View Full Version : Project Reality 2 (Standalone CryEngine 3 game)
Amit
September 22nd, 2013, 04:33 PM
So yes, this is some very exciting news. Project Reality will no longer be limited to just being a mod for BF2/Arma 3. The PR team has decided to use CryEngine 3 as the base for their new Standalone game and I have to say, just based on some of the screenshots, the maps and atmosphere are going to be awesome. PR was as close as you got to realism in a game, but the visuals were not. While it's impossible to get truly realistic graphics in a game, CryEngine 3 is arguably one of the only engines that almost achieves it. So not only will maps be better designed, but art assets will also be way better looking. Just check out these screenshots below:
http://media.pr2game.com/news/announcement/lashkar_gah_01.jpg
http://media.pr2game.com/news/announcement/lashkar_gah_03.jpg
http://media.pr2game.com/news/announcement/fools_road_02.jpg
http://media.pr2game.com/news/announcement/fools_road_03.jpg
http://media.pr2game.com/news/announcement/us_army_soldier.jpg
http://media.pr2game.com/news/announcement/kalashnikov_ak74.jpg
http://media.pr2game.com/news/announcement/kalashnikov_aks74u.jpg
Check out the full announcement HERE (http://www.realitymod.com/forum/f380-project-reality-news/124409-project-reality-2-game-announced.html).
Warsaw
September 22nd, 2013, 06:28 PM
Isn't ArmA already a military simulator? Doesn't that kind of render Project Reality redundant?
Zeph
September 22nd, 2013, 07:53 PM
Isn't ArmA already a military simulator? Doesn't that kind of render Project Reality redundant?
Just because it's in a genre doesn't mean much. That's kinda like saying that Halo is redundant because 'other sci-fi game'.
Ultimately a game is bunch of numbers set up to look a certain way. Using Cryengine as a platform means they get to work up from scratch without having to build off of another method and change it to suit their needs the best they can.
The more I dig into Cryengine and look at how much resources things use, the more I want to gut the code even further and use things more efficiently to get what I want. Last week, I was stress testing animated characters to see how many of them I could get running around at once. A bug left me only able to have 64 of them animated at a time, but I was easily able to get 400 of the sample character moving uniform in formation across a map on the screen at once (meant there was no stuttering between frames). Moving in closer really cut down on my framerate, but through profiling I found that the facial animations were set to be visible from an incredibly high distance. Cutting that down to something like a single meter in distance raised performance like nothing else. Looking at it further, I found that an event listener attached to each game object (inheritance basic entity -> game object -> game object extension thing you want to make) was taking up 95% of the frame for those 400 animated characters even though I have no events they need to be listening for. Should I import a character model with LODs and remake the animated character to not have that event listener, then I imagine I could fill the screen with well over a thousand animated characters running around with Cryengine easily.
If they actually pull in proper programmers to go in and hack out the generic SDK stuff they don't need in the game itself, then they can really get some amazing things going on.
Amit
September 23rd, 2013, 01:15 PM
^ That basically. They did announce a switch to Arma 3 for PR2 (not sure if that's still happening, probably not), but while Arma is considered a "simulator", it's just more of an FPS with a higher learning curve. Arma 2 and 3 play like games with more options and complexity. That's it, really. The Real Virtuality 4 engine isn't all that great anyways. It may work for Arma (and even then it doesn't run well), but it does have its limits for something like the PR experience. Project Reality aims to actually go into the intricacies of the handling weapons, vehicles, and communication. Arma does a nice blend of all of those, but with less realism. I'm thankful for that since I don't think I could play Arma for very long if it's like PR all the time.
Warsaw
September 23rd, 2013, 06:07 PM
If that's your definition of a military simulator, then a flight simulator is just an arcade flyer with a higher learning curve. That's being a bit disingenuous.
Just because it's in a genre doesn't mean much. That's kinda like saying that Halo is redundant because 'other sci-fi game'.
Ultimately a game is bunch of numbers set up to look a certain way. Using Cryengine as a platform means they get to work up from scratch without having to build off of another method and change it to suit their needs the best they can.
The more I dig into Cryengine and look at how much resources things use, the more I want to gut the code even further and use things more efficiently to get what I want. Last week, I was stress testing animated characters to see how many of them I could get running around at once. A bug left me only able to have 64 of them animated at a time, but I was easily able to get 400 of the sample character moving uniform in formation across a map on the screen at once (meant there was no stuttering between frames). Moving in closer really cut down on my framerate, but through profiling I found that the facial animations were set to be visible from an incredibly high distance. Cutting that down to something like a single meter in distance raised performance like nothing else. Looking at it further, I found that an event listener attached to each game object (inheritance basic entity -> game object -> game object extension thing you want to make) was taking up 95% of the frame for those 400 animated characters even though I have no events they need to be listening for. Should I import a character model with LODs and remake the animated character to not have that event listener, then I imagine I could fill the screen with well over a thousand animated characters running around with Cryengine easily.
If they actually pull in proper programmers to go in and hack out the generic SDK stuff they don't need in the game itself, then they can really get some amazing things going on.
Not quite on-target with the analogy.
I thought the point of the ArmA series was to create a realistic and comprehensive representation of modern battlefields. PR is essentially claiming the same thing. As a mod on top of a military simulator, it seems like a duplication of effort. Making two science fiction games, however, is not even remotely the same because they don't share anything inherently except a subjective genre. Basically, this is a realism mod for a realism game, kind of like those many flight model and audio patches are for IL-2: Sturmovik. As games, neither PR nor those flight mods make a whole bunch of sense.
All that said, the last time I played PR it was still a BF2-only affair. It was much more gamey than ArmA was.
Zeph
September 25th, 2013, 09:39 AM
Nah, it's pretty on target. PR2 wants to make the one ring imeangame to rule them all. They're kinda like the people here when it comes to mapping. They're stuck within the confines of the editing kit they're using. People using the HEK had access to the weapon tags, player tags, and structure bsp tags but that's about it. Those are cookie cutters for existing functionality.
If you wanted to make a weapon that pierced certain materials up to certain depths and base damage dealt on what collisions were registered on the model's internal structures, you're out of luck unless you happen to come across the source.
If you wanted to force a player to blink or spasm when they were shot, you've got to hope there's a way you can abuse local scripting to achieve the effect instead of more efficient source code.
They're the same genre, but very different games. I can understand where you're coming from though; you see that kind of view/attitude on crydev. I generally blame it on the lack of diversity AAA production has led to. With huge amounts of money being pumped into a single title, the lowest common denominator was catered to. Selling points to the masses quickly became visuals and cinematic scripted trailers instead of mechanics. Indie dev and digital/self publishing (thanks valve for showing this could be done well before the mass market was prepared) started to correct this and it's starting to speed up with more 3d SDKs. Unity picks up more devs as it continues to update itself and UDK will start to get more diversity when it opens up its source in Unreal4 (face it, unrealscript and kismet can only do so much). Cryengine has a lot of pretty trailers, but very few indies have actually picked it up to the point of licensing. Those who have, are pretty quiet because it's such a mess under the hood trying to figure out how things are pieced together across the 14 years of source code updates. You wont see good community documentation until people put money in their pockets from their release first.
Powered by vBulletin® Version 4.2.5 Copyright © 2024 vBulletin Solutions Inc. All rights reserved.