[Ace]

Bob, what mission is that from? (PM me with the reply when you have the time) Because I've love to see shots from either E1M19 or E2M4.

The high poly Golgotha looks great in those shots, so does the Vanir.

[Bobboau]

um... I think it was... blindsided... yes, what ever mission number that was.

eh, the Vanir could look better, it's all scratched up.
and I have a version of it with it's normals done properly (auto-faceted) so it should be lit corectly as well.

if I have time I'll try to get shots from them, might have to post them internaly though (as there may be confedential materials involved)

[Carl]

me thinks the scorch marks have way too much red in them. limit the red to the very center on the marks.

[Bobboau]

yeah, that's the art issue I was talking about.

and I found something that's probly going to cause that build to crash, so, expect it

[Lightspeed]

Try the ones I posted some days ago, Carl.

[JarC]

Originally posted by Bobboau
oh, and I supose I should mention that index buffers impose a poly limit of 21,845 triangles, yes, I know, it's quite restrictive, but we'll just have to get used to it I guess

that is per model right? I mean, throw 2 HTL-Fenris's and 2 hi-poly Anuket's in a mission and you got 20K+ poly's right there!

[Psychonaut]

Hi Bob,

what campaign or mod are the screenshots from. I don´t think i have seen this ships before:eek:

[karajorma]

BWO most likely.

[Flipside]

Blah blah exits blah blah flamethrowers etc

And it looks like BWO to me, though I'm intruiged by the 'Censored'

[Bobboau]

the limit I mentioned is per texture per submodel,

the screens shots are from BWO, mission called blindsided in wich the Golgotha blows some **** up.

[Lightspeed]

Can you fix the last build you posted so we dont have to expect it to crash?

[Bobboau]

you should probly expect crashes for a while, it has an incredably complex memory management system as a central part. thing is most of the problems seem to go away when I try to debug, so it's makeing it very hard to fix.

[SA22C]

Well I gave this build a whirl.  It seems that the particles have gone all funky.



What's causing this?

Also aside from the ruined particles, my FPS is horrible with this build, going down to the low teens when close to a Capship like the Fenris in the screenshot.

[Raptor]

Originally posted by SA22C
Also aside from the ruined particles, my FPS is horrible with this build, going down to the low teens when close to a Capship like the Fenris in the screenshot.

And thats not even the HTL Fenris...

And Low-teens FPS are not THAT bad.  Mine hovers around 12.5 FPS when I'm chasing a craft with a particle behind it (using IPAndrews released BSG Stuff, which is not really fair on the HTL engine).

Good FPS for me is ~37.5.  Then again I"m only runing 128mb of RAM. Hopefully I'll get some more soon...

[Starfighter]

The prob with the slow downs is easy to fix, just make them go away after say 15 seconds the FS engene can't handel them being there all the time its not built to. unless you make it so only missiles leave marks, or somthing that is not shot 24/7 this slow down prob will always be there.

[Flaser]

They already go away after a certain ammount was fired from the same stuff.

[Taristin]

The messed up particles is the -batch dynamic data flag in the launcher. Turn it off.

[Bobboau]

and that might also be what's causeing your slow down,
the decals are off in there own world, it's completly seprate from the rest of the FS code, and it is built from the ground up to be able to render absurd numbers of decals, the only thing that realy isn't working as best as it theoreticaly could is the burn pass.

[Bobboau]

and I just figured out what was causeing that crash I was mentioning, if a decal should ever fail to create (and it happens due to the culling mechanism being too agressive (need to work on that)) the bit that gets a new decal redy to be made from a model forgets to mantion to the rendering code that it's been changed, and thus it's vertex buffer never gets updated with the new geometry, and when the burn pass trys to access the verts in the vertex buffer for the last decal it's over shooting the vertex buffer, and this is very very _very_ bad.
I also reorganised the whole system (again) it's a lot more modular now, wich enabled me to get decals to work on submodels. the only thing is they won't work on turret barrels, but I might have a way to fix that, but it's extreemly low on the priorities.

nex build I post (or posably the last one I posted) also has an implementation of point sprites, RT was working on this, but the particle system is so unbeleaveably slow I had to do something (don't mean to step on toes RT) I made a thing that will render out a bunch of pointsprites quite speedaly and integrated it into the particle system, also integrated it into the bitmap drawing function so that is why thrusters and stuff are not going to look as good, just bare with it it'll get better, I'm going to try and get a build posted before I go to sleep, but if I can't get it decently stable before I pass out, it won't happen.

[Bobboau]

ok, I turned point sprites off for normal bitmap drawing becase suns were refuseing to draw and I don't have the energy to fight anymore, but particles are still useing my new batched pintsprite thingy, so if you see anything bad with particles this is why (this has nothing to do with the problem above), anyway that's just a bit of fore warning, the big thing is the decals, I played through a mission of BWO a rather intense mission that I died in, hear that I DIED FOR THESE DECALS DAMIT, eh I'm tired, what was I talking about again,... em, play with the decals and post a lot so that when I wake up I can see all the happy people with burning scorch marks on them

http://freespace.volitionwatch.com/blackwater/fs2_open_Bobboau_5-28-04.zip