I expect about 20 cars at most on the grid of a NASCAR race. Because all the cars are modeled quite similarly, and the ovals they race on are simple and not to detailed, I think PD will be able to squeeze an extra 4 cars into the grid. (However, NASCAR does have a couple of road courses on their schedule so it may not work for those courses).
There is no reason to expect more than 16 cars, Kaz has only ever said upto 16 cars per race. Besides, forget about the polygoe, think of the physics, the game has to calculate the physics for each of the cars as well. That said, even thought the pexterior of a Nascar is plaininer and probably less polgon intensive than the average Ferrari, we know that Nascars will be among the cars that have body panels coming off and geting more heavilly damaged. Modelling the cars structure and details under the bodywork = more polygons. With all things tconsidered I'd guess that the Nascars in GT5 will probably use more polygons than the Ferrari 458.
....I'm sorry, but wtf?
The Baja 1000 is a real race with countless vehicles, doesn't mean I think GT5 could do it. That's the dumbest idea I've ever heard in someone trying to justify their claim.
Nurburgring = 13 miles long with lots more detail than an oval course. Fact.
Stock cars are like GT cars (in appearance) but with less variety. Fact.
Less detail in both cars and track = more memory freed up. Fact.
What to do with that extra memory? Hmmm...... More cars?
Like I said, it's basic modeling. If you aren't achieving what you want due to too much resources being hogged by graphical assets, what do developers do? They cut corners that normally wouldn't be noticed as missing. FACT.
Every developer does it. If you don't think PD does that with their games, you're mistaken.
No, his point is valid. The biggest argument supporting the idea (which has absolutely no base to stand on) that Nascar races in GT5 will have larger grids is simply that Nascar in real life has larger grids and thus will need it in the game to be accurately represented. His post simply shows the datness of that line of reasoning. The next biggest argument I hear is about the number of polygons a /Nascar will use being lower which is a much more valid argument on the surface but see my bove reply to sam. If anything Nascars in GT5 will use more polygons than the regular road cars and even if not the PS3 still has to calculate the physics for each car as well. Besides, even if a Nascar did use less polygons than the average road model, the difference probably won't be so huge as to allow extra models to be thrown in without compromising the stability of the game. GT5
already struggles at 16 cars, obviousely GT5 is going to be more optimised but that will be to get 16 cars running as smooth as they can.
All those cars were clearly Chevy, it doesn't mean they'll only have Chevy. What they've shown so far is an unfinished product.
And? We've still been told that GT5 will feature up to 16 cars in a race. Not up to 20, not up to 25, not up to 30, up to 16.
I understand what you mean, and what everyone else means. Hardware can only take so much, not to mention with the polish that PD puts, but they are completely capable of hitting at least 20-25 cars. Just cut some corners...
Such as?
I think it was explained clearly why the same model of the same car repeated over and over on the screen isn't as hardware intensive as the same amount of different models. Nothing hardware defying. Nascar cars have very basic shapes. I don't see why they would make them the same number of polygons as more intricate vehicles. I don't believe every car in G5 is 400k polygons, I think that's the highest number they used on a car. I don't see how a daihatsu copen would require as many polygons as an Enzo or Murcielago.
You're rightif you are simply modelling the Nascars exterior and interior, it's probably half to two thirds of something like the 458 Italia, but since we know body work will come off Nascars and they will suffer more extensive damage they will have more underbody details modelled which if anything could make them a more complex model than the 458 Italia which is unlikely have anything modelled beyond what we've already seen.
Firstly, the game treats each car differently on track, and has to process and calculate differently for each one. Using many of the same models in a race only cuts down on the race loading time, it has no difference on the processing power used during a race.
They are high polygon and high detailed models, as with all cars in GT. A lesser poly count for a simpler model might make sense but it is probably not the case. Besides, Stock cars are rounded, the polygons need to be of a sufficient detail to show the rounded edges etc. The panels are not completely flat. They are also likely to be 'premium' models, which most likely means there is infact more detail in them than roadcars, having to account for the damage model, they have to model the parts underneath the panels etc. so that when parts fall off you are looking at accurate representation of the damage.
As for the size of the car, sure a Copen would use less polys than a big Murcielago, simply because it is smaller. So the same level of detail equals less polygons. But if i'm very much mistaken, Stock cars are quite large.
Most of this is correct except your last comment about the size of the model defining the number of polygons. While the size does play a part, you can quite easilly make a larger but simpler object out of fewer polygons than a more complex smaller one. I get what you're saying but the principle I got from that last comment seemed wrong. Other than that, you're post is spot on, having 16 of the same model in a race will not on it's own ease the graphical or technical burden of processing that race 👍.
, although I do hope GT5 is released before reaching 2,000,000......
( there are some hints on my page for anyone wanting to take a guess ).
)
