Starting this project was a hassle as I did not have a refined plan of what I will be using or doing. Finding a rigged animated object was one of the hard parts for me. I finally got there at the end when scrolling deep into the free rigged websites. I found an animated robot egg that I thought looked intriguing and decided to make that part of my project. On my blog, I gave a short premise of my idea in the 'week 7 Perfecting look' blog post. I talked about how I wanted to have a simple scene presenting two 3D objects, one animated, the other still and an image. So I decided to go for the bin and the graffiti for my other 2 assets in the scene. Finally having an established idea I moved on to camera tracking.
Fig.2 - Lens Distortion
Doing the lens distortion was a pain. I started reading the lens size and downloading the checker. I had to pinpoint every cross to the corners of the square and made sure they were one line not connected to the other. I had to do this process about three times as the lens would not improve my footage. After having soo many issues creating the lens I realised I was using the wrong tools to connect and disconnect. Creating the square was time-consuming, however, I have found a way to do it quicker as I am now more comfortable using the tools at the side of the viewport.
Fig.3 - Camera Tracking
I went into nuke and started camera tracking which has all the details in my personal study. As normal there were so many things that went wrong, however following my blogs and having help from others' blogs, I was able to compare and change my notes to solve the problems that had using Nuke. At this point, I felt more comfortable with the software as I started understanding more every time I added a node.
Fig.4 - Exporting Nuke scene to Maya
After creating the camera tracking, I went straight to blending in the HDRI images and brightening the sky. Thus exporting it into Maya which then became the skydome in helping render my assets as shown above.
Downloading the egg from Truong's website was a very big issue as it would not allow me to save the file outside another file. The scene gave soo many limitations as I could not even change the name of the file. After troubleshooting many times, I was able to find a solution by copying the whole file and moving it to another. Although this worked temporarily it miraculously worked when moving it to the "perfecting the look" file. Thinking that my problems were solved, the scene would not let any other file such as 'FBX' be imported no matter what the file came as. Thus I decided to import the other asset to the camera tracker Maya file and render it from there. However, I had to create two AOVs from this limitation, which was hurtful but a good learning experience for me when doing render layers as I understood more from repeating the process.
Fig.5 - My Final Personal Nodes
Rendering was the most painful received from this project, this is because the computers do not have enough space to render 350 frames. I called and emailed for extra space which was never given to me, thus wasting a lot of my time as my renders would always be distracted and turned into 1KB. Thankfully I have a good PC and I was able to use my own space.
After finally several re-renders that took days, I dragged them to nuke and was able to use my notes and other sources/ references to create the right tree for my composition. Doing the nodes was an interesting experience as I finally had more grasp on understanding where they should be positioned, and how their setting should be used. As shown in my personal studies I used a lot of references and compared my notes to theirs to create my own personal node tree.
Fig.6 - Using After Effects
Overlooking my Nuke nodes and grade nodes, I rendered a video and a phew images to portray contrasts representing the before and after look of RoboEgg. I layered them perfectly in after effect and used a mask for the transitions. Having the experience of after-effects it was easy for me to follow the tutorial, thus making me see the pattern leading to copying and pasting and creating the final breakdown video. When first arriving at the tutorial, there were so many settings that were not visible in my after-effects that I gave up and created a simple video, however not achieving that goal made me feel uneasy. So I woke up the next day and went through other references such as youtube, and went back and forth to the canvas tutorial to solve my issues, thus happily giving me the right breakdown video shown below.
Overall I thought this was an interesting experience as I was able to do a lot of problem-solving, and understand the software that I have never been introduced to. What I would do differently is work more on tracking and understanding how to keep an object completely still with no little movement whatsoever. In my opinion, this was a success following the issues and problems everything came out perfect.
Comments
Post a Comment