Ifurita, you make a good point. Still, I’m dedicated to replying until RivrStyx admits that I actually know - to a degree - what I’m talking about, and that my attempts to help actually occurred.
My convert map Den of Lions: Lair to ET
Thought I’d jump in for the hell of it. From everything you’ve said, kevinski, you’ve demonstrated absolutely no knowlege whatsoever of how the vis process works and how the Q3 engine (which is what ET uses, albeit a modified version) handles portals and clusters. Furthermore, you have demonstrated absolutely no knowlege of how the compiler disects the maps into clusters to calculate the visibility. Not to mention that despite all your comments about optimizing, you haven’t provided one single hint that you know what goes into FPS and cluster vis optimization via Radiant.
First, the game engine has to render stuff that we can’t see on the theory that we will see it shortly so as not to produce the HOM effect. When optimizing a map, you place hint brushes and possibly areaportals (if you have working doors). Areaportals and hint brushes tell the Vis process where to create splits in the BSP so as to have the game engine draw as little as neccessary at any given time. Unfortunately, even if you have hints and areaportals set up perfectly (which almost never happens anyway) the Vis process may just flat ignore them because as much as you’d like a split at location X, the compiler knows that it won’t work. Finally, the Q3 engine has never been good at handling wide open space (because the player can ‘see’ all the way across it so the compiler doesn’t put portals in), so if Res has wide open spaces (I’ve never played it) then the engine is drawing a ton of information and your card is choking (read on). RivrStyx commented about tree models and those can hurt too.
Moving on to your FPS problem in Res, I have two words for you: FILL RATE. You could have a gig of texture memory and that still wouldn’t change the fact that a GF2 MX simply doesn’t have a very high fill rate or a very high memory bandwidth. Lightning and rain and all the other shaders on top of each other only make it worse because each stage in each shader means an additional pass for the video card (assuming non-multitexture. I’m not sure how many stages a GF2 MX can handle in a single pass). A GF2 MX is simply not fast enough to handle that much information at high resolutions. Again, your problem is that area where several models are being drawn along with several megatexels of texture information that’s bein processed 8 or 9 times over by your card.
And if your wondering why I jumped in, it’s because I hate seeing maps and mappers get bashed on performance by someone whose hardware is 5 or 6 generations old now and is technologically and physically incapable of handling that kind of data rate. If you’re going to keep that GF2 MX longer and expect to play the newest stuff, then you need to learn to optimize your system for it.
Footnote, I’ve never played any of RivrStyx’s maps, nor had I heard of him until now, so you can’t accuse me of being one of his ‘fanboys’.
LodeRunner, I’ve stressed the fact that I’m NEW to mapping. Of course I don’t understand all of the details. If you think I haven’t read up on the matters, however, you’re completely wrong. I’ve read about how the compiler breaks the map into smaller clusters, but it was noted that the information was more theoretical, and that it wasn’t necessarily completely accurate as to how the splits would occur. Yes, I know that the compiler calculates what you can and can’t see from a certain cluster. Of course, I don’t know the technicalities, but - then again - I never claim to.
Now, you say that you jumped in because you hate mappers being bashed. Strange thing is, it’s apparent that nobody can see the high number of comments I’ve posted that don’t insult RivrStyx or his maps. I don’t have time for a more lengthy reply at the moment, so you’ll hafta wait until later tonight. Remember, before you spout anymore I-know-this-and-you-don’t babble, take into consideration the fact that I’ve admitted time and time again that I’m NEW to mapping.
I think Loderunners comments on your videocard are worth noting though too kevinski. The general level of graphics in games is always evolving, and now so many users have the hardware to cope with very high poly counts that it opens new doors to mappers. The only problem is having to also cater for those who have no desire to leave the old boxy worlds that run fine on their ancient videocards.
I recently upgraded to a Radeon 9000 Pro 128MB, it’s pretty much a bottom-of-the-line budget card but it can run everything I throw at it with ease. Maybe Doom3 will be a test but until then there’s no games around to really push it’s limits. If I had paid £300 for a top card like the 9800 or something, I’d be a little annoyed that the game developers weren’t really breaking new ground and were still having to make levels that run fine on a Geforce2.
At some point in the future, you’re going to be forced to stop critisizing maps that don’t run well on your GF2. Obviously that time will come, you’ll have to cross that line. That time is pretty much now IMO. Surely you don’t want to be the very last person still hanging onto the old technology and blaming the incredible new games with millions of polys and stupidly high res textures, for them not being optimized. Also I think if you shopped around you’d be very suprised just how inexpensive some very powerful videocards are.
Not actually true - if you ask for a split, a split you will get. Area portals are a bit different, and as I discovered in church, don’t really give the benefits you’d think, and I recall a discussion culminating in the fact that really, area portals aren’t worth it for multiplayer maps. For SP, they work great however.
