Friday 27 January 2017

27/01/2017 - VFX Group Discussion

27/01/2017 - VFX Group Discussion

CJ is not present due to personal reasons.

15:40 - Group meeting begins

15:41 - Denise proposes idea and moodboard for 'TLHM_Footage_QT_Preview_4' : Replace the statue with a futuristic device in the first scene.

15:43 - Manuel proposes idea and moodboard for the same video, showing his ideas : Replace the statue in the warehouse scene.

15:45 - Thomas explains that he also had the same ideas for 'TLHM_Footage_QT_Preview_4'. The three members present, Denise, Manuel and Thomas, discuss which elements would work and how they could be achieved. Thomas proposes that, as a character casts a shadow on the statue, we create a low-detail version of that character to cast a shadow on whatever CG model we choose to replace the statue.

15:53 - The group consults Jin and is told their idea is too complex

16:05 - Preview7's shot at 2:00 is considered

https://drive.google.com/file/d/0BxkkBgxeQMUxQmZRTUlXOExIeEk/view

URL'

Friday 20 January 2017

Animation Practice Evaluation

Animation Practice Evaluation

For this coursework, we were tasked with animating a scene based on audio from monthly audio contests at the 11 Second Club. At the time of planning, the audio in the monthly contest was from Star Wars: The Force Awakens, and this audio was chosen. Unfortunately the final product didn’t turn out as expected. A glaring problem with the animation is the vast difference in animation quality between the two characters on screen. This is a result of rushing the animation due to time constraints and lack of fine tuning of interpolation. The lighting isn’t as good as hoped for as problems were encountered late into lighting and couldn’t resolve them. The final lighting looks very dull compared to what was intended.

If more time was allotted, more time could have been spent making sure the quality of animation is consistent and the lighting interesting.
Lip syncing and animation was one aspect of this project that is inconsistent. As previously mentioned one character’s animation was rushed, however the character on the right is animated more or less exactly as intended. The environment was also rushed due to poor management of time, while a majority of time was spent on animating a single character,  texturing or modeling the environment, and other aspects, were neglected, as a result it relies solely on use of flat colours and Arnold renderer materials. In future projects, time will be managed more efficiently, putting more effort into all aspects of the project, rather than spending too much time on one part. Additions such as bump mapping, more detailed modeling and more interesting lighting would have gone a long way with improving the overall quality and look of the scene.

Another complication in the process was the necessary use of the new Arnold renderer for Maya 2017 which caused numerous issues due to a lack of familiarity and experience, as opposed to Mental Ray, which, due to it shipping with both 3DS Max and Maya for numerous years, has a large amount of resources online for learning. Adjusting to Arnold took quite some time but a lot was learned during the process, allowing for experimentation with many different rendering settings that will be useful for future projects. If given more time, more time would have been spent experimenting with Arnold and seeing what can be done, improving how fast the same results or better could be achieved. Regardless, the experience gained from this project is invaluable and has given the knowledge needed to further experiment with Arnold.

The project was further complicated by render times and underestimating them. Originally the animation was intended to be 60FPS, but upon testing out different render settings, rendering times per frame were reduced from around 6 minutes to 2, Meaning it would take up to 20 hours or more to render the whole animation. Unable to do that due to time constraints, the animation was reduced to 25 frames per second and many render quality settings were lowered to get a complete render in time for the deadline. Depth of Field and Z-Depth were originally planned to be used during this project but since render times were already a problem with the limited time left, they were omitted to focus on allowing enough time to render and upload the animation.

The rendered files were imported into After Effects to be rendered as the final .mp4 file. Unfortunately because of a known bug with the specific version of After Effects used on University computers, rendering the project out was impossible without audio errors and problems.

Ultimately, this project was a learning experience that was held back by time constraints and my own failure to work efficiently in them. If I had been given more time, I’m certain I could easily build upon what I learned during this project and produced a far better animation. I believe future projects will benefit immensely from what I’ve learned during this project, from rendering with Arnold to lighting a scene.

