It has been said a lot of times, but I must repeat it untill it will be clear for anyone: DX10/11 wouldn't change anything. The problem is in gfx engine itself. not in gfx API. newest DXs gives nothing to make current ISIs engine faster, better, more optimized or real-looking. In other words: it is possible to develop gfx engine getting high quality results as on DX9 as well. After that you shouldn't be shocked that rf2 utilizes DX9. you should be shocked that it implements old gfx engine with a lot of issues and wrong approaches (even if a bit improved comparing to rf1)
Yeah I guess that's the problem. But still the quality between cars-tracks is really unbalanced... we have very few guys with talent working on tracks.
I believe many of those talented car modders could do well with tracks too, but it is just so much tedious work with tracks that car making is "easier" choice. To make good tracks, you have to be talented but also completely crazy. I almost only make cars because track making with 3DS Max (great tool for cars) is very frustrating and tedious. Example, when you only need to plant 10 000 trees in specific places and positions you have to do it by hand with 3DS Max because its seeding scripts suck big time. Even that alone puts me off, not to talk about other inconviences with that software. Car making its a lot easier. Maybe my problem is that I don't like modern "parking lot" tracks.
You don't have to do that by hand....you can use scatter/array tools. In my opinion, track modeling is easier than car modeling but you need more time to build the track.
Modelling track takes more time maybe, but gathering data for car takes even more. unfortunately in both areas modders ignore important things releasing cars with fantasy suspension geometry as well as tracks with low quality (just example). Since rF is intended to be the sim, we need better quality in all areas.
Problem is that scatter doesn't work very well, and you end up altering everything by hand. Array is not useful when random placement is required. It is hard to tell which takes more time, gathering track data or car data. Imagine situation where you want to build real life track that existed in 1910's which today is suburban area. Or that you want to make a very car you own in real life. Which would be easier to get it right? It totally depends on the situation. But my point was that actual building process (3ds max) for tracks takes a lot more time, frustration and tedious work than making a any car. Sure, you always ought to have required references what ever you are making. Physics work for cars is totally another ball game, but having its experimental nature, it is nothing compared to suffers of track making.
Ask on the forum perhaps? Look at modding subforums - people are making questions while working on their mods, they get answers, fix problems and move on with development.
Well you need to do some research mate. If the engine was build up on dx 10 or 11 (what i thought it would verry first) it would also run much faster on newer systems. Every GFX card cappable of DX 11 would benefit of that. But you can try that out yourself if you got a DX11 card. Get yourself any Game out there witch uses both dx version. For example F12011, run it with DX11 and you`ll see the improvement---and forget the effect, these are just eyecandys,i talk about the whole performance. Anyway its too late and also just my thoughts when i first hear about rf2 saty on dx 9 wich is a long time ago.
Not to argue, but I have to disagree. When Rfactor first came out I had a video card which supported DX9, but running in dx9 gave me single digit framerates maxxing out in the teens lol. Running it in dx8 gave me smooth racing maxxing out over 100.
Doesn't work that way in either case. DX10 and DX11 are not only more efficient than DX9 but are so removed from DX7/8/9 that they have to emulate DX9 support. Still, a minor framerate increase scarcely seems worth a lengthy and expensive recoding of huge chunks of the graphics engine and cutting out about 20% of potential customers.
First I agree with ethone's initial post. Second, there is one important element I am missing in this discussion: Talk about filters So far you only discussed in absolute terms whether conversion improve rFactor or hurt it. What hasn't even been looked at is what filter to we have and what filter do we need to seperate the good from the bad. rFactorCentral has been mentioned which in my opinion is an example for a very bad filter, because it collects everything, it's rating is worthless and it doesn't give you means to decide if the track is worth checking out. Make modding easier, make it accessible for everyone, have a large tail of weak quality mods, but enable more people to get into modding so that a few more can reach the Olymp of doing really high-rated awesome mods. We have too little people there right now. Have a long-tail and have ways to filter it away. What we need are ways to bury bad tracks and to dig up good stuff. Highlight the great work somebody did, while easily ignore the weak.
