Studio Project!!

Week 1

I was unable to blog for the past few weeks due to the lack of time. Hence I shall try my best to fill this post with as much details as possible. We started off the first day by presenting our idea to MR Ron. Our idea was almost immediately approved and we started planning out the shots, storyboards and looking for reference images. Initially we were thinking of having the Golem form out of pieces from the field but we realised that it would be very tedious to do as well. As we searched for reference images for Golems we also realised that the materials of the Golem should reflect those found in its surroundings so we took some pictures to use as textures on location.

On the second day we finalised the story board and I proceeded to make a Leica reel.

On the third day we tried doing some test shots to plan out marker placement and camera angles. However, we ended up having a slow SD card and the only way we could record video was to shoot at 640x480 in a 4:3 ratio. we were thus unable to perfectly plan out the shot angles on this day due to the wrong aspect ratio. At the end of the day we tried to match-move the footage but the quality was too low and the markers were being blurred out way too much and we had a very difficult time getting any working track.

On the 4th day we tried to shoot the final footage but we were unable to finish our shoot due to bad weather. The light angle and shadows were changing rapidly between shots as well and in the end we had an unfruitful day. We also managed to capture our Mocap footage today thanks to Mr Leong and his his help in setting up the shot and software. 

Finally on Friday of the first week we shot the actual footage and I took over as cameraman on this day. I started trying to Match-move the second main shot we were focusing on but Match mover was having a lot of trouble solving a nodal pan despite having 32 tracking markers. Mr Ron came down and suggested using NukeX to Match-Move but all of us had trouble exporting the tracking data to Maya and we did not manage to figure out how to do so.

Meanwhile the Golem was slowly taking shape though there were a few conflicts on whether to use a high poly, procedurally displaced mesh from from Houdini or to use Mudbox and sculpt in the fine details on a low poly base mesh. We ended up using the former option as Samson found a way to have the geometry displace only at render time and managed to reduce the load of the scene.

Week 2

On Monday, we were still trying our luck with match mover to try and get the Match-Moving done. We managed to solve the camera for the third cg shot we were focusing on but the other 2 were having lots of problems as they were nodal pans. I helped out to texture a couple of pieces of the Golem after Danial taught me the basics of Mudbox. It was a fun experience and I see myself using it more often now after having witnessed its power in texturing. I havent explored the aspects of sculpting yet but I intend to do so in the near future. We were unable to solve the shots at the end of the day and we had no choice but to reshoot again. This time I suggested using red and white tape instead of the regular printed trackers we were using as the contrast was too low and when it was too far in the distance the quality was too low for tracking.

On Tuesday, we reshot the two main CG shots again and we used red and white tape in place of the trackers this time around. I started match-moving my shot and immediately there was a big difference in the ease of tracking. I still had to manually track every 4-5 frames for certain tracks but other tracks were going much more smoothly.

For the rest of the week we started doing the layouts for the shots and started integrating the cg into the shots. While waiting for Samson's simulations, I started Rotoscoping some of the shots so that we would have less things to worry about later. The 3rd match move shot in particular was especially tough Roto-ing Afiq's hair as he was running was very tedious. I ended up doing a rough Roto as I had other tasks to attend to as well.

Week3

We started assigning materials and lighting and started to prepare for rendering. I helped to assign the materials and bump channels for the golem for the 3^rd match move shot. The mesh of the golem also seeps into the ground and I fixed this by grouping the mesh and animating it off the floor on a separate animation layer. I also fixed the same problem for some of the other shots. Finally on Wednesday night I managed to light and add materials for the second match move shot and rendered it. However the following day I found that the render went wrong because the pixel aspect changed by itself due to some Maya bug. I tried locking the attributes and this created a whole new problem where I couldnt view my render settings anymore. After lot of deleting preferences and trying to reopen the scene, I tried to import the scene instead and this time I managed to get it to work and after having made some of the passes again I started rendering only to realise after 40 frames of rendering that the whole animation was no longer playing due to some cache problem. I then reimported it again and luckily everything worked out fine this time. However I was left with very little time to composite at this point as it was in the wee early hours of friday morning. I managed to pull through somehow in the end fortunately.


This was by far the most stressful project I have done and there was always constant pressure piled on my shoulders. Though still with quite a few problems, Im glad we managed to pull through somehow and we could have worked better if we found ways to improve the compatibility between maya and houdini and had more time to tweak the simulations and other aspects of this project.