Hard Light Productions Forums
Modding, Mission Design, and Coding => The Modding Workshop => Topic started by: --Steve-O-- on December 03, 2007, 11:50:17 am
-
ok i'm trying to get that mig model in game to take it for a test drive, every thing is fine all the way through to the load out screen, but the moment i hit commit to start the mission....crash. also when i try to plop this thing down in fred, fred crashes (side note, i used an old mission and just added the mig to the team load out).
both PCS 1 and PCS 2 had no problems with the model itself.
i've gone through the hirearchy time and again, fixed a few errors, reconverted crash.
i've gone through the table and fixed a few errors...still nothing
i've tried scalling the 2048 map and glow map down to 1024, still crashes.
i've even tryed pulling all the TBMs of other models out leaving just this thing...CTD.
its sitting at 18668 polies for the main hull 34198 polies total including polies grouped to detail boxes
i'm running out of ideas. :mad:
-
Could be mesh errors. Earlier versions of PCS2 would save pofs that crashed FSO.
Haven't checked the last two versions tho.
MeshLAb may help
http://meshlab.sourceforge.net/ (http://meshlab.sourceforge.net/)
Look at the first picture on this page.
http://www.blender3d.it/index.php?filename=lab/nonmanifold.html (http://www.blender3d.it/index.php?filename=lab/nonmanifold.html)
Basically the inside and outside get mixed up around these types of mesh joins.
Import model into mesh lab
Render>shaders>xray - and then
Render>colour>colour Non Manifold faces.
Any internal red patches may need attention.
-
will this program save all UV mapping already done? or is it going to toss the UV? ehh...either way, at least it will point the way to the problem. if all else fails i have the model in pre-union form with the UV mapping done. i took precautions on this one, i had a feeling...
-
will this program save all UV mapping already done? or is it going to toss the UV? ehh...either way, at least it will point the way to the problem. if all else fails i have the model in pre-union form with the UV mapping done. i took precautions on this one, i had a feeling...
Just use it to find the problem areas.
If you get stuck I'm happy to take a look at it.
-
do non manifold faces cause FS to crash in game? cuz i found some... the good news for me is that those pieces are easily removed and i have a source of non tainted pieces to use for repair. whats odd is the faces in questions are duplicates, some are fine, these went bad, and the are simple 4 face cubes.
i'm trying to think of anything i did differently on this model that may cause issues.
-
ah there is one thing i did.
i was trying to decide if FS would hate to deal with a bunch a little pieces in one archy group, or a bunch of little pieces unioned as one piece with a light in an archy group.
basically what i did is this. take 3 cubes put some space in between them all so they are not touching in any way, and then boolean union them together as one piece of geometry instead of 3. i was not sure this was sound but i figured it might be worth a try, push the FS rendering engine a little. you know, see what i can get away with. so non manifolds aside is this method a cause for problems?
-
ah there is one thing i did.
i was trying to decide if FS would hate to deal with a bunch a little pieces in one archy group, or a bunch of little pieces unioned as one piece with a light in an archy group.
That is what PCS does. If you have 10 pieces in one group, when it converts, you end up with only one object.
More subobjects in a model = more render passes = takes longerdo non manifold faces cause FS to crash in game? cuz i found some...
Maybe? Although PCS2 should delete(fix) weird geometry.
It would definately makes collision detection a mess. If it can't figure out where the inside and outside of the mesh are.
-
so then if PCS automatically groups several objects into one piece then really me doing before hand is superfluous, right? so then what I'm thinking is I'll just let PCS do it the way it likes. i'm going to fix the geometry errors today sometime and i'll give it another run through the game. lets keep our fingers crossed.
-
i corrected the geometry errors i could find. same problem still. it got through PCS just fine, you can veiw it in the tech room, it appears in the load out screen, but the moment i hit commit to fly the SOB it crashes with no explanation. the only other thing i can think of unless i messed up my archy somewhere and im just not seeing it, is i converted the model with PCS 1 and set up the POF info in PCS 2 originally, and for the fixed model i just imported all the POF data from the original PCS 2 model in PCS 1.
-
i corrected the geometry errors i could find. same problem still. it got through PCS just fine, you can veiw it in the tech room, it appears in the load out screen, but the moment i hit commit to fly the SOB it crashes with no explanation. the only other thing i can think of unless i messed up my archy somewhere and im just not seeing it, is i converted the model with PCS 1 and set up the POF info in PCS 2 originally, and for the fixed model i just imported all the POF data from the original PCS 2 model in PCS 1.
First - PCS1 can't handle more than 10-12k polys per subobject. If you go over (9k is safe) the limit then FSO will mostly crash.
I have found that using the import function in PCS2 can be a problem. If the original model has corrupted data, then importing it to an updated model messes it up as well. This caught me out a few times. And yes it can cause FSO to crash.
Do another version- Just as a model - no engines, guns etc
If you can get it in game - then the mesh is ok - and you need to set the pof data up again.
If it crashes FSO - the the mesh still has errors. - (Since you got it in the tech room the mesh should be ok.)
-
First - PCS1 can't handle more than 10-12k polys per subobject. If you go over (9k is safe) the limit then FSO will mostly crash.
I have found that using the import function in PCS2 can be a problem. If the original model has corrupted data, then importing it to an updated model messes it up as well. This caught me out a few times. And yes it can cause FSO to crash.
Do another version- Just as a model - no engines, guns etc
If you can get it in game - then the mesh is ok - and you need to set the pof data up again.
If it crashes FSO - the the mesh still has errors. - (Since you got it in the tech room the mesh should be ok.)
during one of my reworks i've cut the total polies down to 27545, and the largest poly subobject comes to under 7000, so i think thats rulled out. the mesh itself seems to be stable cuz like you said it fine up until you actually have to put it to work.
i was afraid id have to go and redo all the POF data, good thing i write all this Sh1# down. i'm really starting to think now after your last post thats where the problem has been all the time. even the original test POF i did made it as far as the load out screen. so logic says that the POF data really kicks in when its time to put the model to use in mission, and considering the game scrams when it comes time to implement that data, the well...its screwy. whatch fixing this is as simple as redoing the POF and every thing else i've tried was not needed. ehh...i'll try your suggestion tomorrow, when i wont feel like Mr. Mig just went poo all over my floor.
-
good call Water! it was bad POF data imported from PCS2. i just got the mig ingame and working, now im adding all the rest of the goodies via PCS1.
-
ok now get this... after redoing the whole POF in PCS1 i got it to work ingame just fine. then i added the thruster glows (12 of them) and then it starts deleting or rendering invisible chunks of the model. now bear in mind that i got it ingame once already and there were no mesh mishaps at all, and now its starting to do this.