About time for a new contest?

novolume

novolume

0
About time for a new contest?

Just thinking that with version 6 now released, Cheetah finally getting some publicity in the App store and subsequently new users around it would be a good time for a new contest to get some fresh new images to promote Cheetah with. What does anyone think?
 
Hi David,
I thought about that too. Maybe a dynamics contest would be a good idea. v6 is almost a month out now and I saw almost no usage of the dynamics feature yet. Although so many wished Bullet physics integration.

Bye
Martin
 
I also expected to see a lot more renders since the release of v6. Maybe everyone is just busy creating things at the moment. I like the idea of a dynamics contest though. I'm pretty sure even if you put up a relatively low value prize you would get plenty of entries.
 
novolume said:
Maybe everyone is just busy creating things at the moment.
Click to expand...

yeah... that is valid for me at least: work and family keeping me busy lately (I also stayed away from upgrading to avoid any distraction... go figure!)

but that's a good idea for sure and I do think the contest should be focused on one or more of the new features
 
There is an animation that I've been working on, but haven't had time to finish (is anything ever really?). Or actually it is done, I just did a more basic version to test it out first though. I'll post it up when I'm somewhat done. It won't be an award winner of course. Maybe it'll get the ball rolling though.
 
Had a bit of an idea on this one...

Over on the GreyscaleGorilla site they regularly run 5 second motion graphics projects/competitions centred around a word. The beauty of a 5 second project is that it's a pretty level playing field - its not a huge time investment for anyone that's really busy and it focuses your efforts on creative ideas and excellent execution. Also Martin gets the added bonus of using the better entries to form part of a Cheetah3D showreel.

So how about a 5 second dynamics competition?
 
Okay, well I went ahead and knocked it out. 5 seconds sounds good, but I already had this one set up to 10 seconds. It'll take about 30 minutes to upload to Vimeo though. I do wish I could have made the flag material a little shinier and translucent. But haven't been able to figure out how to mix an alpha image with an material node shader. I did post in another thread about that. Just haven't had more time to play around and test it. This is kind of a straight rip off of the example file for a waving flag. I did use your tutorial video to help me place this though Novo. So you do get some credit. I probably could have done a better job if I had gone into the exif data on the photo and matched the camera to it. It looks decent enough though. I'd like to make the grass look a little bit more like it's growing up around the post. I guess the word for this could be called "work"? It was one way I can get away with doing this some during my work day. ;)

https://vimeo.com/42151580

Any advice after viewing is welcome.
 
novolume said:
Had a bit of an idea on this one...

Over on the GreyscaleGorilla site they regularly run 5 second motion graphics projects/competitions centred around a word. The beauty of a 5 second project is that it's a pretty level playing field - its not a huge time investment for anyone that's really busy and it focuses your efforts on creative ideas and excellent execution. Also Martin gets the added bonus of using the better entries to form part of a Cheetah3D showreel.

So how about a 5 second dynamics competition?
Click to expand...

agreed! :icon_thumbup:
 
Newton's Cradle. Took me a while to figure out how to set up the "rope" properly (edit: spline must be set to "linear"). The wire might not even need to be turned into rope...? Just was guessing it might work better. They still aren't correct though. I need to figure out how to give them more force so it works correctly. No time to upload to Vimeo, so I'll DropBox it for now. I used one of Archie's Parquet materials for the table (thanks Archie!). It definitely needs a lot of work though. Pretty sloppy model with missing details. I'm attaching the file if anybody wants to examine it. Surfing, I think you did a chair once that had that bent metal tubing right? I'll have to go back and see if I can find it. It will help me do the bars the wires are mounted to more correctly.

http://dl.dropbox.com/u/117426/Newton's Cradle C3d.mov
 

Attachments

  • Newton's Cradle Bitmap.jpg
    Newton's Cradle Bitmap.jpg
    67.8 KB · Views: 685
  • Newton's Cradle.zip
    116.2 KB · Views: 900
Last edited:
Hi Martin, I think it might be a thing of people trying to learn how to use the new features first. I'm sure eventually there will be a lot of cool submissions using the Bullet Physics. Thanks for adding it, I think it's pretty great!

Martin said:
Hi David,
I thought about that too. Maybe a dynamics contest would be a good idea. v6 is almost a month out now and I saw almost no usage of the dynamics feature yet. Although so many wished Bullet physics integration.

Bye
Martin
Click to expand...
 
