You aren't too old. But you're talking about an industry that is flooded with lots of new talent every year. It's the latest "dream job" that every tom, dick, and harry are pursuing. The market is currently over saturated. And people are still pouring in by the thousands. You have to be exceptionally good at it to stand out in the crowd. And artists are commonly the first to be tossed out when things are slow.

It's sad that they aren't doing traditional animation anymore. 3D is only cool when there is something else to compare it to. There needs to be a balance of media produced for them to stay interesting. I once read where the animators for Disney were treated like slaves and forced to crank out work under really long hours. Then tossed out when the movie is completed. Today on the news. They were talking about how Pixar needs to crank out more than one movie per year now under Disney. That kind of thing really puts this whole 3D hobby into perspective for me. It's a fun hobby. But I'd really hate to have to be creative while under high pressured deadlines like that. Creative artwork done on a tight deadline is not my idea of a good job. -ScottA

I heard on the news that Disney has just paid over 7 BILLION dollars for Pixar! -ScottA

Poser still uses the same old fashioned coding it had in version 1. If you read up on programming 3D software and rendering engines. You'll find that it's a fairly easy thing to write a program. But it's a very hard thing to make the software efficient at handeling resources. Reboot was produced by a Canadian company that used SoftImage. I loved that show too. SoftImage is an expensive program that has much better coding to handle a professional workload. They also used render farms. Not just one single computer. Something Poser still can't do. Poser is great for what it does. But it's long overdue for an entire re-write. If you use a program like C4D for example. You can visibly see how huge the difference is. The program sort of floats on your machine using almost no resources at all as you work in it. That's just due to good quality programming. -ScottA

It's more along the lines of thick headedness than arrogance. Haven't you ever worked with an engineer before? They're absolutely fantastic at creating things that work on paper in a mystical fantasy world in a galaxy far, far away. But don't work under real world conditions until one of us "morons" comes along and fixes it. ;-)

I guess it's a developers tool more than a users tool. I'm not sure the majority of users will even use it but I'm sure the content developers would find it very helpful.

Wyrwulf: The import UV's are handy when you are setting things up after the figure is done and the UV's are layed out the way you want. Suppose you want to re-arrange the way a model is spliced. Or want to add or subtract a new body part. Just save the current UV's then go ahead and re-map the model so you can physically get at the vertices that were otherwise hidden, to re-name them. Then when you are done. Just import the saved UV's and the previous UV mapping will go back to the original and you won't have to re-do them. And you also now have new body part names to use to update the figure. That's a handy trick for getting at hidden UV's without losing the current UVmap set up. It takes a lot of work sometimes to get UV's laid out well. That trick helps you make changes without losing your previous work. That doesn't really help me with cutting up figures after I alter the geometry since the UV's vertice count won't match. Like I said. I just found it strange that I can create a figure and set up the bones by hand. But I can't simply re-use those "same" bones to re-split the "same" figure with some minor changes if I change the base model later on. Without it getting turned into a chopped up mess. With more pre-planning in my modeling program. I can get around most of the repeat set up and re-use the existing .cr2. It just would have been a nice option to have.

"why aren't you using the obj file that the CR2 is referencing, which is by its very nature already grouped?" Because the .obj files have far too many points to work on. And if the model is grouped and you want to alter it where two parts connect you'd drive yourself crazy dealing with all of those connecting vertices. "There isn't anything wrong with autogroup. It does exactly what it was designed to do. It gets you about 3/4 of the way there and then you need to fix the rest. It is there so we don't need to start from scratch in our grouping." If that was true Anthony then I wouldn't have a problem with it. But the AutoGroup mangels my figures so badly that it would take longer to fix it than to cut it up by hand. Connecting two body parts making chest parts where the hip is. It makes a real mess. I can make named selections of the body parts inside C4D and use the RipTide plug-in so that I can edit the nurbs model and make the conversion process to Poser a bit faster so Making changes isn't such a chore. I just had to change my workflow and do more work in C4D. But the more work I do In C4D. The more I wonder why I'm using Poser at all. I just wanted to make sure I wasn't doing something wrong. Thanks for the suggestions.

Thanks Doc. But that isn't really what I'm talking about. Pick any of the Poser figures and group all of the body parts into one group in UVmapper. Then run it through the Setup room using the same bones for the figure in the library. And you should see that the AutoGroup function fails to properly split up the mesh. In fact It makes a horrible mish mash out of it. Since realistic organic models are made in one piece without doll like ball joints. This makes the bones feature useless unless you split the model by hand exactly like it was before. Very time consuming. And not really worth the effort. The AutoGroup is a nice feature idea. It's just a shame that it doesn't actually work very well. Frankly, I don't see how it can work properly because there are no values inside the JP's that tell the program where to split the mesh. So it just takes a guess where to split it. And that guess is usually way off from where it needs to be. That's really a shame. It would be a handy tool to be able to split the model automatically and accurately after you've already taken the time to do it once.

That's what I was afraid of. I was just curious if I was doing something wrong.

Thanks MR.
I found what I was doing wrong. I cut the leg in half then relaxed it and stitched one side together.This is wrong.

The key thing I have to remember is the three piece rule:
The bodypart must be in three pieces before stitching together. As well as all of the seams facing eachother in the same direction before joining them.

I have a much clearer understanding of the theory now.
But DeepUV still has some things I can't figure out.
I can't spread large numbers of clustered points enough with the relax functions. I wish it would spread points instead of constricting them during the relax.

I'm also getting weird things like a seem that the program wants to join to more than one parent.
Example: The right thigh has been split from the hip. Yet when I relax the rThigh and get it flat to my liking. For some reason when I select it for joining. I get a few stray points that show up on the opposite side of the hip that the program says it wants to join to.
Hard to describe. But it's getting confused as to where the original seam was when I try to join them back together.

I can't relax an entire body part. If I do that. It collapses in on itself. So I have to select the inside of the part to relax it. Leaving a lot of close unrelaxed points around the circumference of the part.
If I could relax those points outwards, instead of inwards, I could get much better results around the edges of the parts.

Message edited on: 11/18/2005 17:29

It's not going to work. There's just too many problems trying to use DeepUV. I can't even lift apart the two halfs in DeepUV because one or two points stay connected. There's just too many odd glitches getting in my way. Although Cinema4D makes nice clean quad based meshes. I still create a fair amount of points and polygons because I don't use splines. I model by slicing up a cube using a small amount of hypernurbs for smoothing and create polygons that ultimately creates too many points to try to manipulate afterwards. If I want to be able to create a nice UV map without pulling my hair out. I guess I'm going to have to make the model using Nurbs edges. Trying to edit even a simple .obj polygon model is just way to complicated. I'm never going to pull it off this way.

I scaled the model up 10,000% and tried the DeepUV tutorial again.
This time the model didn't explode. But when I try to use the Relax function. The model gets twisted into a mess.
The picture shows the before and after relaxing operation.

Message edited on: 11/17/2005 14:24