Source vs UE3

[Big Fat Duck]

HOLY CRAP !!! Valve better update HL2 with a patch or something to include real time shadows or im going to shoot myself !!!

 

[The Terminator]

.......lol

 

[Impulse147]

i watched the UE3 demo movie, but i can't remeber them saying how powerful of a machine would have to be in order to run all that stuff( that is if game designers even choose to implement all of its features).

 

[Devilphish]

But the same can be said for Source

Don't be so defensive. I wasn't saying Source couldn't do neat mods too. I was just replying to the comment that Doom3 mods will be limited to Doom3 type games, which is untrue.

I'm not saying one engine is better then the other, i'm just saying that there is still enough reason to choose Source and make a game with it, even if it does come out in 2006. Look at Max Payne 2. Max Payne 2 has the exact same engine with a few updates to it, and it looked awesome when it came out, thanks to those updates. Same could be Source

Ofcourse Source can be updated to look better than it does now. No one here claimed anything to the contrary. But, I don't think you realise the difference between the time of the Max Payne engine and the time Source is hitting the scene. The technology changed very little between the release of Max Payne and it's sequel, while Source is coming out right at the beginning of a paradigm shift, and the beginning of the end of Source/Max Payne type rendering technology. Because of the paradigm shift in the way games are being rendered, Source will have to be changed in a major way to compete in the coming years, unlike the Max Payne engine between the release of Max Payne and it's sequel. Yes, it's possible Source will be updated in time to still play a significant role in this upcoming generation, but that is not certain. It's still in the air, could go either way. We can only wait and watch. I hope it does. Gamers can only benefit from the addition of one more great engine in the next generation. But to assume Source will save the day and come out ontop isn't very realistic at this point.

 

[MultiVaC]

You need to accept the fact that unified lighting isn't the only thing that matters. Currently, unified lighting looks a lot worse and hogs a ton of resources. Of course it has it's advantages, and id had chosen to accept and work around the downsides. The same can be said with lightmaps. Lighting aside, the Source renderer is acutally quite competitive; great shader support and excellent flexibility. From what Valve has said the Source seems to be extremely modular, making parts of it easy to add and replace. If so, an new lighting system can be implemented at the whim of a developer (probably not for HL2, but for other projects). It's been done with other engines, I'm sure it can be done on Source. Nobody here is assuming that Source will come out on top of this generation, you are saying it has no place in the future, which is not true at all.

 

[OCybrManO]

The method used in Doom 3 isn't exactly a mystery that only id Software can solve (UE3 is more complex but it's still no secret). Hell, it's even been done on the original Quake engine by amateurs in their free time. Valve just chose not to use it. By the time engines with the features of UE3 start to pop up Valve could probably rewrite the renderer and fix the light placement of the maps to compete with them (if Source is as modular as they make it out to be).

Right now, Doom 3 lighting just doesn't seem to fit Half-Life's style (its results are too sharp and uniform), but when the lighting is realistic enough and not too demanding most games will move over to a unified lighting solution... if only to cut down on development time. As the amount of detail in models and maps increases every year games keep requiring more and more work to stay top-of-the-line in terms of graphics. If developers can avoid huge map compile times with an efficient unified lighting system (like Doom 3's maps) they can make more maps in the same amount of time (or the same amount in less time) and they will be more likely to fix tiny bugs that would otherwise be a waste of compiling time.

For now, Source is fine. I'll start worrying if they don't do anything about it in a couple of years.

 

[crabcakes66]

By 2006 everyone will have a DX9 video card.

:dozey:



Id say less than 20% of pcs have DX9 GfX cards.


Mabey less than 30% in the general gamer community. less than 40% in the "hardcore" gamer community.



Those are just my best guess but Id say they are close or if anything they are high. Even if the numbers double in two years.....thats still only 1/3 of people with DX9 hardware, and i even doubt that figure...considering most people currently have DX6/7/8 lvl hardware.

 

[blahblahblah]

According to the steam survey they did a while ago, roughly 27% of Steam users having DX9 Hardware. That number isn't completely accurate since a 10% or so are Geforce FX 5200/5600 which is not fast enough to run HL2 in DX9 mode.

The anticipation of HL2 and Doom3 has done wonders for the adoption of DX 9.

 

[billbo]

According to the steam survey they did a while ago, roughly 27% of Steam users having DX9 Hardware. That number isn't completely accurate since a 10% or so are Geforce FX 5200/5600 which is not fast enough to run HL2 in DX9 mode.

The anticipation of HL2 and Doom3 has done wonders for the adoption of DX 9.

It was enough for me to finally upgrade.

 

[Mr Neutron]

I'm betting the UE3 demonstrations were run on NVIDIA's previously unannounced SLI setup. If true, this changes my previous impression: I thought, 'Hey, the 6800 with it's 3.0 shaders is a powerful board, but damn, they must have done some very clever programming to get those results!' Now I think, 'OK, hardware with more than 4x my 9800pro's horsepower? Well that's what I'd expect, really.'

I'm in no position to dis the talent and hardwork of Tim Sweeny et. al., but it seems the days of hotshot lone programmers like Sweeny and Carmack getting a generation ahead of everyone else are over. Forgive my lack of sentimentality, but this is a good thing: the cycle of progress is starting (I emphasize starting) to get shorter with each turn .

 

[Rupertvdb]

aww, i need benchmarks, i have had a crisis of card and now don't know whether to get the x800 or the 6800....

 

[The Dark Elf]

awww lets not turn this into a which card thread.

 

[^Ben]

Just remember to use this new fangled SLI you need 2 ultras expensive mondo!

You need a motherboard which supports 2 PCIE slots only one of them out at the moment. Expensive Mondo and what processor can that motherboard support.... Xeons Super Expensive mundo!

 

[HeadcrabOnMyAss]

The nForce4 Chipset will have dual PCI-E slots, and will be a smidgen cheaper than Tumwater, or whatever the hell the only current dual PCI-E solution is right now.

But you still have to shell out upwards of 500 USD on each of the 6800's you'll have to use. As of now, I don't believe the lower end GPU's have the bridge connector for use with the SLI bridge-chip, and I'm not sure ATI's cards are capable of SLI right now.

You'll also need a massive power supply, and that'll cost too. So unless your made of cash, or are Bill Gates, I'm thinking you probably won't be able to afford this gfx solution.