View Full Version : Oh my god!

04 April 2002, 07:24 AM
eh eh, I started 3D not long ago and was using Max. I finally got the XSI 2.0.2 learning CDs and can only say one thing : wow! Maybe XSI is a bit longer to learn but all I can say is I'm kinda hooked! Who needs Max now? :D So much cool features. Just the f curves deformation are enough to switch... Cool... :thumbsup:

04 April 2002, 08:59 PM
I second that! I use max too and I prefer XSI for so many reasons!

Bernard Lebel

04 April 2002, 03:43 AM
Glad to hear that you are having fun.

04 April 2002, 05:34 AM
Thanks Ed! I'll be looking into your tutorials pretty soon. Nice stuff! :D

04 April 2002, 07:21 AM
Thanks a lot. :)

04 April 2002, 07:58 AM
[note from softimage guy] Thanks, your check is in the mail [/note from softimage guy] :D :D j/k

04 April 2002, 03:05 AM
haha that's hilarious rendermonkey :)

04 April 2002, 07:31 PM
before i got to college i was playing with max during highschool last years and got kind ok in it, then went to college and started learning maya4 and xsi 1.5.
yah 1.5 sucked compared to maya but i still liked it more though learning maya was easier for me. but now with 2.0 i don't see how xsi is worse than maya. so what there is no paint effects, they are just for fun anyway and no one takes the seriously for serious projects. so what xsi doesn't have live trackign system, in studeos they still do it somehow so there is a way. and so what that maya has lil bit more features. just wait and see what XSI3.0 will be like. probably another quantum leap in 5th dimention :)

04 April 2002, 03:12 AM
Aren't they going right to XSI 4.0 and skipping 3.0 totally? Maybe I missread that somewhere but I thought it was funny when I read it. Something like they want to appear to be closing the gap with Maya in terms of revisions.

I like, well, I use XSI and I have to say it is great but man o man our group and the other group in my class have run into some really nice annoying problems that sadly our professor can't help us with. He's a maya guy and so were kinda outta luck. He agrees that XSI is very cool but from what we have encountered it still needs a few more minutes on the ironing board.


04 April 2002, 08:19 AM
Tellerve can you give us some examples of what unresolvable problems you encountered? just cuirous

04 April 2002, 07:02 PM
Tellerve directed me here, I thought i may give you some examples...

Check out the attached animation, I would have liked to give you all 40 frames so you could see that it is fine the rest of the way through, but alas, 50k isn't much to work with. Youll see that some of the vertices are being ripped out of the ear. That is where a shape clip is tilting the ear forward to a more alert postition. Only some of the clip-effected vertices are tearing... the rest are functioning perfectly. The clip wieght is not animated, just set to 1. Its the rig animation that triggers the distortion. There doesn't appear to be any rhym or reason to it. I thought perhaps the verts were straying into other envelope territories, but that isnt the case, there is no bone to pull them that direction at that time, and the distorted vertices seem to be random, not proximity based.

Iv'e tested many theories thrying to make it work and this was the best result i found. The big problem is that the whole character has been bandaided along our 6 week animation adventure and we have no instructor to help us through the technical side of xsi. Not to mention the deadline which is a terrifying month away. ..we would start over with a fresh 'model', but we do not have the time....

Symetry is also unresponsive and rotating clusters go berserk when in proportional mode, and only in proportional mode. And why wont the D@mn vertex colors go away when you remove the envelope and clusters from the mesh?!

Xsi has amazing modeling tools and im even getting used to 'her' render tree, but 'she' is so damn specific, and if you dont know precicely what she wants, she takes you out...

Any help would be appreciated, Tui

04 April 2002, 09:33 PM
And why wont the D@mn vertex colors go away when you remove the envelope and clusters from the mesh?!

You might be experiencing a known bug that keeps in the scene a temporary cluster. To get rid of it, export your scene in dotXSI 3.5, import the .xsi file in a new scene, open the Explorer, expand the object's clusters, and delete the cluster named "EnvelopeSeletedVertx".

Hope this helps
Salutations - Cheers
Bernard Lebel

04 April 2002, 10:20 PM
merlin9876: ok, since you have now gone trough this enlightenment why don't you tell me where i can order those learning CDs so i can see the bright light too, huh? :p

05 May 2002, 04:21 AM
errr, well... Since I live near Softimage headquarters I could go knock on their door... :rolleyes: Not sure they'd answer though... ;) If this doesn't work, I'd try their website. :thumbsup:

05 May 2002, 05:22 PM
Its very frustrating working like this... after posting this as a last ditch effort, i gave up... (still, any help to correct the problem would be great)

I decided that we would have to work with a simplified system of facial animation. We are going to try and just use static expressions at different 'cuts' of the story. You know, like he falls and bounces down a hill with a cute little grimace on his face, then when he shakes his head, we will cut to a replica scene where ive modeled his face into a more dazed look. It should come out seamless and we can take advantage of active moments to hide and 'in-shot' swaps.

Well, you would assume that if you edited the mesh, ie: model a grimace, that the mesh would still deform properly on the envelope (all the face vertices are enveloped to the same 'head' bone). Well, as he walks away from 0,0,0 in the scene, the edited points stretch out, like pinochio. This has got to mean that there is something wrong with the transforms, or something (i tried freezing the transforms...). I also froze the move component on the mesh, but that would delete the envelope one way and distort it even more when i reordered the operations stack.

This same issue accured when i tried cluster wieght mapping. There must still be options and i will continue to search. I will update anything that might help you help me :)

Thanks in advance, Tui

05 May 2002, 05:56 PM
I was reading another post and was inspired... i think i found the problem...
The first pic is the way it should look on the rig...
The second pic is what happens when i 'disable from here' the envelope op in the explorer. (obviously a problem and probably the cause of my shape truama)
The Third is the result after reseting the transforms, freezing them, and 'enabling from here'... im not suprized...

Any suggestions on correcting the transforms are welcome :)

CGTalk Moderation
01 January 2006, 05:00 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.