I think the problem with hint brushes is once again people think they know how they work, but really they don’t. I’d say to start with steer clear of hints - beginners should simply try to get their map running at a decent framerate without them - leave hints (which if improperly placed simply do nothing than increase compile times) alone.
Finally, the Q3 engine has never been good at handling wide open space (because the player can ‘see’ all the way across it so the compiler doesn’t put portals in),
Again not true - try loading up the portal data in radiant. Lots of portals! Let’s not forget over and above any BSP splits placed by you as hints, or created by structural geometry, q3map / q3map2 automatically splits the BSP up along the major axes (you can fiddle with this with the _blocksize key in worldspawn, although not recommended if you don’t know how it works). The thing is, you can split your big area up into as many portals as you want - the final vis data is determined by what portals can see what other ones. Twice as many portals in a wide open space doesn’t help cull anything - THAT’s where the problem is - and upping the portals with hints again (in this instance) simply does nothing but increase compile time.
The Q3 engine CAN handle open terrain - mostly it doesn’t care what a polygon is, be it interior detail or part of a landscape (OK there is some hit with lightmapped terrain cos it’s usually another texture pass than most normal shaders, but that’s only a small point here). What goes wrong with terrains / open spaces is people make 'em, then try to fill them with as much detail as if they were indoors - this is where it all starts to go wrong - particularly lots of free standing structures - because the VIS stage tends to love tunnels for it’s efficiency because of it’s portal-based visibility tests.
There are always ways to optimise that people don’t implement, but usually it shouldn’t be necessary - it’s only really anal people like me who go for that extra 20% performance. Of course, my maps run better as a result, but the number of people who are likely to complain about my maps running speed are fairly few because I always design my maps with slow cards in mind. Others don’t - if they don’t, then that’s their choice.
RivrStyx has stated his minimum specs for his map, and if yours doesn’t match up, then there’s not really much you can say about that, other than ask him to make it work better in future. Given his maps are being upgraded from RTCW to ET, somehow I doubt it, cos the official maps in ET are slower than the official ones from RTCW.
That many custom mappers appear to be using this as an excuse to be lazy in their mapping is not something I’m very impressed with, but other than pointing out where they go wrong, there’s not a lot I can do about it. I wouldn’t be so bold as to assume RivrSytx will be lazy though - I prefer to comment on results once they are out, not hypothesize about how he may or may not do it. But bear in mind that ET itself isn’t inherently any slower or faster than RTCW, so a conversion from RTCW is likely to run about the same speed on ET. If, once again, your card doesn’t meet his minimum specs, then you’ll just have to accept that 
Finally, have you TRIED running these maps with everything turned down to minimum? You said turning it down to minimum doesn’t seem to affect much, but then you state you did turn everything down to minimum for Market garden and it helped - I had to accept long ago with my crap card that I was going to be running things at minimum just to get a barely reasonable refresh-rate. Minimum res, minimum texture sizes, low geometry, 16 bit all the way. It helps, and if it’s the difference between eye-candy and smooth (-er) framerates, then that’s a choice you’ll have to make and accept the consequences.
Just TRY whacking everything to minimum and see if it makes a difference
For example, it looks to me from your screenshots that you have texture detail at high, with lightmapping enabled. Try turning lightmapping off and running with lowest detail textures. It will most likely help - maybe not get you a constant 200 fps, but better than what you have. Gotta be worth a try?
G0-Gerbil, well put. I map for Q3, and my maps have been small enough to get along without hacing to hint them because of how I use hall ways. As for wide open spaces, that’s what I meant: people make them then fill them with a rediculous amount of detail and don’t set up any vis blocking by using blocks of caulk under hills. As far as vis, yes I was wrong in what I said, hints will get you splits, whether or not they do anything other than up compile time and bloat the bsp is another matter. Space maps in Q3 are the best example that I am familiar with.
As for your comments ziege, do you have any concept of how much faster a Radeon 9000, even budget one, is than a GF2 MX? I would care to point out to you that a GF2 MX is a severly stripped wodn version of the GF2 standard, with limitations programmed into its core. Your Radeon 9000 is most likely running on 128MB of DDR at AGP 4x )even at 2x it’ll still be able to handle that). The Radeon 9000 has a much higher fill rate than the GF2 will ever have.
Kevinski, perhaps I came off too strongly. I was not here to do an “I know this and you don’t” speel, I could have phrased most of it better (was a little late at night for me, but that should not be and is not an excuse).
You made no attempt to be nice about it, you offered no suggestions as to what to do, and you acted as though you knew mapping inside and out. Perhaps a better way would have been to comment, “RivrStyx, is there a way that you could optimize the map more for lower end cards?” And yes, I hsould practice what I preach about trying a more polite way to say things. And quite frankly (I’m just being blunt, not sarcastic, and trying not to be rude, but there’s no other way), you can’t use the “I’m new to mapping excuse” if you’re going to make statements that make you look like a know it all ass (I know it doesn’t work 'cuz I did it when I tried my hand at Q2 mapping years ago. I’m glad the forum I was at no longer exists.)
