MOCAP Sessions 6 + 7

The 6th mocap session was productive, as the cameras were connecting up to the computers, and so a successful ROM test was run. Simon is now in charge of cleaning up the data on the performance day, and Jack is to work with Motionbuilder. Since the notes taken when being taught by Geoff were vague, it was hard to understand what to do when it came to putting the data in Motionbuilder, and so the session was ended early.

Lucky for the group, a Tuesday session was booked in, and during tuesday, wednesday and thursday sessions, two staff members who are familiar with the MOCAP room, were available for help. This was beneficial as it helped explain the missing steps that the group missed when running through the procedure. There were certain steps in the procedure that were essential, that was not written down in the notes, and these were then updated. Cleaning of the data was simplified, meaning that more time was had to work in Motionbuilder. This would become useful for the group, as the time before the performance will be limited.

Since working with real-time was explained too quickly when Geoff explained it, the notes taken were very brief and very hard to understand. The staff members in the room were around to help, but they had not covered the subject much as well, so experiments were carried out to try and get the feed live. After many attempts, and mistakes of removing an actor from Motionbuilder, the feed finally worked in real-time. The problem was that whenever the real-time plugin was started, the actor in Motionbuilder would just shrink into a heap on the floor, and not match any ROM data. The solution is explained in the revised notes, which can be found here: Mocap Room Documentation. Because of limited time, the marker set was not set up properly, meaning the character jolted and became jumbled up in itself when working in real-time. By creating a proper marker set in Motionbuilder will help keep this data smooth, and hopefully cause no problems. Since it took a while to find a solution to importing the character that Jack and Alex had made into Motionbuilder, it meant that little time was had to experiment in real time, but now this was working, it was proven that the group may be able to achieve a good, working performance, and next time, know what to do, so the group can show Alex Counsell. Even though this step was meant to be achieved a month ago, many problems were encountered. The group have two months to get the performance ready.
Tasks for this week include:
Frances - Work on the animations for the start, end, and inbetween levels. Animations will also need to be created if the character falls, or gets hit, or dies.
Alex - Finish the aesthetics of the ropebridge level, add the textures, and continue working on the mine cart level. This should be aimed to be finished in 2 weeks time.
Jack - Continue with texturing of character, and finalise rigging.
Simon - Help out with cleaning up of data, and texture the cave level.
Helen - Continue updating website
Greg - Continue production process

MOCAP Suite 5 - 13th March 2009

A few sessions had been booked before this, but these had both been unproductive, as the two computers in the MOCAP room could not connect, losing valuable time for the group. An email was sent to Alex Counsell to fix this, and by this session, it was completed. Since the group did not take any pictures or remember where each of the markers went on the suit, it made it very difficult to place them. It was found out in the end, by loading ROM test data from another group, and finding where the labels went. This cost us time, and meant that we could not clean up the data in time. Since it had been a while since we had cleaned up data, we had forgotten what settings to use, to fill in the gaps, but this was to be solved by simple practice. Because of the little time we had left, Greg had to email the MOCAP room managers and book in more slots. We had found out that one of the other group's had booked in monday's and fridays up until the 30th March. This meant that we'd be left with little time and would almost mean the group would fail the presentation. It was found out that Tuesdays and Thursdays could be booked in, only in certain circumstances, and these were explained as little time left, and due to technical problems. Permission was also given by Alex Counsell, and the group managed to book in another 3 sessions on a tuesday and thursday.
Because we only have 3 hours in the MOCAP room, it gives us little time to setup, capture the data, clean the data then work in real-time. To get this to work, alot of practice would have to be put in, so that by the end of the 3 hour session, we would be able to have a real-time character in Motionbuilder.
Alex had finished the ropebridge level, and showed us. The group approved of this, and would make a perfect level for the game. It would be that if the character was to walk on a certain plank of wood, it would crumble away and the character would have to go another way. This would be done by using an animation trigger. Since Frances was in charge of the animations, she had preferred to create them in Flash, so this may have to be done in Flash, but it will be sorted soon.
Simon had completed the first cave level, and now matched the original design/layout of Frances's original concept designs. If possible, we would have pieces of a key to collect in this level, for the door at the end to open.
Tasks for this week involve:
Alex - Work on creating the mine cart level. Greg had already practiced the animation of a platform to move in Motionbuilder, and should be simple to get it working.
Jack - Learn to texture in motionbuilder, create textures for the character, and advance on the rigging of the character.
Frances - Work on the flash animations (Start, End, Between levels (maybe) and animation triggers (if character fails etc)
Helen - Continue on website
Simon - Help with cleaning up with data in MOCAP room
Greg - Help with cleaning up with data in MOCAP room and continue production process