For polishing the animations after creating the first iteration we had to conduct a lot of testing which also involved playtesting with different groups of kids to get feedback from them in whether they can understand Tom’s character and if the animations help in making the shows interesting and engaging. Polishing and Testing therefore played important roles in helping us make Tom look more alive and make the shows more engaging.

For polishing the animations, we began by reviewing the first iteration of animations on the animatronic as well as in Maya to ensure that the animations are smoother. This test helped us in knowing if the changes that we made to the animations worked well as the first iteration included the implementation of the different speeds and acceleration using the different range of motion for the animations. We also were testing out the different frame rates for the animations as per recommended by Tim Eck, our alumni adviser. Therefore, we had converted the animation files for frame rates 30, 45 and 60. After testing we determined that the ideal frame rates would be 24 frames per second or 30 frames per second and we decided to go ahead with 30 frames per second.

After reviewing the first iterations on the animatronic, based on the observations Niharika started working on polishing the animations using 30 frames per second. We then decided to test out the second iteration for playtesting with different groups of kids. However during our play test the animations looked really smooth and the kids were able to identify Tom’s character and that he was there to perform for different shows however they did mention that they wanted him to move more as he looked really slow. We also noticed that the animations looked a lot slower and made him look old and more humanly rather than an energetic puppet. Based on the feedback Niharika then started working on the 3rd iteration.

During the 3rd iteration Niharika started polishing the animations, increasing the speeds and making the actions for all the functions faster. We also decided to get some reference videos with our creative director Dustin acting as Tom which would help Niharika in adding performance to enhance the animations. But after our play test Tom had to be taken apart for reassembly to implement the mechanical changes made by Alex Gobeler our mechanical engineer and since there are just four team members and none of us have a mechanical engineering background we decided to focus on the reassembly so that the animatronic is ready for shipping and also help in electrical. Due to this Niharika was unable to test the 3rd iteration till the final day of shipping.

On the final day of shipping after testing the 3rd iteration we discovered a major problem that had not been identified before. The animations for the torso and the scepter were switched which meant that the animations for the scepter were playing on the torso and for the torso were playing on the scepter. This was the main reason for the slower animations during our play test with the second iteration. Another thing we discovered was that after switching the torso and scepter animations back to the correct channel in the show controller software we tested the 3rd iteration of animations with 24 frames and 30 frames per second and realized that the animations looked a lot smoother with 24 frames per second. This helped Niharika a lot as she then decided to go back to 24 frames per second as the use of 30 frames per second was making the animations a bit slower and robotic after we fixed the issue with the scepter and the torso. But since we were not able to test all the animations properly as the animatronic then had to be shipped, Niharika started working on making half of the show files at the speed using 2nd iteration and half of the show files at 3rd iteration speeds so that she could test onsite and compare to determine which speeds work the best for 24 frames per second.

At Give kids the world village we then tested the 2 different shows to see which looked better and based on everyone’s feedback especially Alex Gobeler our mechanical engineer and Tim Eck our alumni adviser, we finalized that the second iteration would be the best along with a mix of some animations from the 3rd iteration to add different acceleration speeds for the animations. This would help represent different character traits like Tom would react suddenly to certain things or when he is thinking he would move slowly which would help in making him look more alive and like a performer. The testing at onsite was the most crucial one as we discovered a lot of problems like the interactions of Tom with audience, volunteer and the other characters in the castle was not working as Tom would not look at them and was bending forward a lot which made him look more towards the floor rather than the character. We also discovered the issue of snapping on transitioning from the animations as with the user interface it had to be made easier for the volunteers and due to the idle animations being longer at the end of each dialogue a rapport could not be created between Tom and the volunteers which was crucial for our show.

We then began fixing the issues discovered onsite. We started first by fixing the line of sight for Tom for looking at the audience, volunteers and the other characters. Since this was happening due to the torso bending forward, Alex Gobeler had already created a smarter mechanical design which could be modified to fix issues like these. She then suggested to move the torso a bit backward in terms of the mechanical design as the the design can be modified and she had added 3 different positions as options for the torso to be either moved backward or forward. After implementing the suggestion we moved the torso one 2 positions backward which helped in fixing the line of sight for all of them and she had to polish the animations a bit for father time. But in this way by testing they were able to identify the issue, fix it and test it again. Niharika had to also keep in mind that since kids are shorter then Tom would always be bending forward while interacting with them and also the volunteer but with Father Time (another character at Give Kids the world) since he is at a height Tom had be in his resting position so that he could appear as if he was looking a Father time. Similarly with Joel the troll (another character in the castle) since our main story was that Tom has staring contests with him, it was necessary that Tom would stare at Joel whenever talking to him. Another major issue was the scepter hitting the head during some of the animations and to fix this Niharika had to limit the scepter movement towards Tom. Based on these observations we then finalized our speeds, accelerations and range of motion for the animations. Niharika then worked for the second week of install on making all the animations consistent in terms of speed so that she could then test the final iteration.

For the final iteration we then started testing different shows as part of system testing integrating audio systems, scepter lights and fixing the snapping issue to test out the timings, interactions during the different shows. We then also invited the volunteers and asked them to operate the shows which helped us in gaining further feedback regarding any minor changes that were needed.

Therefore it is necessary that after polishing the animations, the different shows should be tested out on the animatronic as testing helps in identify any problems that may occur due to new changes made which is crucial and helps in fixing these issues that we may not notice if we do not test the animations.

Categories: Weekly Blog