Re: .xsi format trashes UVs

[Luke <luke(at)rsp.com.au>]

Hi Raffaele ;-),

I'll walk to your desk and chat to you about this in a minute ( now you 
are in the same room! ), but it is worth posting so everyone else knows 
about it.

XSI incorrectly stores UV data within its SDK ( based on the examples 
from the SDK on how to access it ) and as a result, there are a LOT of 
issues saving and storing UV data between it and other packages.

We have been running into a lot of issues lately with this and it is 
especially apparent when you use UV's in the C-SDK.  From Python the 
problem is easier to solve.

Only way that we have ever been able to get this working is to export as 
a .obj and import using .obj

From what I can tell, any manipulation of a mesh fries the sample 
indices in such a way that they are, from that point forward, useless 
from the SDK ( the indices get scrampled and the elementArray that 
"should" unscramble them isn't correct ).  Since the dotXSI format must 
have been written using the C-SDK ( makes sense for speed.... ), it also 
suffers from this problem.

The only reason I think that .obj handles itself properly is that when 
it exports the mesh it traverses it in the same order that the SDK 
examples indicate it should, which means the UV's are written out as you 
would expect if you were accessing them upon import or manipulation from 
the SDK.

It's a big pain in the arse but it can be worked around.  Once I get 
some time myself and Moritz will send XSI a repro for the problem, since 
it is not the easiest thing to demonstrate.

Good luck.

Luke

P.S.  If you are up for a bit of a coding excercise, another workable 
solution is to write a nearest neightbour comparison between two meshes, 
one with the OK UVs before the export to maya and one with the fried 
UVs.  You can find the same points in space and then copy over the UV 
data deviod of issues.  It is a simple thing to code that should take 
you about an hour or two.

ozu's mailbot wrote:

> dotXSI will only keep the UVs if there is an image map attached to it.
>
> oO
>
> Raffaele Fragapane wrote:
>
> > it would help to know how you're transfering data.
> > anyway... I've transfered heaps of stuff between maya and XSI in the 
> > past, and when it's only geometry and UVs you want to move across OBJ 
> > will do the job perfectly.
> >
> > Matthias Worch wrote:
> >
> > > I'm having the hardest time trying to get a UVed model from XSI to
> > > Maya 6, the UVs get lost during transfer no matter what options I use
> > > (the model doesn't have any UVs once it's imported into Maya). Has
> > > anybody experienced this? If so, how did you fix it?
> > >
> > > --
> > > Matthias Worch
> > > http://www.langsuyar.com
> > >
> > > ---
> > > Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in 
> > > body:
> > > unsubscribe xsi
> > >
> > >
> > >
> > >  
> ---
> Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in 
> body:
> unsubscribe xsi



--
luke emrose
technical director - luke.emrose(at)rsp.com.au
----------------------------------------------------------------
rising sun pictures - www.rsp.com.au
redefining visual effects delivery
----------------------------------------------------------------
direct line +61 8 8400 6400
mobile ph +61 408 893 505
----------------------------------------------------------------
our adelaide phone number & address has
changed, please update your records..........
----------------------------------------------------------------
adl ph +61 8 8400 6400 - fx +61 8 8400 6401
level 1, 133 gouger street, adelaide, 5034
----------------------------------------------------------------
syd ph +61 2 9338 6400 - fx +61 2 9338 6401
15/16 charles street east redfern, sydney, 2016
----------------------------------------------------------------
rising sun research - http://research.rsp.com.au
----------------------------------------------------------------


---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi