Jump to content

grafikzeug

Members
  • Content count

    4
  • Joined

  • Last visited

Community Reputation

4 Neutral

About grafikzeug

  • Rank
    Peon

Personal Information

  • Name
    Felix Möckel
  • Location
    Germany
  1. Optical Mocap Solutions

    Hey Emmanuel, no experience with Vicon (though quality-wise nobody ever said anything bad about it), but I set up and operated an Optitrack-rig that covered about the same area as you need. After a lot of consideration we decided to go with Optitrack Prime 13 and 13W cameras; 16 cameras in total. The most important reason to choose this over Vicon was the price. The Prime 13 cams don't have great resolution, but the price is unbeatable and unless you want to capture fingers or face markers (if so, do yourself a favor and get higher resolution cameras!), you should be fine, even with the 13 series. Quality of the hardware and software has been very high and the compatibility between the different Prime cameras has always kept the door open to enhance the stage with a few Prime 17s if need be (we have never bought one, so I can't say if the claim holds true). Generally the cameras do a great job at capturing at 120 fps, I can't really recall doing higher framerates, but I believe the 13s go up to 240 fps, which is very fast if you want to capture human motion. We mostly did regular motion, nothing too fast, but in the few jumping/ falling recordings the data showed amazing detail. We did some realtime tests outside of Optitrack's Motive software, feeding the data live into Motionbuilder, where it was directly retargeted. It was a bit fiddly to setup, but worked fine afterwards. The latency wasn't too great though; maybe a lag of about 1/6 s. We never tried any optimization on this. If you would like to go into more detail on what you're planning to do (and what space you have available), I might be able to give more feedback. Cheers, Felix
  2. VEX Robot: movement & rotation issue

    Thanks a bunch, f1! I just read this after making it (sort of) work. Your approach is definitely more straighforward, even though I did end up using the detail attributes to store the data for the whole creature on. It took me a while to understand that the thing was ripped apart because not all of its points had meaningful values on the necessary attributes that I was trying to use for locomotion. Since this is a project in order to get going with VEX, so I'm always thankful for feedback or hints! Here is the current status. I've included the hip if someone is interested. Right now I'm happy that it moves at all; the next steps being a general sense of direction (so it doesn't just turn back when it looses the trail, but is a little more determined ;)) and then see if I can make it avoid obstacles (and other ants). Cheers, Felix Ant_v14.hiplc
  3. Procedural robots

    Hey Ezequiel, really cool looking designs and thanks for sharing a bit of your process in the video. Very inspirational! I had to think of the cool micro-detail subdiv stuff that Simon Homedal built (there is a tutorial doing similar things here: https://www.youtube.com/watch?v=mXhKjB6-ENc). That might provide some cool fractal texture that's fit in nicely with the overall structure of your bots. Cheers, Felix
  4. Hey guys, new here, nice to meet y'all! I'm new to Houdini and have been dabbling with VEX recently. It's a love hate thing. Being interested in social insects I got the idea to build a little "ant robot" that could follow a trail. It has a body (point) and three points that serve as sensors. These sensors sample the ground's trail attribute (red, works), compare their values, find the "winner" (works) and then the whole thing should rotate accordingly and move along a vector that goes from the body to the best sensor point to stay on track. (Works not so well). That's the status and at least it's doing some following the path, but I can't get the robot thing to rotate around a custom position. Instead, the matrix rotation always uses the world origin as a pivot which looks increasingly bad with distance and at a certain stage just breaks. Also, the body promptly severs itself from the rest, which is a tad unfortunate, as I wanted it to follow the sensors (stay attached) and serve as a the pivot for the rotate(). I know this is a little messy but maybe one of you would like to have a look at the hip and point me to the obvious mistakes I made. Much appreciated! Cheers, Felix Robot_movement_issue.hiplc
×