Yes, that was my whole point. I bought a cheap budget card, and it still heavily outperforms his one. He should upgrade.
Wonderful.
Hey let’s not dwell too much on the ‘upgrades are cheap’ path - he may not have the money, he may not want to upgrade just for a couple of maps, it may not be his computer.
If it’s that cheap - why don’t you guys chip in to get him a new graphics card? 
Anyway point is, it’s his choice for whatever reason to not upgrade, but by the same token it’s mappers’ choices how they do the map.
Personally, I must admit, I’d say his card is EXTREME bottom end for ET (I should know, I had one myself until a few weeks ago), so I’d be suprised if many custom maps will play OK on it. For RTCW, well… I dunno - maps should really run at a reasonable framerate on it - certainly if it’s just you cruising around on your own. But that’s my opinion, and obviously other people think otherwise.
I’d only really bitch if it was a level I was desperate to play and it ran crapply - if, according to the impression I get here, the map runs badly but he doesn’t want to play it, then well, live and let live 
True, he might not have the money to upgrade. I certainly put it off for enough years. But what I’m saying is that at some point he will have to accept that graphics have advanced further than what his card can handle. I don’t believe somebody should be complaining about FPS if they’re willing to continue using outdated technology. If all maps are created for a GF2 MX, forever, then games will never evolve. At some point he will need to either upgrade or stop complaining about FPS. I think we’re at that point already. Before I upgraded, I accepted anything new would run poor on my card.
If someone’s happy with a GF2, and wants to stick with it, great. Clearly when they start complaining about low FPS, they’re not happy with it, but instead of pointing the blame at the mapper every time they should be looking at thier own system.
Depends entirely on the game.
Official maps on RTCW worked fine (not great, but fine) on a gf2 mx.
ET ones don’t. So granted, if he bitches about ET maps, I’d have little sympathy (although I will be quick to point out my own large maps coming soon SHOULD run plenty fast enough on his cards, and they are bigger than any current custom offering).
But we are covering an RTCW map here that unlike many maps, doesn’t run fine on his computer. He’s not ok with it - the map-maker is - hence a stalemate really - especially since he’s hoping the ET version will run better. I’d argue that’s highly unlikely really, and although it may be a ‘symptom’ of the original map, for ET it’s a non-issue - ET chugs on slow cards, statement of fact.
Also, if you think RivrStyxs doesn’t optimise his maps (which I’m sure he does try to do), then the chances of a (presumably) n00b mapper copying his map and making it faster are slim indeed!
I dunno just about Kevinski, but RivrStyx should be worried about this conversion too - I know I am! I’d be highly suspicious that anyone could do a much better job than the original mapper, and the only reason I’m not bothered about my footy map is cos it was so simple to start with it couldn’t really be done ‘badly’ (he said, bearing in mind with the mirror in fidel’s map I was getting <20 fps :/, but I’m not particularly protective of the map, and if the worst comes to the worst I’ll help him sort it out myself).
Of course there are MANY maps that someone could do a better job than the original mapper, but what I mean is that anyone wanting to convert someone else’s map obviously doesn’t have the ability to do their own, completely different map, and therefore is highly unlikely to be as skilled as the original mapper.
So fear this conversion Styx - you may need a gf4 to play it at low res 
The map that started this thread 50 posts ago was not mp_resurrection, it was rivrstyx’s den of lions: lair map, which evidently kevinski doesn’t have problems with.
I for one am still concerned why this question remains unanswered!
Indeed, whatever you think of styx maps, it is rude at the very least to convert them without permission. Even more so if he didn’t release a source map for it.
Come to think of it, hijacking a thread to resurrect (pun most certainly intended) a flameware from another forum is not exactly the model of tact either…
Well sin… you going to respond or hide. Not a good thing to do since i’ve seen you promoting your website here and probably elsewhere. Not good for business really. Aren’t you the guy with the map website?
Just looked and don’t see it there… so what are you intending to do with it?
all i know is RIVRSTYX IS DA MAN! REMEMBER ME?!
rivr is god guys, bow down to him, do what he sais and dont talk back
lol just messin kinda, but yea i dont see why ppl hijack others maps, just lame
As noted elsewhere he’s also the one putting up private releases of maps publically without permission. Didn’t get a (coherent) responce there either :banghead:
Objectives for axis & allies:
There will be purpose both (both of; both; both of) team by 3 flags 3 minute each agog < take-over >. This backup of mode is with (from) UT, Domination.
Portion soon screens!
from www.poltran.com
Hey Diaper Eater…i remember you…:). Whad up
Sin…lets not do that…make your own map ok? You don’t have my permission. Whats with you dude?
I really have no idea what you said… but whatever it is… no.
It seems like he could care less about modifying someone else’s work without permission. Bailiff, whack his pee pee :twak:
OMFG! He’s using an online translator to communicate with us. There is no hope!