In what Brecht showed me there was no pinning involved, just seams and impressive results. Maybe it was pelt mapping after all…
I will keep you posted on this one. I own Zbrush and this kinda just came out
http://pixolator.com/zbc/showthread.php?t=029628
It drops the model to the screen and sends the image to photoshop for painting and then back to zbrush. Zbrush has either you own custom made uv’s or there own AUV auto mapping unwrap. So it basicly means zero unwrapping needed. But a pelt map is better to have for the smarter visuals…
Either way a direct to gimp tool would work wonders for blender…
Probably ABF+, even better results normally than LSCM because it respect angles in the mesh.
And i already said it, but it is worth repeating, LSCM is perhaps a bit more manual work, but it produces the guaranteed best result for one set of pinning and seams.
From what I’ve seen on the 3Ds demo, there is quite some level of stretching involved. there is a lot of hype around this algorithm because it is 3DS, it produces nice results too, but all demos i’ve seen (for the 3Ds tools) needed a relax UV phase afterwards. this means the mapping is not that good initially
[quote=“lukep”"]
errr but tuhopuu also has a uv smooth feature to help in the lscm stuff…
On another note, Just cause we dont have something does not mean we dont need it and can do the same exact thing with what we have in the same amount of time. That is mostly the point, time, in the pro biz all you have one dead line after the other 
Learning how to use a tool is great, having to tweak and tweak and tweak and then give up and start over sucks… But it goes with the territory, BUT it would be nice if it could be fixed bit by bit instead of naysaying it.
That said, if you check the modo forums they are not really having that great of time with pure absolute unwrap magic either. Though still much faster than we have, they are more likely to have high stress and paying jobs. So if we could get what they have I am sure we would be super happy as is…
Ok now you can start taking my post out of context here
just to add to the discussion - this tool is for the moment maya specific, but I’ve entertained ideas of trying to port it to a java based standalone (although I don’t know when this will be).
the springs on this implementation are not just “on the outside.” each vertex is connected to contiguous vertices by a spring, and the user has control to increase the number of springs to even non-contiguous vertices. I don’t know how 3dsmax’s implementation works out, but in general pelt mapping is dependent on the placement of seams. the actual spring calculation is relatively trivial, and increasing the spring count in the base mesh can go far in keeping down uv distortion.
pelt mapping is thought to be a point of departure for finessing uv’s, but i’ve had a fair bit of success generating also my final uvs through the pelting simulation.
cheers,
sunit
Aw sunit ! I dragged ya from the other forums 
Have you kinda ever thought to invest some time in giving a proposal to Blender CVS team or a python script ? I dont blame you if it is not an ideal situation for the time being, I hear blenders sources are kinda perplexing and time consumeing.
But anyway you can releaseing the tool to the open source world would be sweet. In my view LSCM is a good tool but really like a developer made tool still not optimised for day to day artists folk that just want to get stuff done. Adding in your tool could give Blender a super boost in the work flow.
I think it can be done easily with the new rip tool and some hooks and soft bodie, going to try now 
Ups, I found a bug, can´t try it now 
http://projects.blender.org/tracker/index.php?func=detail&aid=3285&group_id=9&atid=125
I think it can be done easily with the new rip tool and some hooks and soft bodie, going to try now
I already tried it with hooks, but it didn’t work. But using the wind field works!! Here is an example of pelting a very simple mesh in Blender:
Blend file: <edit>Blend was done in 2.37a</edit>
Just press ALT-A (cursor over 3d view!) to start the animation.
I manually ripped apart the Mesh, extruded all the vertices on the ripped seams, and placed them in the area of the wind fields.
The pelted mesh is of course only geometry and not a UV map, but that should be a piece of cake to convert in a Python script.
The rest also should be easy to automate. The only problem is deciding in which direction the extruded vertices should be pulled. Else it can happen, that two veritices from close places on the mesh get pulled in completely different directions.
But else it should be easy to implement for more complicated meshes with proper seams.
But we already have LSCM, so i won’t bother anymore 
what did you do to rip apart the mesh ???
what did you do to rip apart the mesh ???
I did that manually, by making a new edgeloop on top of the other, moving the vertices that should be ripped, and then removing doubles. Though there will be a new ripping tool in 2.40
Let me repeat, that this is only a proof of concept that this method would work with softbodys. It would have to be automated a python script of course. Done manually, as it is know, it isn’t useful in any way.
Done manually, it is certainly more work but the results could still be better than LSCM. For subdivision UV maps, you have to make a heavily subdivided duplicate anyway to avoid texture stretching so you can rip apart the mesh as you like. Then if you get it to the state as you have done, you just do a planar UV projection. But I agree for a complex mesh, doing that manually might be a little too much. It also might take a while to calculate on slower machines.
mmm, no way, it got to work with hooks, just place the hookes below the soft in the modifier stack
[quote=“ZanQdo”]
mmm, no way, it got to work with hooks, just place the hookes below the soft in the modifier stack[/quote]
Yea, I guess you are right. I didn’t notice hooks were addad as modifiers.
Well, when 2.40 is released I’ll check the Python API to see whether everything we need is there, and then we could think about writing a script.
…
Ok tested it now. I just can’t get it to work with hooks. 
But maybe thats only for 2.40alpha1
I read in the developer mailinglist, that softbodies will work with armatures. Maybe that will also be the case with hooks. Anyway, there will be something better than these wind force fields.
Hi all,
Just to update - this tool is now released:
http://www.sunitparekh.com/pelting
MOVIE DEMO
http://sunitparekh.com/pelting/movies/demo_0_2_h264_mid.mov
It’s been in beta for about two months, and in development since early this year. There are a number of studios that are using it for production, and a group of individual users who’ve been part of the beta team. Here’s a quick run down of some of the features.
FEATURES
- Mesh and Component Part based pelting
- Multiple pelt groups in one scene
- Direct connection to original, uncut mesh
- Two different pelting frame algorithms
- Per vertex spring controls
- Multiple model views - checker, baked normal
- Edge coloring on cut and border edges
There are a bunch of videos on the site, as well as how-to tutorials.
regards,
sunit
maya… Will you try and port it to silo or blender or java ??
WOW! THAT is soooo grossss and coolll for mesh…
I find it irresistbly funny that this is just being created now. When I first got into 3D art, I was appalled that this was not in the feature set of any major 3D packages. Nice work, but I’d say it’s about damn time!
That looks like a really useful tool. The fact that it’s free is even more incredible.
cept ya need maya to use it
Well, I guess the MEL script could be rewritten in Python with some help of the python team (i think all of the used softbody functionality is present in Blender), but the question is “do you want that?”… Because I really think Live-LSCM is MUCH easier! You don’t need to fiddle with all kinds of sliders, settings etc, you just need to know what to pin and how to move - and that’s not difficult at all. While pelt mapping gives a great initial result, tweaking may become quite hard after all: with Live-LSCM you can do it right straight from the beginning!
What lukep is talking about is AFB++ btw, which you can find a paper on here: http://www.loria.fr/~levy/publications/papers/2004/ABF_plus_plus/abf_plus_plus_temp.pdf
if that’s really what Brecht is working on, that would be spectacular!
Besides that. 