Lightwave .lwo reader SOP
#13
Posted 19 October 2012 - 01:06 PM
#14
Posted 16 November 2012 - 02:38 AM
#15
Posted 16 November 2012 - 04:45 AM
I could probably work on a matching shader that would try and duplicate the basic lightwave material, with basic procedurals and such.
#16
Posted 18 November 2012 - 12:16 PM
#17
Posted 21 November 2012 - 04:49 AM
Will try to pin it down some more, over the weekend.
Check it out, as is, for the moment, but last I checked, there were quite a few cases where is fails.
It's all python, so If anyone has the inclination to chase bugs, be my guest.
eetu, on 16 November 2012 - 04:45 AM, said:
That would be dead cool.
I went from this, to attempting to do the same while leaning C++, and got quite far.
I figure ,Ideally this should be done with HDK, but my skills in that regard are questionable.
I have a good grasp on parsing the .lwo file, so perhaps with some HDK guru's guidance we could make this sop solid.
Attached Files
Edited by rom, 21 November 2012 - 05:27 AM.
#18
Posted 22 November 2012 - 12:34 AM
on call asset and on load geometry gives a mistakes : ))
#19
Posted 22 November 2012 - 08:47 AM
The cabin one errors on load.
The film noir one loads the geometry but errors in new_Color and doesn't load the material info.
The Halloween one loads in ok.
Seems like it wants a c:/tmp folder, too.
I'll torture it a bit more
But, cool in any case!
#20
Posted 22 November 2012 - 11:15 PM
eetu, on 22 November 2012 - 08:47 AM, said:
If I remember correctly, the c:/tmp thing is where a log file is written, in order to trouble shoot. I think there may be a DEBUG tab, where you can change the logging features. As a Mac user, I probably hard coded that path to log file (my bad), I'm sure if you track that down, you can check the log as to where the Sop fails. I figure that this will help solve the 'boundary conditions'.
After going down the rabbit hole on parsing all the material data, It did appear to be far less stable, so this is really the area that needs attention.
I will do my best to give this some more attention soon.
#21
Posted 23 November 2012 - 02:24 AM
#22
Posted 23 November 2012 - 10:38 AM
since AD products do not oscillate on the same frequency as my brain and logic does , i really hope to see this idea take a solid form .
so thanks in advance to anyone involved even in trying for it .. =)
{ yep , seems a bit 'strange' to see LW's layout of parameters inside Houdini , but as i said ; this might be a good candidate / reference for 'the_ubershader_material' ( i think ) Houdini needs }
--
Quote
on call asset and on load geometry gives a mistakes : ))
hi skazkaPiter ,
accidentally im trying lwo.otl in the same H12.1.33 ( WIN_7x64 ) and im having no problems .
.cheers
#23
Posted 27 November 2012 - 01:11 AM
Turns out it was failing mostly due to vmap names not being compatible with houdini's attribute names
currently the node parms for surface layers are only set up for the color channel, so it will also throw exceptions when trying to set layers in other channels.
Although it should be able to load large files, it is not currently optimized for them.
The initial intention for this was not to just load old lwo files, It was more to use Lightwave Modeler to quickly create elements that could then be subjected to houdini's procedural treatment, providing the best of both paradigms.
For now it is probably also wise to 'cache' as a native houdini .bgeo as the lwo is not reloading, when the scene reloads.
Attached Files
Edited by rom, 27 November 2012 - 01:16 AM.
#24
Posted 27 November 2012 - 06:51 AM
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users