Ok, I've removed all but the ball and wire from the scene for now so I can focus on the physics. I realized my original file the wires were angled poorly. That caused some problems with the kinetic force not passing through the balls correctly. I've mostly solved some of the problems I had in the original. My main problem now is the anchors for the balls to the wire. They are still behaving more spring like than I want. They should be more rigid. So the kinetic force is still not passing through the balls as it should. Increasing the mass on the balls helped some, but not enough. Also after reading the wiki article about it, I realized that in the real world the steel balls actually have some slight springiness to them when they hit each other. So I increased the "elasticity" of the balls a little to mimic this. I'm attaching the updated file. If anybody has any idea how to anchor the balls more correctly so they don't move about so much, please let me know.

Thanks!
 

Attachments

  • Newton's Cradle 2.zip
    27.2 KB · Views: 836
Swizl said:
... If anybody has any idea how to anchor the balls more correctly so they don't move about so much, please let me know.
Click to expand...

Hi,

Just use the 3 points as anchor points for the ball instead of just the middle one. That way, the ball has to stay at an equal distance of the 2 upper points.
 
Hi Francois. Yes I'd already tried that way. All the objects fall when all three anchors are set together as a group in one anchor tag. Not sure why. I just followed the example files that came with C3d anyway. I referred to the "Soft Body Anchor" and "Rope" to try and figure out how to do it. Maybe there's a sub-setting to get all three anchors to work in one group? Not really sure. I tried changing the "influence" on the anchor, but set low it makes all the objects drop. Set too high and they all just disappear when hitting the play button. I might have to make a "Y" shape with the spline with the bottom leg folding back on itself so there is a bottom point to attach the ball to. I'm open to any other ideas / alternate ways of getting this to work right.
 
Ah thanks Francois! I guess I didn't think about using the same points in multiple anchor tags. Definitely works better than it did. There still a problem with the balls drifting apart right as the animation starts. I have no idea what would cause that though. Shouldn't they remain perfectly still until something hits them? I tried putting an attraction force on each ball in the middle, which helped a little, but then the kinetic force is lost early on. It would be better to figure out why the balls are drifting apart in the first second of animation. There's one of these cradles done in Blender. I'll have to go take a closer look at how it was done there. I know it won't translate over exactly, but may give me a better idea around some of these problems. Thanks again for the input, it's greatly appreciate it! :D
 
Hi,
Swizl said:
... with the balls drifting apart right as the animation starts. I have no idea what would cause that though. Shouldn't they remain perfectly still until something hits them? ...
Click to expand...

Martin could better explain that than me but that has to do with the accuracy of the simulation, object and margin overlapping.
In the help file, have a look at Dynamics/Accuracy, Rope Tag/Margin & Solver iterations as well as Rigid Body Tag/Margin and Activate.
Try with a Dynamics Accuracy of 8 and you will see that the movement is a lot smaller than with an accuracy of 2.
 
Yep, maybe he'll have a minute to take a look. There are so many different settings to adjust that it made my head spin yesterday trying every combination I could think of. Most of them I have somewhat of an idea of what they do, I'm just not 100% sure exactly how they work. I didn't change any of the Dynamics accuracy, but I will when I get a chance.

Here's a tutorial for the same thing done in Blender that works pretty well.
http://www.youtube.com/watch?v=ozipVZ9MkZU
I tried copying some of what he did, but it's still not exactly the same. The scale of the whole model does have some affect on the physics. I guess the Blender physics work on about 1 Blender Unit = 1 Meter. I would guess that C3d maybe is about the same? I even added a gravity tag to the whole scene and I'm not sure if it really helped or not. Like I said, too many variables going on for me to be sure what is working and what isn't. Thanks again for the response. I'm determined to get as close as I can to solving this one though. I nearly got it working right, but the problem is that it would loose momentum and only give me about 3 swings before stopping completely.

Edit: I guess I should also add that I figured out my second model the balls weren't lined up evenly and the pivots on the points on the splines were all wacky. I checked them in a side view. I did fix those, but it only seemed to help a little bit.
 
Last edited:
Increasing the Dynamics helped a huge amount! It's still not perfect, but probably good enough for now. Maybe Martin has a few more suggestions to tidy it up? Anyway, I've uploaded the video to Vimeo and I'm attaching the updated .jas. Waiting for other submissions. I already have some ideas for some others I'd like to try. Just a tip about the Dynamics settings. If you copy and past the model into a new file, the bumped up settings will not come with the file. They will default the standard 2 instead of the max of 8.

https://vimeo.com/42518531
 

Attachments

  • Newton's Cradle.zip
    113.2 KB · Views: 904
Hi Guys,


I was playing with the dynamics and made a little test for animating a fish, swimming through an anemony, waving in the water...

see video here

This is just a test for now, I'm planning to make several improvements and add a complete rigged and animated clownfish...

Any remarks or suggestions are welcome...


Kind regards,
Filip
 
Last edited:
You must log in or register to reply here.
Back
Top