CG vs. Non-CG

[camsinny]

I apologise in advance because this thread is not in a light mood and may anger/upset some people, but this is my view, feel free to post yours.

Recently Racer has come leaps and bounds and now rivals the graphics of GT5 and IMHO can be better. So it really irritates me when I see all the crap that is being released these days. I'm not going to deny it, this post really rubbed me the wrong way:

concept is cooler than creat real cars .. real cars i prefer to convert, it is easier and practical than waiting a good person called angel to do from scratch a car that i like very much. After Tiberius's jag and my booster, harey's caravans and anothers, we have no more new car made from scratch for here..

This is in no way an attack solely on AMGfan, I know he's good at converting cars but therein lies the problem.
When cars are created for a game the artist creates the model around the game engine. Racer currently uses OpenGL and CG (In it's latest guise) and this is COMPLETELY different to basically all games out there. Taking a car from say, GTA and converting it to Racer will end up with ****, however you polish it, unless you spend a lot of time on the car, that time you invest should be equal IMO to creating a car from scratch, otherwise you're releasing a half-assed effort that will inevitably be ****.

I can't speak for all artists but it really angers me when you spend so much time and effort modelling a car, taking note of the nuances of the car body, the shapes and flow of the car body and ending up with a car mesh you're really happy with and proud of only to have some guy post a 2 hour or 2 day conversion which ends up **** but gets all the attention.
Artists put hours of care and attention into their meshes and when your work ends up getting shunned for some ****ty conversion, it really disheartens you and makes you think twice about releasing content again.

Now, I've never lost the urge to release content, and I am extremely proud of my work, this is why I cannot understand why people are still releasing content and planning content to be sub-par. Why would you release a car or track not wanting it to look the best it can?
This is where the CG vs. Non-CG debate comes in.
CG is clearly the way forward. I can within 5 minutes, load up a website and order a graphics card capable of running CG for less than 50AUD.
In a week I spend more on petrol than that. I cannot fathom how 70 or 80% of the Racer community can't spare $50 to buy a graphics card.
Bear in mind that with this you could play all the new games coming out, you might need to upgrade your CPU in the end but it's a good starting point for a new PC.

I've been lurking around these forums since the move. I was at RSC for a few years before the shutdown and I believe I'm one of the longest members of the Racer community.
In the time I've been around I've learnt a lot about cars and modelling, physics and have since started my career in 3d art.
My career is important to me because I love what I do. If I'm not content with what I'm producing I'll dump it all and start again because I want an end product I'm proud of.

It really angers me when the community is so ungrateful of all the time and effort Ruud puts into a game you get for FREE that's one of the best platforms you'll ever experience.

No, it's not Ruud's fault ATi cards are **** with OGL, it's ATi's fault, stop blaming him.
On that note, stop complaining that you don't have a Non-CG edition of the latest version of Racer to play with. If you want to keep up with the development and progress of this great simulator fork out $50 and buy yourself a new GPU, you'll thank yourself for it.

I'm making no apologies for any of my opinions here, I stand by everything I say and believe that Racer has so much potential that's yet to be recognised.

Ruud is trying to move the game forward. The community is holding Racer back.

 

[Knut Omdal Tveito]

And perhaps we should have a prefix for converts?

That would work:good: I need to learn more about modeling and this would be a neat way of identifying the scratch-made threads

 

[Cobra]

Looks like I missed one hell of a party!
All of this is over CG vs. non CG? Well, I know there is more to it.....lots more! I love Racer and appreciate all of the cars/tracks on here. I may not be able to enjoy them (CG stuff) yet, but one day.
Maybe designating the converts and the scratch-made ones would be a good idea. But, we would have to do the same for the converted and scratch made tracks. (It may have been already mentioned-I only read the latest post before my 'reply').

 

[Ruud]

Although I haven't read all, I think it's hard to keep quality the same everywhere. Conversion can be ok, but Racer is quite complex these days. That's what you get when trying to play catch up with GT5.

One of the reasons I had for creating racerdownloads.com is to have more control when v0.9 comes out. I'll add a compatibility field so that you can skip non-0.9 content. Really for the best experience you should add Cg shaders, since they can make material look a LOT better.
But the community is big, quality will vary. Popular cars will automatically tend to float up though. I respect the 'LEGO' type of attraction has for a lot of users; just playing around with Racer and your 3D models can be fun. I do target at high-end graphics quality though.

As for non-Cg, v0.8.10 will have a dedicated racer_nocg.exe which loads track_nocg.shd and car_nocg.shd instead of the regular ones. Lowering the threshold for people to have a look at Racer is still important. If they like it, they'll be more inclined to go through some effort to get the Cg version working.

In the end, the most beauty is always in the detail. Cars specifically designed for Racer are easier to maintain, than starting from a converted model. You can go very deep & technical in Racer, that's not something that all people want to do though. I still aim for those who want a free and open environment for creating content that matches those of the high-end games though.

