Hi,
In this topic there is a small showcase of a jME user displaying his nice air scenery, followed by Azatoth's Rover game. Azatoth mentions that he used a high poly count model but reduced it's amount of poly's. Why would someone do that? I can understand that the high amount of poly's eats alot of resources but 1 very detailed modell shouldn't matter, should it? So I was wondering if it perhaps was a limitation.
To make a very unfair comparison,
the jeep on this picture is much more detailed.
http://www.freakygaming.com/gallery/action_games/battlefield_2/jumping_jeep.jpg
Now I was wondering if it was a matter of using the right textures or if its a limitation in the hardware or perhaps software?
In other words, if I would use the exact seem modell from the picture, would the game be playable?
Might be a silly question but im trying to figure out how others worked.
Thanks!
Sure you could replace the model, and it would still be playable just the same. It's mostly good textures (though not that good). The impressive thing about BF2 is you can have a dozen of those jeep, a bunch of explosions and half the map, all in the same view.
That said, I think the jeep of Rover run actually looks pretty good compared to this one. Matter of taste I guess 
llama said:
Sure you could replace the model, and it would still be playable just the same. It's mostly good textures (though not *that* good). The impressive thing about BF2 is you can have a dozen of those jeep, a bunch of explosions and half the map, all in the same view.
That said, I think the jeep of Rover run actually looks pretty good compared to this one. Matter of taste I guess :)
Oh yea! I didn't want to criticize anyone on his work but I was just wondering how the makers of BF2 did that.
So you're saying the reason why it looks so nice in BF2 and still runs with a good frame rate is all thanks to the textures?
Actually, I wondered the same thing about BF2, the explosions and action and still the decent frame rate. What do you think is behind that and is it possible with in jME as well?
Not that I want to make a BF2 clone or anything, but im trying to determen if they have very special high-tech techniques or if its just "the average engine with alot of good modellers" behind it. :)
I'd love to make a BF2 clone. :-o
darkfrog
Well yes, I think it's mostly the textures (maybe with some bump mapping, but that's basically texturing too). It doesn't look very high polygon to me.
What's behind the high framerates is not one single technique, but several… (and I don't pretend to know them all). Using and reusing (to save memory) good looking textures is one, compared to other games BF2 uses much less detailed models.
It however probably involves a much more fine grained resource managment than what the current structure of jME lends itself for, but that's simply because noone is making a game like that. So to answer your question: yes they use very special high-tech techniques. That doesn't mean you can't implement most of them in jME.
llama said:
That doesn't mean you can't implement most of them in jME.
But doing so will require weeks of research, coming up with a likewise technique and then trying to implement it in jME?
So it comes down to "It's possible.. but it'll require alot of ingenuity and dedication" If i hear you right :)
But the fact that its possible is nice to know nontheless.
And the alternative is to create a mod inside Battlefield 2? :-p
Really your only valid alternative is to likely spend even more time and/or money in order to develop it in C/C++.
darkfrog
Yes, and time. More time than you as single person has. The people behind BF2 have been working on this type of engine since back when most of us were still playing indoor/BSP style games like Quake2. I'd say, for this type of gameplay, the BF2 engine is still one generation ahead of anything else I've seen that you can actually buy in the store today.
darkfrog said:
And the alternative is to create a mod inside Battlefield 2? :-p
Really your only valid alternative is to likely spend even more time and/or money in order to develop it in C/C++.
darkfrog
Well its not really BF2 im interested in but when I look at the game I see a rich environment, nice vehicles, nice models, nice effects and a fast framerate. Im evaluating how this kind of "stuff" would be translated in jME, which I find a very interesting discussion. Basically I get the idea everything is possible if you implement it, now im beginning to get curious what already has been implemented but that would require myself to delve in deeply into jME I guess.
When I look arround I see the cloth effect, particle effects and bloom effect I get excited because those are, in my eyes, "high-tech" techniques to make the world more realistic.
For the rest I can only agree with llama. The reason why I'm so interested in these techniques about jME is because I want to make a showcase demo about 3d gaming techniques. jME already has a couple (is there perhaps some documentation which ones jME has? I can search for their definitions on the internet) and perhaps I can help extend jME with some. Eventually I hope to have a nice demo of the techniques and pour those in a game for my college. If anyone has any tips where the best place is to start, i'd be thankfull! :)
I could write a book of tips…That is, only in the graphics making side fo the things.
