Star fox (krystal model) Needs skinned and rigged and compiled
604 replies, posted
What you have just witnessed is called "Thread CPR". As you can see, the thread has fully recovered from its near death experience.
I was here when the thread was first concieved, I remember it well, it would be interesting to see this come to full expectations.
IF it comes to full expectations.
I'm running into that p4lib.dll error when trying to compile under the orange box engine. (Did Valve forget something?)
Edit: I think I see why this wasn't filled sooner.
Model compiled. but the phy file is missing.
The thread was never dead.
To quote the Cthulhu Mythos, "That is not dead which can eternal lie. And with strange aeons even death may die." These aeons have been pretty strange indeed, so the death of this thread is in fact dead.
I know, it's a conundrum. How can you kill death? The answer: You can't kill death, but you can kill Death. Maybe Old Castro, one of the cultists of the Cthulhu Cult, was referring to death with a big D, but messed up his grammar a little.
In other words, threads can't die unless locked, and even then, some have the power to unlock them.
If I can't get Krystal to compile useing Orange Box settings, I'll have to reduce the poly count.
Finished work on the stockings. Almost satisfied with how the materials turned out.
[img]http://img144.imageshack.us/img144/6492/previewkrystal91k.jpg[/img]
Unless someone can get Krystal's clothes to reduce in poly count under 25000 along with Krystal herself, The best you can expect would be her nude.
Edit: Well that and maybe her clothes as separate objects.
Are you using the regular 36470-poly mesh, or the 12132-poly LOD mesh?
The 12132 poly LOD mesh. I figured it would be the most workable within the compiler's limits. I did have to reduce the poly count on her eyes and hair though, but her eyes suffered the least. For that matter the texture still works just fine.
3dsmax has a tool in it called optimizer. basically it eliminated edge or faces that weren't needed. and by increasing the ratio, you could take off polys to a targeted amount. the problem with this tool was it almost randomly removed polygons anywhere, and it destroyed the UVs for the textures.
anyway I was talking to someone, telling them I found that you can retopo the UV on the optimized mesh using the render to texture projection in 3dsmax. and they mentioned that no, there's an easier way, there are tool on the market that do this way easier.
so I googled around a bit, and sure enough. there's a pretty handy tool for 3dsmax. I'm sure there maybe better tools out there, but this one I found right away. "PolyCruncher"
you set the ratio to meet your target, and the UVs stay intact. :)
also I was reading on the SDK site, it says the MAX limit of the source compiler is;
"[I]maximum vertices being roughly 10,000 x 3 (32767)[/I]"
this includes the useage of bones, and your collision model as well. so be sure to optimize your collision model prior to compiling. though in my own experiments I found that I could get closer to 36 - 38 thousand polygon models compiled fine.
anyway we can say that 20k is max for a single SMD, and that if you use a split SMD, you can get to a total of 32k safely.
by the way the compiler error for vertex limit is;
ERROR: short conversion out of range 34206 = too many polygons
34206 is the number of polygons I tried to compile, and the compiler failed. reading on the SDK site, seems that this limit is because of how the polygon strips are written in the file.
"[I]studiomdl.exe is trying to write out a triangle strip with more than 32767 vertices, which is more than Source supports[/I]"
what does this mean? it means that the actual limit to the amount of our models is controlled by the amount of poly strips written at one time. there may be a way to break apart these polystrips so that larger models can get in. because SDK says 32k is max, yet me and fakefactory have compiled models around 36-38k
enough babble, but we have the ablity to easily stick in retardily hipoly models.
O.o I only understood about half of what was said in the last few posts... I need to get back into modeling if I am this rusty...
Didn't FakeFactory use a hacked version of Studiomdl? (from what I heard anyways)
so what is the situation on the gmod port? is it coming along?
[QUOTE=LDragon;13988476]Finished work on the stockings. Almost satisfied with how the materials turned out.
[img]http://img144.imageshack.us/img144/6492/previewkrystal91k.jpg[/img][/QUOTE]
That mesh looks like it's built from smoothed out boxes if I'm brutally honest. Those edges are so hard and not natural.
[QUOTE=r0ck16;14044558]so what is the situation on the gmod port? is it coming along?[/QUOTE]
Latest Test compile is a success.
Textures not yet converted, had to throw in some from my personal collection.
Still needs rigging
Still need custom Collsion model
Final stats: not finished
When finished will be Gmod 10 and up only. (My first.)
I'm not sure but he might be working on porting a few things right now so it might be a while.
This model is one of those "few things". I just take on way too much.
Current Projects:
Krystal
PKT's model
Swatbot
Community Chest 4 map (DOOM)
Sprite comic Dimension Sonic
Caramelldansen part
School work.
I have a habit of doing that too. It the reason I'm now trying not to give out details on my projects so I don't get rushed.
[QUOTE=PLing;14044965]That mesh looks like it's built from smoothed out boxes if I'm brutally honest.[/quote]
My first attempts at a stocking used the box modeling approach. This one, however, was built from a tube.
[quote]Those edges are so hard and not natural.[/QUOTE]
Which edges?
Latest status.
College paper out of the way!
Textures still not yet converted.
Rigging has begun
Still need custom Collsion model
Final stats: not finished
If successful I may look into normal maps for future projects
So what's the size of the model, about Alyx's size?
yes. Although when I imported her into max, she was the size of a Micro Machine.
Poser has used this ridiculously-small scale for over a decade. Causes no end of problems with floating-point precision and the like.
Still, it's nothing you guys can't sort out. After all, you've got good skills.
[QUOTE=LDragon;14051645]My first attempts at a stocking used the box modeling approach. This one, however, was built from a tube.
Which edges?[/QUOTE]
The character mesh, not the clothes, I meant.
Well, that one's beyond my control, but I'll pass your critique along to CharleyFox, if you wish.
I think it'd help him so you could do that, yeah. There's nothing wrong with box modeling but those legs and feet just look too bulky and square in my opinion.
Believe it or not, I'm doing my Swatbot that way.
The thing about box modeling is if you don't know much about the curvation of organic things or at least keep them in mind they can turn out like that. I had that problem lots when I was working on Sally. I have a rounded cube style model which I'm now using as my base for organic models. Since Vanilla anyhow[saying it here since someone already guessed it was her] I do think box modeling helps you form quick and easy models with some practice. When I was making Mina I used a combo of box modeling starting from a flat base combined with a mirror modifier and a bunch of extrusions. With Sally I subsurfed a fair bit to smooth things out. Now I subdivide smooth along edge rings for that effect and then adjust vertices when I can. One of the biggest problems I had with Mina was I tried to model out too many parts which I should have UV'd The hair and shorts show this problem the most.
Sorry, you need to Log In to post a reply to this thread.