 

[Some1]

Well, I speak only for myself, but I feel that content shouldn't really be created for Racer BETA versions. Things tend to change quite a lot sometimes between betas and something created for previous version won't work correctly in the next. And always keeping your past work up to date will mean less time for new work to be created!

While, having backwards compatiblity using "_nocg.shd" is great as an idea, it adds more work for the content creator.

Content specific Cg shaders might add some flexibility and playroom for content creators but also make the content Racer version dependent. I really like the rFactor approach, where you just select a suitable system shader for your car body or windows material, specify the textures to be used and be done with it.

So, I think that until a new Racer final comes out people should not create, update and convert too much stuff. Until Racer has standardized and correctly renamed, documented and cleaned up its system core shaders and other stuff, I, for one, won't be wasting time creating cars and tracks from my own free time for Racer.

 

[AMGfan(BPM)]

Well, I speak only for myself, but I feel that content shouldn't really be created for Racer BETA versions. Things tend to change quite a lot sometimes between betas and something created for previous version won't work correctly in the next. And always keeping your past work up to date will mean less time for new work to be created!

While, having backwards compatiblity using "_nocg.shd" is great as an idea, it adds more work for the content creator.

Content specific Cg shaders might add some flexibility and playroom for content creators but also make the content Racer version dependent. I really like the rFactor approach, where you just select a suitable system shader for your car body or windows material, specify the textures to be used and be done with it.

So, I think that until a new Racer final comes out people should not create, update and convert too much stuff. Until Racer has standardized and correctly renamed, documented and cleaned up its system core shaders and other stuff, I, for one, won't be wasting time creating cars and tracks from my own free time for Racer.

Indeed, but if new "final" racer wont be fine as we wish, I will still using my good 088 cause for my deals it is perfect. (as ruud has no plan to do traffic, blinkers, etc.. well, my deal is a simulator of real car, not a trackday car, or something like this. yes, sometimes is nice to run in some circuit, but i prefer 90% of time to do free ride at open roads.)

 

[QuadCoreMax]

I apologise in advance because this thread is not in a light mood and may anger/upset some people, but this is my view, feel free to post yours.

Recently Racer has come leaps and bounds and now rivals the graphics of GT5 and IMHO can be better. So it really irritates me when I see all the crap that is being released these days. I'm not going to deny it, this post really rubbed me the wrong way...
Ruud is trying to move the game forward. The community is holding Racer back.

I agree with you, I never saw so much crap from modders. I've been testing from http://www.xtremeracers.info some dozen cars & tracks, & well I'm disappointed.

I'm new here, quite famous as a Shift Modder currently & when I do/release something, believe me, my stuff is completely debugged, almost perfect, like Racer generally speaking...

Been learning from the best worldwide engine : Crytek Engine (SB2) & been doing 3D (almost all sectors : modeling, animating, rendering, dynamics, custom scripting...) for 5 years now.

I'm really happy to work/learn/develop for Racer, I hope I'll bring some decent tracks / cars / new ideas/workarounds to the community. I love this project from Ruud, real top application & tools available !

I'm working on a surrounding system powered by 2 ATI cards, 3 LCD 19" in DVI-D & a G25 coupled with a G15 on a quadcore Phenom BE PC. Racer 0.8.10 runs perfectly !

Thx Ruud & Steven...

 

[luthobu]

A long time ago, there was this 3D game, and it was called DooM. It was at the time
one of the best games I had ever played (and I was a die-hard D&D gamer then).

As time went by, addons, wads and total conversions were being made, some really
good ones - like Ravenloft, others not so good at all, and some really "crappy" ones.

Now, the thing is, that no matter how much "crappy" stuff people made for DooM/II
the game itself still remained a hit for a very long time. Why? Judging by many of
the comments here, one would think that the "crappy" content would put down DooM
and render it silly and useless. But that never happened, in fact, quite the opposite
happened, and DooM is still very popular, its engine has been rewritten, improved
and extended.

Again, why did not the "crappy" wads have any effect on DooM? My theory is that
since DooM was a /release/ - i.e. complete with engine and content - no matter
how much "crappy" stuff people made, the /real/ game itself had nothing to do with
it, it was community made, by and for the community.

My point? I think that if Racer was a complete release, engine and content, then
the community stuff would be just that, community stuff. And since Ruud on his
own pages state that "Cars, tracks and such can be created relatively easy", that is
bound to happen, people will make tracks and cars with varying results in quality.

It seems as if Racer is moving towards an official release, and that is really good!
But, if Racer is to be kept "separate" as a complete distribution, the released content
has to be good, varied and enough to keep a new user occupied long enough to want
to try the community stuff.

My point here being, it is silly (imnsho) to blame users who willingly and on there own
spare time create content for Racer for being the ones holding Racer /back/ - for this
is in no way true. What is in fact holding Racer back is the lack of an official release
with decent content. And I know Ruud is working towards that goal now, and that
hopefully soon he will release the long-awaited, breathtaking and utterly awesome
Racer v.1.0

Sorry if I offended anyone. As always, it is not my intention.

