Quake physics in Unreal Engine
#11
Posted 01 August 2015 - 04:13 AM
UE4 changes this by first adding what would be considered a real world model as to everything that includes how things move as well as materials, cameras, and lighting that you would expect in a high end rendering system like Mental Ray.
Another consideration most things in UE4 are hardware rendered so using 125 FPS as the benchmark as to performance is not as critical as to things like draw calls and fill rates. In other words that really expensive high end video card will give you an advantage where it really does not do much to add to a game based on a 15 year old engine.
#13
Posted 02 August 2015 - 11:53 AM
So lets hope our team finds a coder now or our old SoF2 coder is willing to team up and try this out as its a dream of him aswell.
Btw, sorry for the late response but I couldn't acces the forum for a while.
#14
Posted 02 August 2015 - 05:44 PM
BladeKiller, on 01 August 2015 - 09:03 PM, said:
wait just a second! then why did zenity just say
Zenity, on 27 July 2015 - 12:15 AM, said:
this makes no sense now. by what you just said and what he just said, means that players in ue4 now jump higher than players in urt 4.2 currently?
#15
Posted 02 August 2015 - 08:03 PM
Dark-knight, on 02 August 2015 - 05:44 PM, said:
Depends on what you mean by jump.
Do players jump higher in UE4 as they do in 4.2 as a function of code?
No
That's just a math problem already part of the movement component where a few values can be typed in that tells the component how high and how fast the player can jump. As Zen mentioned this can be pulled directly from 4.2 so it is the same.
Do players in UE4 jump higher as a visual?
Yes and no but. Yes they seem to jump higher because UE4 input is much faster and reactive, I suspect due to client side management once served up by the server, and no because the visuals are still bound to the rules of hard coded game speeds that requires adjustments be made to make sure the players actions keep up and syncs.
#16
Posted 03 August 2015 - 01:16 AM
Dark-knight, on 02 August 2015 - 05:44 PM, said:
if TwentySeven fixed it years ago?
this makes no sense now. by what you just said and what he just said, means that players in ue4 now jump higher than players in urt 4.2 currently?
Nah, just a matter of semantics. :) 27 did indeed fix it so you jump the same height in 40 FPS as in 125 FPS, but he did so by essentially running your physics always at 125 FPS, not by actually fixing the underlying issue that causes this (which may sound weird but it makes sense, because the issue is related to Q3's networking optimisation). So without breaking the physics the same way (which would be a bit silly), we cannot replicate that exact jump height. Instead we tweaked the jump height a little to match.
We have tested this with with a box you can just barely jump on and one which you can't, which differ only by a tiny amount in height. There are still going to be hairpin differences in jump height, but it shouldn't be noticeable or affect the average trick jump.
#17
Posted 17 August 2015 - 11:03 AM
#18
Posted 17 August 2015 - 12:32 PM
KarlMariaSeeberg, on 17 August 2015 - 11:03 AM, said:
Agile movement seems to be all the rage among console shooters lately (Titanfall, Advanced Warfare, there's also been Brink and Dirty Bomb already). It's a positive development IMO and should make those games a good bit more interesting.
No AAA developer is touching Quake style movement physics though, I expect that will remain the domain of indie games for a long time to come.
#19
Posted 17 August 2015 - 01:10 PM
This post has been edited by KarlMariaSeeberg: 17 August 2015 - 01:10 PM
#20
Posted 17 August 2015 - 10:50 PM
What really separates the games? It's mostly the aesthetics, price, and all the fluff Call of Duty puts on top (levels, killstreaks, etc). Also (modern) CoD is heavily focused on gamepad controls, whereas Urban Terror is still a pure keyboard and mouse shooter.