Animation Practice: References

11 Second Club: http://www.11secondclub.com/

"11 Second Club - The Monthly Character Animation Competition". 11secondclub.com. N.p., 2017. Web. 20 Nov. 2016.

Both characters are Max rigs provided by 11 Second Club on their website and belong to Peter Starostin and James Hunt. Their original website and the rig can be found here:

http://www.bockstyre.com/james/maxformaya/

Animation Practice: 11 Second Club November Animation

The finished animation can be found here:


Animation Practice: Z-Depth

Animation Practice: Z-Depth

Rendering Z-Depth for still renders and animations is a very useful technique. It allows you to save the depth of field data from a camera used in a scene and use it for VFX work in other programs such as Nuke and Adobe After Effects. There are multiple ways of doing it but as I'm exporting my frames as .exr, I can make use of the format's ability to store Z-Depth information.

Rendering Z-Depth is done by turning on depth of field for your render camera and activating the Z-Depth AOV

The main appeal of rendering Z-Depth for use in software other than the 3D package you're using is that you can control depth of field on the fly and animate it easily in a program like After Effects without having to re-render a frame with each change. While 3DS Max or Maya will have to re-render everything just for you to see how different depth of field settings look, the process is simplified and much faster when depth of field is applied to the rendered frames in After Effects or Nuke as it is, put simply, a mask for the blur.

Lower focus distance


Higher focus distance
Just adjusting the focus distance was immensely faster than doing it in Maya.
You can even change options like where the camera is focused to achieve great looking focus movement and blurring. The blur options are very simple yet powerful, and can really improve the workflow of any animation's post-production.

Thursday 19 January 2017

Animation Practice: Environment and Rendering Process

Animation Practice: Environment and Rendering Process

Adapting to rendering with Arnold hasn't been easy, with frequent problems encountered trying to do even basic lighting set ups, I was questioning if I'd be able to make good use of it. Eventually after spending some time learning from online resources, I managed to get a hang of using Arnold and grew to appreciate how powerful it is. Features such as the real-time render window are very useful for quickly previewing changes.

I set out to make an environment that met certain criteria:

- Be designed in a way to make it clear characters are in motion.

For this, I decided to use the windows with the metal beams running down and the outside buildings for this. It's much easier to establish the characters are moving when there is a background that is clearly moving behind them.

- Be visually interesting, make sense in the context of the animation but not too over complicated

An office corridor fits this criteria, the audio relates to a job and a corridor like this provides some really interesting lighting opportunities such as sun light casting interesting shadows thanks to the window beams and the ceiling lights.

Quick and rough concepts for the environment.
The black/white tiled floor and window beams were always planned from the start, however the placement of doors, lights and length of the corridor changed over time.



The first version of the environment with lighting.

For the first pass of lighting I focused on the lights, at this point I wasn't sure if I wanted the environment to be night time or day time and used the lights to get the hang of Arnold. The lights themselves are simple meshes with an material. In this screenshot and for the next few, the lights are simple Maya point lights.
Experimenting with glass material: Higher specular value

Lower specular value
After eventually figuring out the correct settings to allow light to pass through the glass, I began working on simulating a strong sunlight that casts visually interesting shadows.


A few changes to colour correction and gamma later, the scene looks far more interesting and appealing
The same render except with the fog and volumetric environmental render settings enabled.
With the two characters added in, rendering can begin!

Friday 13 January 2017

Animation Practice: Animation Review

"Alike"



https://www.youtube.com/watch?v=PDHIyrfMl_U

An 8 minute animation featuring stylised yet incredible rendering and lighting. The animation featured in this short film represents the principles of animation, from appeal to anticipation. Making good use of lighting and movement, the animation adjusts tones frequently as it goes between the monotone, colour-devoid streets of the city contrasting against the beautifully rendered colourful nature. The composition present in this short film is very effective, splitting the screen in half to show how the day of the two main characters differ.




Overall the film tells an interesting and charming story through the smart use of colour and composition., with a fitting soundtrack to compliment it.