 

[Mr Whippy]

Doom looked "relatively" crappy to start with though, so crap stuff didn't really stand out. I remember making map tweaks for CounterStrike/HalfLife and replacement sprites/anims for Red Alert in a few hours...

A single person could recreate the artwork in Doom themselves in a few weeks. Today, for a high-end game, it takes a few weeks just to make a good quality mesh for a car!

Just look at GT5 screenies. Racer is capable of that today. If you want to put NFS4 'mod' level stuff into Racer that is fine, but it really isn't making the most of Racer.


To be honest with you, a person downloading Racer today gets a fantastic base. What they then download to use in it is up to them...


Racer has NEVER had decent in-built content. It is up to YOU the user to make it... the level you choose is up to you but there will be a pressure to up your game when good content starts to be delivered for a fixed version in the future again!

I don't see what is so bad about wanting to make good content. I for one think it's fantastic we have a GT5 level game wrt visuals that we can develop for in a nice open friendly environment, so why limit your own imagination and development of skills?

Dave

 

[Stereo]

I don't see what is so bad about wanting to make good content. I for one think it's fantastic we have a GT5 level game wrt visuals that we can develop for in a nice open friendly environment, so why limit your own imagination and development of skills?

Dave

Well, PS3 is a relatively stable platform to develop stuff on, compared to the huge variety of hardware people use for Racer.

Right now my setup is happy to play Carlswood/Murcielago at a fair framerate with all the graphics settings turned up, and my Aronde with around 15k polies is fine, but throw in higher poly content or file sizes and it bogs down a fair amount (for example, ~5 fps with AMGFan's C-10 pickup, I don't know an exact polycount but I'd guess it's around 60-70k)

The Aronde is of course relatively low-poly to the stuff in GT5 and so on, but that's because my computer is happier with the smaller amount so that's the way I build content.

 

[luthobu]

I don't see what is so bad about wanting to make good content.

Neither do I, in fact I encourage it. That was not my point. I just thought that maybe
this way of thinking might put an end to the many "quarrels" on the this's and that's.

I still think an official Racer distro, with good content (more than just one track and
two cars), and optimized settings might make the needed distinction between "official"
content and "community" content.

I do my things.. I do it for fun, and if people like it, cool, if not.. cool too. At least /I/ am
having heaps of fun, both with making content, fixing content, and driving..

 

[Some1]

I still think an official Racer distro, with good content (more than just one track and
two cars), and optimized settings might make the needed distinction between "official"
content and "community" content.

And if not "official", when can always create a community distro

 

[luthobu]

And if not "official", when can always create a community distro

That is also a great idea, and of course your tracks would be on the top of /my/ list for sure

 

[Cosmo°]

Stereo: Those are two different things though, trying to make the most out of the development platform and going for sheer numbers in high polygon counts and texture resolutions. There's nothing advanced about a 100k polygon mesh, or anything that makes it inherently superior to a 20k polygon mesh. Using 2048² textures with a single detail in the centre and empty space around it doesn't improve it over the same thing on a 128² area on a tightly mapped 1024² texture. The idea is to push yourself and not (just) your hardware and I'm sure that's what you mean as well.

 

[Mr Whippy]

There is currently a massive amount of wastage on some cars.

Why people can't draw inspiration from the 2005 ish vintage high-end Racer cars... there really are some fantastic looking ones out there.

Bumpers DB9 and Some1's C6 are perfect examples of mid-poly count by fantastic visuals through just high quality authoring!


Of course, pushing the boundaries is good, but not just making poly counts really high. After about 10m you can't tell anyway. GT5 and FM3 for example use only the high-quality sub-d meshes in photomode. In-game they quickly go to a relatively modest mesh level!

Dave

 

[Some1]

Well, after I created the Supra for rFactor, I'm always going to add 3D bodylines, because I want to model my cars in parts (doors, bumpers hood etc, not just one solid body mesh). I can't remember exactly, but the Supra has around 20k - 25k triangles, and rFactor didn't have much of a problem running 10-15 cars at the same time.

Modern GFX cards have quite a lot of polygon pushing power. :good: ... which doesn't mean you can go on wasting polies all over the place.

 

[Mr Whippy]

Yep, right now in Racer with one car and screenshot taking, mega polies is ok.

BUT, when we have nice tracks with big view distances suddenly polygons get costly again and saving them is important.

Being efficient is ALWAYS good practice.

BUT, I plan to model polygon panel edges too... just which method. I like the idea of a nurms 'safe' edge with a smooth group over that edge, which makes a nice smooth normal blend at the edge, but is relatively light, AND you can just bump meshsmooth up a level for a super-hq model for photo mode style stuff (super close LOD model!?)

Hmmm

Dave

 

[Some1]

I made my bodylines the old way - just chamfering.

Of course, LOD's are important, and 3D bodylined model should be used only when camera is close to the car.