This week besides working on my Learning Goals, I also tested out some of the ideas that might be relevant for our project. The main goal was to see if those ideas are overly ambitious or not and how feasible they are for the product.
Shaders
Since we discussed the possibility of creating a winter environment for our product, I tested the feasibility of this idea. As a result I found out that creating a snow shader was possible and accessible within the timeframe in case we decide to follow this idea. Additionally, while creating the shader I learnt more in depth information about nodes that could be used for other shaders as well (for example "Absolute World Position" node which allows to extract the data about the position of the object in the world and do manipulations using that position).
Since for our client it is important to create a widely accessible product, I also started researching into how to optimize shaders, so that we could use our product on low end devices (devices that have a low processing capacity). My current research insights are that there are 3 optimization methods:
Deleting the unused parts of the shader
When possible changing the heavier nodes with less heavy math combinations
Packing multiple grayscale textures into the RGBA channels (refers to the red, green and blue channels of the displayed pixels) of a singular texture
Combining multiple nodes in order to avoid specific nodes to be repeated and processed multiple times
Real time effects
I continued building up on my effects creation knowledge. This week I focused on movement, aspiring to create either natural and accurate movements. Within this context I created a generic movement effect and a dandelion effect which could be used as a base for a potential atmospheric effect for the product.
Houdini
This week I started learning about particle simulations and procedurally simulated geometry . These are the easiest simulations to get started with and would allow me to progress into creating more advanced simulations such as fire or smoke.
Comments