I don't think there can be a perfect filter for everyone. For one thing (and as this thread has shown) everyone's personal quality standards are different. For another, different people believe in different measurements of quality. So a proper solution from my point of view would be a service implementing different filtering methods and allowing users to chose from them. - A simple and uncontrolled user-rating system like rFc had is probably the one filtering method we all agree is crap, since we saw how it failed. - A seemingly objective measure like number of downloads can be manipulated as well (and seeing what happened with the rFc rating, it can be taken for granted that it would be) and in any event would not properly indicate quality as the user would define it. Some tracks like those run by F1 will be by default more "popular" than a quite possibly better quality track that is only run on a single weekend each year by the SCCA. - A authoritatively prescribed quality index likewise is subjective (the antithesis to a good filtering) and can easily descend into favoritism - Screenshots, while not an obvious filtering method, would allow you to self-filter based on the shown graphical quality. It would be preferable to have these screenshots done by the same person/group of people to have comparable quality standards (Full level of detail, Shadow quality, same FSAA settings, comparable camera views). - My personal favourite filtering method is just trying a track I'm interested in and filter it myself. If it turns out to be a bad track, I just uninstall it. I thought of doing occasional "track reviews" on my blog but hadn't bothered so far as I couldn't hope it to be anywhere near objective. I guess though, as with conversions, having anything is better than having nothing in this regard.
To be honest, i would love to see RF2 finished in 5month even if i know its not even possible to get to the product the most of us are looking for. There is way to much todo! We have seen what they can do within a month and therefore its nowhere possible to see a final product this year. Thats really not a Beta stage we`re all into. I´ve said before, it would be good to call it a Alpha for sure. And against the pm`s i get about that im a sms fanboy, come on. Even if the handling is not like rf2 its even yet a more complete game then rf2 will be in 3 month for sure. The updates sms deliver are way more complex ( remember they do weekly updates also), and you can drive more cars on more tracks in that Pre Alpha. So please BE serious now. If pcars get to the handling (wich they will) Rf2 needs a very very good reason to stay on. If you compare just some screens of pcars with the latest rf2 build 60 there are worlds between them. And why do i buy and build a high end Pc ??? To look at Nvidia`s 8000 series graphics,,,come on. And at least i can very eays mod my rf1 to look way better then rf2 does!! THATS NOT AGAINST ISI!!! But the support on this forum is bad and so is the rf2 progress right now. ISI change your mind please and take more care of your customers. Even try to answer a mail. Modmate
RF2 has exactly the same chance in terms of graphics to be as good as CARS is, like CARS in terms of physics to be as good as RF2 is ;-) But to be honest I believe that new shaders and some lighting improvements will help reach RF2 much higher graphic level, but I just don't belive that some tweaks and improvments can bring to CARS such a good physics and FFB as RF2 has.
Sorry mate, but ISI`s engine is far away from being able to do Graphics and effects .... like the Shift2 engine does. They can get a bit better but thats it. Instead of using such an old engine they would have gone better to buy one Even the dirt2 engine has much more to offer trust me. I make a bet with you that at least on the graphics rf2 will stay way behind pcars. And like i said , if pcars dont get the handling rf2 has to offer i will stay here, but just if...you`ll see what happens. Dont forget Support and Updates!
Remember, pCARS is not a modding platform. That's the main difference between it and rFactor. Physics wise... again - with rFactor 2 you can check by yourself, what it offers. With pCARS you have to believe what Ian says, because you won't be able to check it. And ISI gfx engine not beeing able to look like Shift 2... what forces ISI to not improve their gfx engine? It uses 9.0c API. Now it's a matter of using all the best from what that API has to offer.
@modmate I agree with you, it will be difficult to reach CARS level in terms of graphic. But it will be much more difficult for SMS to reach rf2's physic level. The good thing is that we have a choice. I prefer advanced physics And that's all I wanted to say.
I have read the discussion carefully. And I have to reply to one specific argument: I don´t think that even bad conversions would hit the here discussed programm seriously, or even let it look bad in the community or even let it die! That´s far to exegerrated! Think: The whole simracing-community is - in the end - one big soup. Everyone who favorites a special simulation or game, knows about the others or even tested it him/herself, what normally will be the case. So also everyone knows about conversions (not only because they are marked as this in the readme-file or online-description) for everyone of us and them tested a few in his/her life in different games. So really everyone will, or just do know about good and bad tracks! And that is something we cant 100% stop anyway: bad (looking) tracks! We can decrease theire number of appearance, but there always will be the one or the other noob who does c*** - eeeh, no good work. So for me it´s not a question of conversion or not conversion but of a most possibly high level of quality in track/mod-making. greets and thnink deep! Pete P.s.: pleeeeeease no more bull**** about DX9!!!!! It´s said often enough that DX10/11 does nothing better!
They're much closer than you think. They just need to drop the old way of making materials and implement something more after reality.
