For me the content quality and detail level of r3e can stay as is, as it's pretty and crisp. Only finding a nice weather and dynamic light solution that is consistent with the achieved quality is critical. I'd hope the game keeps its current style, and doesn't turn into pcars. Nice to have are improvements to in race fx (small car damage, dirt/grass interaction, oil on road...) Bonus points for improved shading of interior and more animated characters, garage... Pre/post race. That would matter in a VR future 2-3 years from now.
Did anyone watch the AMD Capsaicin event at GDC this week? They explained the benefits of DX12 much better then I did. I can't find the full webcast but the following is a pretty good start.
Ok, the graphic of R3E is not really state of the art in comparison with Project Cars but the isiMotor engine offers everything what a small sim studio without own engine development like Sector3 is needed. The rFactor sim which use the same engine since 2005 offers dynamic day-night-day transitions, different weather conditions, … so I think there is no need to change the engine. There are a lot of other programming work like a pace car animation, flags-system, realistic damage model, … which are more important than a better graphic.
Not the same engine. Rf2 = ISImotor 2.5 and AFAIK RF2 is the only game that use the ISImotor2.5 R3E (and GSC/AMS) use previous version.
Even further beyond that... ISI motor is comprised of a graphics engine and physics engine. rFactor 1/2 use both of those in combination. Same with GSCE/AMS, it uses both in combination which is why it looks identical to rF and feels the same. R3E uses a highly modified version of the physics engine, along with it's own version of a graphics engine, independent of the rF graphics engine. So it isnt that Sector3 can simply flip a switch to have rain(or change a value in an .ini file like "enable_rain = 0" and make it "enable_rain = 1").
indeed. And there's different licensing option that allow different levels of access to the motor. For what I understand, Reiza went a step further into their ISImotor licensing allowing them to modified it even more. And they licensed the latest version for their upcoming title (Reiza2017), but with limitations like "no modding allowed"
Sector3 has the full IsiMotor engine license, from what I've been told, there isn't much left of the original engine code, it's been modified to within about 20% of it's life to allow for a lot of the physics we wouldn't otherwise have and that you don't get on other IsiMotor engine based games.
This part isnt quite correct. Reiza licensed the source code for ISI motor 2.0, which is what GSCE/AMS is built on. The most recent ISI motor is 2.5, which is only used by rF2, because ISI doesnt want to license it out yet. So, basically Reiza now has access to the source files to the original rFactor 1 engine, to modify however they see fit. Previously, they couldnt modify the source files, they could only use "code injection" to make changes. Which is a very sloppy way to make improvements. Reiza 2017 is still using ISI motor 2.0, but with the improvements that Reiza is making now, that are being implemented into AMS, like tire damage, flatspots, a pre-defined track grip, etc. Yep, Sector3 has done remarkable things, but they have been at it with the code for quite a while now. It will be interesting to see where they can go from here, as the improvements they are making, and some yet to be released are really moving RaceRoom forward. Honestly hope they can keep on going and stick with R3E for a long while because it is really turning into something great. And if there is ever a new graphics engine update (which certainly isnt coming soon, and as mentioned, isnt the focus at all) or for a future game, they will have my support
This whole motor licensing thing give me headache It's even worst that the Porsche Licensing deal to understand But at the end I don't care what engine is under the hood (not talking about cars here..) as long as I enjoy it a lot
Hm, I think I just quote what I just posted at AC forums on RR visuals: To me, RR right now offers the niciest environment looks there are in the genre. The engine may be old - but the looks in RR still are giving me a blast, I just love them. I understand however that they need a new engine probably if they want to do more wizardy with dynamic weather, night time and multiple light sources. And
Well there you have it - the fate of your city depends on you upgrading your PC. Not sure i've ever heard a better excuse to upgrade
The look of the game might not be the only reason to a drastic update of the graphic engine. Sometimes underneath changes are needed to allow the game to reach futur goal and improve itself. A Dx9 engine might be good for know but it has it's limitation that may block futur improvement, support of futur hardware (good VR, etc), when a Dx11 engine allow to go further with less downside. We already can see this now, when racing game based on dx9 engine suffer from stuttering even when you have a high and solid frame rate.. while title based on dx11 run more smooth even at lower frame rate. I can compare AMS/R3E with Dirt:Rally (or Asseto). There's always times/moment when I get some stuttering (even with some solide 90FPS on a G-sync monitor) in R3E/AMS ... in Dirt:Rally I run it between 55 and 75FPS and it's smooth as silk. I read somewhere that iracing suffer the same issues between it's old dx9 engine and the new Dx11 one . When you have an always developing game like R3E (I really don't think there will be R3E 2 or any new title before a looooong time, R3E has for me the same dev schedule than a iracing), there's point in time you should ask yourself if it's still worth the pain to tweak some old engine built on old techs or moving forward to be more futur proof and get a cleaner code to deal with.
^ I will not object to that. I think it also is about limitations in the current engine not allowing mutliple light sources for car headlights, lightposts, etc. Virtual headsets are another good explanation for their move, yes. BTW, I never have any freezes or stutterings in RR, never, and its always running smooth as silk, always.