I work hard in a company now were we don't have even accelerated 3d of any kind, all done in cpu, something like a pentium 120 or the likes, no ram (think of 4 megs for EVERYTHING) , no opengl, no lighting , no floats (there's way too many cpu features lacking, is not then really comparable to a pentium 120 or the like) , no texture correction, no mipmaps, no more than like 1500 tris in total of scenery, and characters…Downloading crazily textures res, and textures memory…Think of making a 500 tris model and forcing it look real (at a meidum/near distance, but even so…)
Well, even so, you would just not believe all what can fake graphics when using brain and tricks in the graphics making part. You get used to that, and even take those habits as a vice…
That's a mobile 3d engine.
But before I was at a PC game company, AAA aiming (in graphic quality)…
And generally, there are lots of things you discover… Like…LOD is crucial, a good tetxure can fake pretty everything, normal maps are good but imho, the human eyes get fooled easilyin many other ways…
The AI, FX, shadows can kill more permformance than any polygon count…depending on the technique and way they are put it…
Crazy stuff like making way too much meshes is a bottle neck, can be a bottle neck simply faces put together if it causes some lock to the engine…When doing a level, there are Max plugins that allow automatically studying what a camera path would never show, and so select or directly delete the never seen polygons…usually silly stuff as never ever model the "floor" of a barrel or structure unless it have physics applied…etc, etc, etc. Books and books can be written. And no, I never ever trust anymore on the "great engine" need to make a great game. You need a great team. Or a great aim and knowledge if you are alone, and know you can only then do an small, but very clever and cool game. And to me, it can be of the same valu of an AAA. As at the end what is important is the game, how fun it is, how carefully it is done.
Normal maps badly applied, for instance, have killed way too many games. And also, games withno normal maps at all, have become great hits even today.To mention that normal maps, well done, mean a huge team.Thing of making a hollywood movie and you get the idea. Is work aaaaall in hi res meshes. Like cinema. Think of zbrush for every mesh. crazy. Medium and small sized companies had to close for reasons like this…Yet though, usuall some medium companies do what they only can: dirty normal maps, often just using so quick photoshop bump making technique for walls and objects, and u find many enemies not well normal mapped, only mainplayer. The ps2 can throw like 60k tris easily, so it often just trust on that…andhave usually ridiculously low mem for textures , so they get to be as small as 128x128 or 64x64, often. And yet ppl get amazed…as it's low res. TV is VERY low res. (mobile way less 
)
To me, the gfx part of jME doesn't worry me. You can make an amazing game without even any particles or fx at all. You can fake all themwith good graphic tricks. You mainly need alpha sprites, good shadows (yup, that's my only minimal worry in jME) , smoothing normals, some basic multitexturing(lightmap)…and…little more.
No doubt an incomplete list… but you could make a demo I'd think.
Basic OpenGL effects:
Environmental mapping
Fog
Render to texture (camera effect)
Bump mapping
jMe effects:
Physics
Particles
3D Sound
Basic shadows
Bloom
Sketch
Outline
Lens flare
Cloth
Fade
jME effects not in CVS but with source:
Water (MonkeyWorld)
Explosions (Furballz)
Video to Texture ( http://www.jmonkeyengine.com/jmeforum/index.php?topic=2908.0 )
@Snaga
Thanks for writing that informative post. It made me think more about the subject and you're right on the graphics vs gameplay aspect. Gameplay always wins over graphics, in the early days (Pac-Man, Space invaders) and even today (various online multiplayer games with horrible graphics but rocksolid and well-thought of gameplay). On the other hand when you see games released with alot of potential but it being worthless due to bugs, low fps on all machines(other words: bad code) its pains me in my heart. Note that I make a mix up between bad design and poor graphics. I think a game should be well thought out off and developed (even though thats impossible when you have a publisher who wants you to finish at specific deadlines) how did you manage at your former company?
Also you mentioned they aimed at AAA quality graphics. Dit they succeed and how did they initially started?
Im hoping to start something for myself in the near future as well, im a bit ambitious and know about 5 people that are willing to join in on the graphics and sound department so im curious about other succes story's. I have a good idea how to deal with everything but it never hurts to evaluate others that walked the path before you
@llama
That's a nice list I can start with. The bump mapping is the one im most interested in since it makes textures look very realistic. I wonder what suprise the other techniques will bring.
Everyone thanks alot, this really is a great helpfull (and funny -> Darkfrog+llama+Mr.Coder dialogue's) forum
Darklord said:
@Snaga
Thanks for writing that informative post. It made me think more about the subject and you're right on the graphics vs gameplay aspect.
Actually...my point is more ...."you can do great graphics, without eating all cpu and resources, and you can fool yet a lot the user. What is more, often bad concept or lack of base knowledge of making graphics makes errors way more noticeable for that reason that for technology lack."
