orientation to the discipline more generally, immersion in a learning environment, the social benefits at age 19 of being at college, and the acquisition of learning skills more generally. A subsequent theme addressed the primary relevance of coders and developers and whether the future is perhaps primarily determined by tech-savvy individuals. One panelist's response did endorse some truth to this and highlighted how writers may possibly have less presence. Another panelist responded how there is a strong need for the writer to provide the primary backbone of the production, and the crucial need to further cultivate this talent. The final theme addressed the relevance of AI to the VFX industry, which also does not interest me.
Bibliography
Autodesk Maya (2018). Arnold Render View is Black in Maya 2018.
https://knowledge.autodesk.com/support/maya/troubleshooting/caas/sfdcarticles/sfdcarticles/Arnold-Render-View-is-Black-in-Maya-2018.html
Eyal G, Verhoog MB, Testa-SIlva G, Segev I. Human Cortical Pyramidal Neurons: From Spines to Spikes via Models, in Frontiers in Cellular Neuroscience. available at doi: 10.3389/fncel.2018.00181
Klatt E, (2015). Hematopathology, chapter 3 in Robbins and Cotran Atlas of Pathology, Elsevier Saunders p 79
Kominami R, Shinohara H, Yasutaka S, Kishibe M. (2013). The human hippocampus observed by scanning electron microscopy (SEM): The dentate gyrus is made of an array of the neuronal lamallae, in Okajimas Folia Anatomy Japan, 89(4):157-164.
Maestri G, (2018). LinkedIn Learning, Maya 2018 Essential Training, Create Materials: Work with Arnold Materials, available at:
https://www.linkedin.com/learning/maya-2018-essential-training/opacity-and-refraction-in-arnold?u=57077561
Malik P (2018). Rendering Volumes and Effects with Arnold and Maya. available at: app.pluralsight.com/course-player?clipld+ef01336a-b240-4317-83ff-3b0c22b9528