Well, let's have a thought. You ride looks like this: Would you take this? Or would you take this? Make your choice!
for what is worth, reiza also mentioned they are/were experimenting with unreal4 for graphics. there is a lot of propaganda around dx12, one hw vendor tends to benefit relative to dx11 more and one software vendor wants you to upgrade os for it It is very useful in some particular scenarios when you are very limited by the dx runtime/driver (all CPU side). However, it doesn't add anything majorly interesting in terms of GPU usage vs the last variants of dx11. So if you are GPU bound, it's not gonna magically save you. Article by game developer on their dx12 porting http://wili.cc/research/northlight_dx12/GDC16_Timonen_Northlight_DX12.pptx For the similarly minded Vulkan I had written & illustrated an article about what to to expect https://developer.nvidia.com/transitioning-opengl-vulkan There comes a lot of extra responsibilities with the low-level apis, that used to be inside the driver. Getting this "right" and well for multiple GPU vendors is a good deal of work. And so far most dx12 titles were lucky to be close to dx11, and the developers were experienced veterans. Though they are up against the veterans in the driver teams on the dx11/gl side. That said if you can design from scratch with all the modern approaches fully utilized (as the ashes guys did) and you can make use of all that extra CPU time for something else (or draw tons of tiny units in an RTS game), then it's definitely worthwhile. And in the long run people should design with stuff like that in mind. This style of api is relatively fresh, so we may see changes along the way. Personally I think track racing games are not that "crazy dynamic" to benefit a lot. The main benfits lie in some of the features useful for VR (replaying command-streams), but other VR related stuff will be available via driver backdoors for dx11 by the gpu vendors, too. As I am not an actual game developer, the following impression could be wrong. The majority of the content in a racing game is rather static and unaffected by the player, unless you want to be able to destroy the world beyond track limits with your or your opponents car Shading wise you also don't need super much variability and quality, the majority of content is rather far away, you are not gonna notice lacking bump-map detail on some barbecue in the Eifel. Cockpit and garage are the major things that benefit from fine detail, but it's also a controlled environment, you are not gonna set your cockpit on fire (garage maybe), you are not gonna see crazy customized race driver avatar as in some RPG games, or the cars being uniquely modded a lot like in underground racing. Other effects like tire marks, marbles, oil, weather.. are taken care of by specialized systems, that handle a high N of such elements well, and the jump from dx9 to dx11 capabilties is huge in terms of GPU generating and modifying content without CPU usage. One aspect where we really have unique situations that could add a lot of variability is car damage/destruction. Let's say you want a big grid with many different car models and all of them partial destructible as in dirt rally, then it's gonna be a tad more challenging in dx11. Because whenever you want to draw a lot of stuff that has its own state permutation from many different shaders, many different textures, many different geometry buffers, then it becomes harder and that's where the new apis shine and where you need to be a bit more creative with the old apis. But the old apis don't add a lot of other tasks on your burden, so it's a balance that must be well understood. Good thing is we may not need to be able to strip the car down to the crashbox, if we think of 24 hour races, it's not like the cars keep driving completely broken, or that all cars explode at once What we need is a few more dents, duct tape, smudges, flappy things, flat tire damage... Down the line, in a few more years dx12 or vulkan at whatever version they are then, will be more "practical" relevant, and it's great to have it on the developers mind where the direction is going and build up the experience. edit: just realized the cockpit is probably getting some dynamic weather influence in an open-wheeler. so not super static
Never MP, and the only two tricks I always use is to have AA and AF, iv available , never set to maximums like 16x, but only to something like 4x, since that already gives most of the effect visible, but saves tremendously on the gfx board resources, if their is a shadow slider allowing low, medium and high settings, never to move beyond the medium setting. Two golden rules: very easy, and very effective in most situations.
This is great news without a doubt. And I think it'll bring the game a big step forward in the right direction. Virtual reality is on it's way and it's important to jump from the DirectX 9 train as soon as possible, not only in matter of graphically fidelity. BUT, as a customer for years who invested a lot of money in this game, I honestly have some fears in terms of when we truly will be able to see (or use) the first results of this new engine stuff. Will it take another years? Is this the next beta thing we have to deal with, even if the "main" beta game is still not "finished"? And by the way, will it ever leave the beta phase? Where the heck is the 2015 content of ADAC GT Masters or WTCC? It's 2016 now and still no sign of it – except from the screenshots and the "soons" we see all the time. Okay, I know the Bentley stuff was holding the content back. But seriously, why not releasing the new pack with the note, that the Bentley will be added later for free? Speaking of ADAC GT Masters, will we get truly "new" updated content or will we still have to deal with car models from 2012? R8, I'm looking at you! Please don't get me wrong! The game is one hell of fun and I told several times how much I love it. Or loved ... I made the decision some weeks ago to delete the game. Announcements are important and also to "talk" to your customers. But even the biggest patience has it's limits. Almost every announcement for new content, features or even for fundamental things comes with some kind of aftertaste: silence. There are too many construction zones replaced by even more announcements and not getting touched for an unknown amount of time. Old (payed for!) content becomes obsolete because it's (still) not updated with new features, physics for example. Maybe I'll give it a try again when everything is done and the game is finished or has all it's features implemented. Unreal Engine 4 is really something and looks breathtaking stunning (watched some Kart Kraft footage). So maybe in 3+ years then?!
If R3E is in fact still in 'beta' then it's surely the most expensive, long-running beta I've ever experienced. Much like the 'free-to-play' tag, the whole beta thing needs to be dropped too. It's a continual development game much like iRacing and rFactor2. Also...agree with everything else you said.