PDA

View Full Version : cc sds 'disconnection' - still not fixed?



greent
07-17-2007, 01:52 PM
are there any chance that you fix problem with cc sds and edge weighting when adding new geometry, that object does not 'disconnect' itself? new version is about to come out but this annoying bug is still present throughout two major releases.

Surrealist.
07-17-2007, 03:34 PM
It is not rendering like that is it?

For me it is usually a display problem. If I up the CC Subdivision Level in modeler then set it back, the problem goes away.

Also if I have parts of the model hidden and I toggle subpatch it can do the same thing from my experience.

What are your renders like?

greent
07-17-2007, 03:48 PM
well, it is not display problem nor that trick with subdivision level helped. this is what is going on: http://www.greent-online.com/lw/ccsds01.jpg.
this problem will appear in any geometry detail add operation.

JamesCurtis
07-18-2007, 08:38 AM
Have you reported this to the fogbugz bug report site. If not, you should!

mav3rick
07-18-2007, 02:55 PM
that is old habbit

Surrealist.
07-18-2007, 09:29 PM
well, it is not display problem nor that trick with subdivision level helped. this is what is going on: http://www.greent-online.com/lw/ccsds01.jpg.
this problem will appear in any geometry detail add operation.

Sorry, I misread the first time. I was seeing it as two separate issues. I get you now.

PixelDust
07-19-2007, 10:46 AM
It also happens if you edge weight one half of a model and mirror the geometry. The weighted polys on the mirrored side will be broken like that too. It's extremely frustrating.

Sensei
07-23-2007, 09:02 AM
TrueArt's EasySplit v3.0 is the only tool either 3rd party or NewTek built-in that properly preserves Catmull-Clark Edge Weight while adding new geometry..
It can be seen in action on this video:
http://www2.trueart.pl/Products/Plug-Ins/EasySplit/Graphics/Movies/EasySplit_Edge_Weight_Preserved_1.mov (please Save File As.. and watch off-line)

evenflcw
07-23-2007, 02:31 PM
Well done, Sensei! It's a shame NT didn't get this right from the start themselves.

greent
07-23-2007, 06:44 PM
no offense to newtek intended, but i think that you should contact engineers at nt and explain them how to program such things since they don't have sligthtest clue how to fix this to be functional.
sure, that would mean that they have to fix most of tools to work with cc sds surfaces, but then again, why advertise such feature if it doesn't work like it should?

GraphXs
07-23-2007, 08:01 PM
Sensei

What version of Easy Split works with CC's. I have version 2.7 and it breaks the polys when I split it. It will reconnect the edges if I cut over a broken edge, but it also makes that edge hard.:thumbsdow

See image.

voriax
07-24-2007, 12:38 AM
sure, that would mean that they have to fix most of tools to work with cc sds surfaces, but then again, why advertise such feature if it doesn't work like it should?

Sad but true.. To me, catmull-clark subd's in LW are more of an embarrassment than a feature :(
When they work they are great, but they are just such a hassle. I'm still surprised that even after the entire 9.0 beta cycle AND so much of the 9.2 beta cycle that it hasn't been addressed.

mattc
07-24-2007, 02:22 AM
Sensei

What version of Easy Split works with CC's. I have version 2.7 and it breaks the polys when I split it. It will reconnect the edges if I cut over a broken edge, but it also makes that edge hard.:thumbsdow

See image.

Version 3.0, just like he said.

M

evenflcw
07-24-2007, 03:37 AM
that would mean that they have to fix most of tools to work with cc sds surfaces

This is probably (I hope) the main reason it wasn't done. It's just bad if they have to fix each and every tool to get it to work right. It's likely that they are still designing/developing ways this could be handled internally somehow or via one simple function call. Unless they do it that way, they do need to fix every tool specifically and every 3rd party developer needs to develop their own solution (=few who develop for free will bother). They do need a proper solution!

Maxx
07-24-2007, 05:52 PM
Personally, I'm hoping the next point update focuses on Modeler - we've gotten lots of nice shiny new layout bits, but modeler has been woefully neglected so far. I may just be wishful thinking, but I'm hoping for tool consolidation, actual CC integration, and better UV work flow in 9.3.

greent
07-24-2007, 06:08 PM
i hope so too. but there's also ca problems to be solved so i think that nt should employ some more staff :-)

GraphXs
07-24-2007, 08:01 PM
Do'h! My Ed"Ah"ma"K"shun is kicking in. Sorry:twak:

Karmacop
08-01-2007, 01:03 AM
no offense to newtek intended, but i think that you should contact engineers at nt and explain them how to program such things since they don't have sligthtest clue how to fix this to be functional.
It's not that they don't have a clue how to fix it, they are fixing it the right way (fixing the old modelling core) instead of creating a temporary fix.