project reality header
Go Back   Project Reality Forums > Developer Blogs > Dr Rank
21 Nov 2019, 00:00:00 (PRT)
Register Developer Blogs Project Reality Members List Search Today's Posts Mark Forums Read

Rate this Entry
Sangin - its getting there
Posted in: Uncategorized
Posted 2011-04-05 at 19:28 by [R-DEV]Dr Rank

Hey everyone, apologies for the lack of updates on Sangin. As ever I have been working hard to get this map out to you guys as soon as possible (work/life permitting). Thankfully I'm now getting close to the final stages, although I still have a lot of work to do on it before we hit release. A massive thanks has to be said to Rhino for his on going support for me on this project; for the last few months he has been working solely to produce statics/vegetation that I need to finish the map. Without him the map just would not be happening, so make sure you give him some lurv next time you see him about

Right, where am I now with it? Well, I finished roughing out the Green Zone at the beginning of January this year. Since then my main focus has been on optimising the map to ensure it performs as well as it possibly can for minimal to no cost to visuals. To that extent I have replaced, by hand, thousands of trees and hedgerows. Basically the way the bf2 engine works means that the more objects you have in the map the worse it will perform. Logical, yeah, BUT if you take say six trees, so six individual objects, weld them together and export them as one object and then put that one object in, despite having the same number of polys as six individual objects the engine will handle it a lot better. Now if I'd known this earlier in the development of the map, I probably wouldn't have put 2 hedgerows down for every strip of hedgerow to make it thicker but would have asked someone to make me a new one (which Rhino has now done). So in order to maximise performance I have had to, by hand, replace thousands of pieces of hedgerow

Anyway, I've now replaced most of the hedgerow I will hopefully have to replace and replaced all the shrublines bar one. I am currently in the process of tweaking all of the fields, after which I will go through the map and replace a load more trees I've used as bits of shrub to break up hedge lines etc. Once that's done I'll do Bastion, sort out the surrounding terrain, and tweak bits and pieces then do the compounds and detailing.

Here's the minimap taken at the beginning of the year before tweaking, some stuff have changed but most of it wouldn't be noticeable in the minimap anyway since the majority of the work has involved replacing objects (certainly a contender for the most boring job in the world XD):



Here are a few more in game shots:





Views: 20212 | Comments: 33

« Previous Entry  |  Main  |  Next Entry »

Comments
Total Comments: 33
  1. Old Comment
    charliegrs's Avatar
    If 128 players ever becomes official, then I think this map would be the perfect candidate for it.
    Posted 2011-04-06 at 01:35 by charliegrs charliegrs is offline
  2. Old Comment
    USMC scout sniper's Avatar
    Is that Afghanistan?
    Posted 2011-04-06 at 02:06 by USMC scout sniper USMC scout sniper is offline
  3. Old Comment
    Just_A_Thought's Avatar
    Quote:
    Basically the way the bf2 engine works means that the more objects you have in the map the worse it will perform. Logical, yeah, BUT if you take say six trees, so six individual objects, weld them together and export them as one object and then put that one object in, despite having the same number of polys as six individual objects the engine will handle it a lot better.
    Would there be an advantage to extreme welding, like welding all/most of the statics together just before final release--when you know they are exactly where you want them? (Is it possible or desirable?)

    Other than being the world's most boring, time-consuming job, what disadvantages would this have?
    (From these questions, you can guess how little I know about all this, but I am interested.)
    Posted 2011-04-06 at 06:52 by Just_A_Thought Just_A_Thought is offline
  4. Old Comment
    dtacs's Avatar
    Most realistic looking map to date, excellent work.

    NOW GET IT DONE!
    Posted 2011-04-06 at 09:18 by dtacs dtacs is offline
  5. Old Comment
    mockingbird0901's Avatar
    Truly epic work then
    Can't wait to have a go at it
    Posted 2011-04-06 at 10:26 by mockingbird0901 mockingbird0901 is offline
  6. Old Comment
    Rhino's Avatar
    Quote:
    Originally Posted by USMC scout sniper View Comment
    Is that Afghanistan?
    Nope, that's my back yard


    Quote:
    Originally Posted by Just_A_Thought View Comment
    Would there be an advantage to extreme welding, like welding all/most of the statics together just before final release--when you know they are exactly where you want them? (Is it possible or desirable?)

    Other than being the world's most boring, time-consuming job, what disadvantages would this have?
    (From these questions, you can guess how little I know about all this, but I am interested.)
    In the way your suggesting it, no. Would be too many unique objects and having very large poly heavy objects isn't good for LOD management, hence why the vBF2 Carrier and other large statics are split up into different sections, as if you had them all as one static, even thou your standing right on the front of the carrier, if it was all one object you would still be rendering the rear of the ship, and it's interior all in its highest detail, even thou your 500m away from it and can't even see the interior, let alone the fine detail on the other end of the ship where with it split into different sections, when you stand on the front of the ship, the front of the ship is rendering in its highest detail but the rear of the ship is rendering its a much lower detail, like lod1 or lod2 meaning your GPU doesn't need to process as many polys and your CPU only has to feed though a few extra draw calls

    There would also be a problem with too many unique statics with what your talking about, its easier for your CPU and GPU to process the same exact static over and over again, rather than lots of different objects so having the same tree be rendered over and over isn't too bad, until you get into extremes like what was happening on Sangin then the CPU gets clogged up with draw calls so merging 5 to 10 trees into one common object, means its much easier to process still and you still have a common object to be placed around the map which also has good LODs
    Posted 2011-04-06 at 10:35 by Rhino Rhino is offline
  7. Old Comment
    TheAssProtectors.Nr1's Avatar
    Simply amazing. Fantastic effort.
    Posted 2011-04-06 at 15:57 by TheAssProtectors.Nr1 TheAssProtectors.Nr1 is offline
  8. Old Comment
    TheOldBreed's Avatar
    Quote:
    Originally Posted by TheComedian View Comment
    Finally, a map no sniper will like.
    Only if you're a shit one.
    Posted 2011-04-06 at 19:53 by TheOldBreed TheOldBreed is offline
  9. Old Comment
    TheComedian's Avatar
    Quote:
    Originally Posted by TheOldBreed View Comment
    Only if you're a shit one.
    Well not exactly, all snipers are useless in flat terrain. Plenty of static cover and foliage to go around too.
    Posted 2011-04-06 at 21:41 by TheComedian TheComedian is offline
  10. Old Comment
    CareBear's Avatar
    Quote:
    Originally Posted by TheComedian View Comment
    Well not exactly, all snipers are useless in flat terrain. Plenty of static cover and foliage to go around too.
    actually its more of a snipers paradise if you know what your doing and not as flat as you think
    Posted 2011-04-09 at 22:47 by CareBear CareBear is offline
 

All times are GMT. The time now is 05:15.