Category Archives: Uncategorized

The Proposal

Posted on by

Aim

To build on the skills I had developed in 3D visual effects during the last semester and to produce a short 1-2-minute Showreel as my final artefact, with the purpose of expanding and improving my portfolio of work, to attain this I will be utilising Houdini FX, Maya, Photoshop and Substance Painter, as well as some work in Premiere Pro to compile my work into video form for my final artefact.

Objectives

• Further develop my skills within Maya, Photoshop and Substance Painter by producing a large scale, high poly model

• Fully texture and composite my model into a still image by utilising the Arnold renderer

• Build foundational skills in Houdini FX to create and render procedural terrains and simulations • Fully document my process for display on my blog

• Conduct research into the aesthetics and critical theory behind my work

Strand

I have decided to select the portfolio strand for this project, my reasoning behind this is that the research and development behind this project will be quite robust as I still have much to learn within Maya and Substance Painter, and also this will be my first attempt at learning the Houdini software package. Therefore, I would like a heavier weighting on my R&D rather than most of it being placed onto my final output.

Abstract

This project is intended to show further development into my modelling skills within Maya, as well as building a strong set of foundation skills in Houdini, mainly through experimentation with the robust set of shelf tools available within the software. The project will mainly consist of practicebased research as well as some critical theory work, which will then lead into my final Showreel
which will be my final artefact for the project. I will be sticking to a photorealistic theme throughout my work, as to meet the industry standards on what a Showreel should consist of. Whilst I have experience working with Maya and other modelling packages, Houdini and FX work will be something completely new to me, my reasoning behind wanting to work with Houdini is that it is currently very in demand within the industry, therefore I would like to get a base knowledge with the software, so as to better prepare me in case I work with it in a professional context.

Audience

The audience for this project will be both potential employers and educators in the VFX industry. My aims and objectives for this project are based on the role of CG Generalist, which typically requires knowledge of both Maya and Houdini, therefore I would like to show my learning capabilities and problem-solving skills in these packages.

Platform

A short 1-2-minute video of my compiled work which will be posted on Vimeo, as well as various screengrabs, sketches and time lapses of my work which will be hosted on my blog.

This project will extend my creative skillset by further building on my knowledge of the modelling process (Modelling, Texturing, UV Unwrapping and Rendering), as well as requiring me to learn an entirely new software package (Houdini) alongside of this. The biggest challenge of this project will certainly be learning Houdini, which has a notoriously high learning curve. Whilst this will be a difficult process with certain failures, I intend to use these problems as a learning experience as well as developing my problem-solving skills at the same time.

Initial Ideas – Semester B

Posted on by

For semester B, I wanted to use my experiences and knowledge gained from my previous project to dictate where i was to place my efforts going forward. While I did enjoy my work in modelling with maya, I wanted to further branch out into 3d and try my hand at something similar whilst also wildly different and experimental.

Recalling a conversation late last year with escape studios, one of the key points made was that currently, the industry is looking for FX artists, more specifically people trained in Houdini.

hou_BuildingUserInterfaces_lesson001

The Houdini interface.

After some initial research into the Houdini software package i found where it differs from other 3d packages is that it focuses on procedural generation. Whilst standard modelling tools and procedures are available within, where Houdini shines is its ability to produce complex dynamics, particles and volumetrics, which are typically used to create simulations (E.g, Fire, smoke, water, Etc.)

Looking into a few houdini showreels, it becomes apparent just how powerful the software is.

Doing some research into the learning process, which will take up most of this blog, according to various VFX forums, Houdini a notoriously complex piece of software to learn. Employing the fail-fast approach from last semester, I will begin immediately working with the software and attempt to have at least one output, no matter how small at the end of each week, failing this I will go into detail what exactly didn’t work when trying to achieve this goal.

Thinking towards an output, I have considered at this point making a compilation of simulations made within the software which I will work on throughout the coming weeks, working on my feedback from last semester I will try to include all of my experiments within this output.

Final Showreel

Posted on by

Reflection

Camera Mapping

This shot is definitely my least favourite part of the showreel, while it does work, it doesn’t do well to show off the textures since it wasn’t properly rendered and is just a recording of my viewport, which also prevented me from adding any post processing effects such as depth of field, which would have greatly improved the scene. The improvements to this scene that I would make would obviously be to figure out the method of getting my renders to work, either by trying a different standalone rendering engine, or updating my version of Arnold / Maya.

Tracking

Very pleased with how this one turned out, since I managed to get my solve down to 0.18 as well as accurately lighting the scene and correctly applying my textures on the signpost. To improve this scene I would definitely consider adding more models and doing further work in texturing to make it look more realistic, but for the purposes of showing off the accuracy of my track I think the shot works quite well.

Modelling

Incredibly proud of this shot, whilst the topology of my model isn’t very clean, the final render and textures do very well to sell the realism and enable to model to retain a lot of detail in close up, I’m especially happy with my edge work around the analogue sticks, as I think they do a good job of reflecting the light in the scene. To improve I would first clean up the geometry, making sure to get rid of any triangles or ngons, and also I would do some more texturing work, such as adding some more normal detail to create a groove around the base of the controller.

Final Shot – Camera Map

Posted on by

Problems Encountered

The main problem I found for this scene was that when rendering none of the rendering engines for Maya were working properly for my scene, either the render came out as a completely black screen or alternatively looked fine within the render view but got ridiculously dark when saving the image sequence. To remedy this I first attempted to google the problem to see if any other users were encountering the same problem as me, as I found, there were multiple posts about this exact problem, solutions ranged from changing the gamma settings to reconfiguring the lighting of the scene, I tried most of the solutions given but none seemed to work on my machine. I then took the files and attempted to render on my housemate’s PC, only to find upon pressing the render still button, Maya crashed multiple times. I was very frustrated during this process as I had made a few test renders of previous camera maps which had worked fine.

I then contacted my tutor to see if he could help me in remedying the problem, after looking for more solutions he found a forum post stating that a user with the same problem had contacted Solid Angle who informed him that it was a known bug with Arnold renderer and that a future update will fix the problem. My tutor then recommend that I look into screen capturing my scene and animation as a workaround.

After some messing around within the software I found that I could just use the playblast feature within Maya, which renders out an image sequence using the standard viewport, while the sequence was obviously not as high quality as an actual render, it worked quite well and was sufficient enough to be placed in my final showreel.

Reflecting on this, I have to take responsibility for leaving my render so close to the deadline, as well as not properly researching render engines in Maya, in the future I must remember to look into this at a much earlier point. Whilst it was a very hard lesson to learn, it has definitely affected how I will approach future projects.