An example. Metal slug. great gfx. With very little resources. Those kind of games were all about 8 bit palette graphics as much...
I am wasted to see lots of renders of humans with big anatomy errors, or non expresive faces. that will kill the game no matter how much graphic technology is in it.
besides, I consider optimization itself as an art.
And is not a matter of some crazy type of graphic geek, fan of the old school...you allways need to optimize. If you work for rendering movies for a TV serie, you need to work in the subdivided cages often, and in many cases, with lights precalculated (yup, sort of lightmap, but different) Why? Time. A render throwing polies and fx without worries, or antialiasing at maximum quality no matter if is just for fast tests, will make sink your company, as you will wast tons of rendering hours for a frame, or, working with subdivided meshes, will make work less accurate and slower.
usually it provides later on better quality, ironically.
i do believe bad graphics do kill totally a game. true that a tron or arkanoid are fun by thenselves, but generally, you need graphics to make their function but...no more! You don't need to demonstrate how powerful your 3d card is unless it really makes the user get more into the game, and not out of it.
Gameplay always wins over graphics, in the early days (Pac-Man, Space invaders)
With a difference. By then, nooone had seen battlefield or doom3 graphics ;)
People now expect and demand a lot. My point is: The clever position is not over-use technology, and don't feel forced to, if you can achieve the needed effect on the player with other ways. The tech often makes people forget the real aim.In that I'd agree.
An example, Unreal Tournament4 has triggered huge lots of fans all over the worl, as well as Worl of Warcraft, and imho, to this moment, you can consider those outdated. yet though, the graphics are good, to me, are even better than some of those eating the whole of the 3d card to paint a logo ;)
and even today (various online multiplayer games with horrible graphics but rocksolid and well-thought of gameplay). On the other hand when you see games released with alot of potential but it being worthless due to bugs, low fps on all machines(other words: bad code) its pains me in my heart. Note that I make a mix up between bad design and poor graphics. I think a game should be well thought out off and developed (even though thats impossible when you have a publisher who wants you to finish at specific deadlines) how did you manage at your former company?
My everyday tends to be constant changes of route by bosses, and adapting myself a lot. All time is new harder tasks. Is also a very risky and unstable job.
The workarounds in game art...many. Mostly, being practical above all things. And remembering passion about games is totally secondary there. One learns which are the likes of the art director, wether wrong or not, and tries to find a intermediate way between their confused concepts, and what it should be there to not loose quality, or not enough as to be strongly noticeable.
i learnt in first game company I was that was all about not anymore do GREAT, unsurpassable graphics...but the best you could, in thegiven time, with the given resources (software, refs or time to dig for them) , and mostly, the fighting bosses's ego around the place. (the terrible combination is where there ares everal all with same hierarchy and fight between em...about YOUR graphics... )
So, I do my job. That's all. they allways end up happy, just not me. ;)
Even so, the games are good, we work like beasts, so how could it be other way... ;)
The coders I usually work with are godlike level, luckily.
I am pretty satisfied, even so, with what I achieve with all the obstacles, in my graphics. but I tell u: is not the career of my life. Have worked at...3 or 4 of games, or more or less, and some other of web design, pc maintenance...and othe rkind of job, very differnt of all that. At the end is all much the same. Boss/empoloyee. better do what you like in free time. I prefer to do an uglier job badly, than my favourite task badly...
Ehmm...not to sound depresive or something...
Also you mentioned they aimed at AAA quality graphics. Dit they succeed and how did they initially started?
I choosed carefully the verb. No they didnt. Company sunk, folded. Too small , but too ambitious people without feet touching floor...
But was fun. As raw workers, non hierarchy ones: coders, and another gfx man, were really tallented. Exceptional pros I will never work with again, I am afraid...each took his route.
Im hoping to start something for myself in the near future as well, im a bit ambitious and know about 5 people that are willing to join in on the graphics and sound department so im curious about other succes story's.
Oh, well, a group of friends is totally different. It is possible, but very hard , you all must maintain with no money for long time...
I have a good idea how to deal with everything but it never hurts to evaluate others that walked the path before you :)
my path is of one worker inside companies not vet well carried, and too small and unpractical. Not very useful.
Try to be practical and do good research before on how and what can you sell. There are possibilities, but they must be studied quite a lot before jumping into stuff.
Though sure you know of busines much more than me :)
Each situation is a different world though. i sound negative for the late years course of things. You will draw a different story.
Also, not all places/people/money are the same.