Major Project: Soap Bubbles Update & Texture Detail Tests

A few weeks back I began struggling with my soap bubbles when I decided I wanted to render them separate from the environment. Due to this, I decided to postpone working on this scene until all of my other shots were sorted. Some of my problems included dark spheres appearing in the reflections of the bubbles despite having enough samples and rays and the transparency not working the way that I wanted to once the passes were brought into After Effects. I brainstormed list of what I could do to fix this - but as I opened up the file I decided to render out some test frames and bring it into After Effects for one really simple test. I changed the 'Mode' on the bubbles from 'Normal' to 'Screen' - and they worked the way that I had originally wanted. 

It was a very simple change, which is frustrating but also relieving. It seems to work fine this way - the bubbles are rendered on their own but also contain the reflections from the room. This should allow me to blur the background without blurring the bubbles as well as apply depth of field blurring to the bubbles only. I still need to test this to be completely sure that it'll work properly, though. I will set up a render tonight to see how it works.

I also decided to do some more reading on the bubble preset and found this article about how to animate the texture so it looks like the soap iridescence/specularity is fluid. I decided to do a few tests on this just in case it adds an extra layer of detail and realism to my soap bubbles. I played around with some of the settings to see what results I could get.

Bubble Fluid Animation Test #1

Bubble Fluid Animation Test #2

Bubble Fluid Animation Test #3

I quite like how it looks, but I need to test this further since I'm using particles and an instance to animate my bubbles. I'm unsure if the animation on each bubble would be exactly the same. Luckily, if I render the bubbles separate from the environment anyway, it wouldn't take too long to tweak the bubbles. I will do an initial test first and see if there are ways of randomizing it by having the particles rotate if need be.