Hollywood Physics *no spoilers!*

theholygod

Newbie
Joined
May 6, 2007
Messages
136
Reaction score
0
I was actually astonished how awesome buildings being blown to shreds looked in episode 2. I read quite a lot into the physics engine and really wasnt sure what to make of it, but i thought in the end it looked completely believeable (albeit no fire) and one of the most impressive sights of destruction in any game to date.

What did you guys think?
 
Yeah, there was some destruct-ee-un in HL2 and E1, but not on the same scale as E2.
 
It was seriously real looking. They hit it right on the spot.
 
My jaw hit my desk watching the bridge collapse at the beginning...
 
No loss in frame rate either which is lovely
 
I'm guessing physics for the cinematic physics are calculated at compilation of the maps, stored as an animation which is played in the game. This is because the objects don't interact with anything that wasn't there during compilation (debris flies through characters and such, not that it's a problem at all). Or perhaps the physics are stored like they are stored in a demo file.

I wonder if the breaking can be calculated by the compiler too. Like if you can see "I want this wall to break realistically" and Hammer calculates the breakpoints and pre-destructs the wall for you. Or if you gotta create every piece of debris as a seperate brush and link them together with constraints, which would be incredibly tiresome to do.
 
I read on wikipedia that the physics runs partially as an animation
 
I read on wikipedia that the physics runs partially as an animation

It would make sense, something as awesome as the bridge collapsing at the beginning would of definatly lagged my system using normal physics with lots of props.

It was amazing, not seen any other game close except maybe the AGEIA PhyX game where you had Telekinesis but it sucked and lagged REALLY bad.
 
No loss in frame rate either which is lovely

I actually don't think Episode two lagged significantly even once during the entire play-through. And that's with a not-so-impressive rig, running on higher than recommended settings and with a slightly dodgy graphics card.
 
I was completely impressed. I wish I had read less and viewed less "preview" footage as that stole a lot of the awe factor ingame. Regardless, each event looked realistic and dramatic... more so ingame then was experienced through videos.

Awesome work.

--

I would also agree with the exceptional performance even on older rigs. I have one of the first AMD dual-cores and an older SLI setup and pushed all the settings (save resolution) as high as they would go. I didn't see a single hiccup.
 
I actually don't think Episode two lagged significantly even once during the entire play-through. And that's with a not-so-impressive rig, running on higher than recommended settings and with a slightly dodgy graphics card.

Ooh, describe your rig... I've got an older system, and I'm wondering whether to hold back until I upgrade, which will be forever, or just go for it..
 
Ooh, describe your rig... I've got an older system, and I'm wondering whether to hold back until I upgrade, which will be forever, or just go for it..

AMD 3500+ 2,1 ghz, 1 gb ram, ATI x800XT 256mb (wich as I said before, does not function as well as it used to).
 
Yeah, the physics in the game were/was? absolutely amazing
 
I was really impressed with how well Episode Two ran with higher-than recommended settings. No stuttering or serious loss of framerate even during the huge battles or the cinematic physics spectacles.

The only time I noticed a slight loss in framerate was when I turned the flashlight on.
 
The only time I noticed a slight loss in framerate was when I turned the flashlight on.

Yeah, that's quite annoying. It's due to the fact that VALVe added some new sort of shadows in ep2. Not sure exactly what though.
 
Nitre said:
Yeah, that's quite annoying. It's due to the fact that VALVe added some new sort of shadows in ep2. Not sure exactly what though.

Dynamic shadows. But I don't think it should be as bad as it is. I don't think it's normal, look at for example, Doom 3. It doesn't hurt performance as much I think.
 
I think comparing the half life series to anything by ID is a little unfair.

I love both VALVe and ID, but VALVe seem to focus on realistic phsyics engines whereas ID focuses on getting the most incredible graphics ever running on a 5 year old graphics card.
 
Dynamic shadows. But I don't think it should be as bad as it is. I don't think it's normal, look at for example, Doom 3. It doesn't hurt performance as much I think.

I noticed that the flashlight slowdonws are less evident in closed ambients (corridors and so on, and Doom 3 is all corridors). Since open ambients don't need flashlight, in my opinion this is not much an issue. But I keep thinking that dymanic shadows can be optimized.
 
I love both VALVe and ID, but VALVe seem to focus on realistic phsyics engines whereas ID focuses on getting the most incredible graphics ever running on a 5 year old graphics card.
That's just plain wrong. Any HL game runs better on a DirectX 6/7 card than a Doom 3 game (if at all it runs). I've tried it firsthand. Quake 4, for example, is unplayable on a ThinkPad R60.

Valve makes awesome, beautiful games, and id makes bleeding edge engines.
 
Back
Top