• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Project CARS | OT | Made by 80.000 slightly mad developers

Mascot

Member
Not since I got AC, the physics, in most cars, are just such a turn off now I've played AC.

I liked AC's physics but I don't dislike those in pCARS enough to sacrifice dynamic weather, dynamic time of day, good AI, massive grids, loads of tracks, the race weekend experience, telemetry apps, Crew Chief... oh, and fucking WHEEL SUPPORT.
 

danowat

Banned
I liked AC's physics but I don't dislike those in pCARS enough to sacrifice dynamic weather, dynamic time of day, good AI, massive grids, loads of tracks, the race weekend experience, telemetry apps, Crew Chief... oh, and fucking WHEEL SUPPORT.

Yeah, I like a lot of hotlapping, so those things, while good, don't really bother me that much, physics are the most important thing for me.

Hopefully PC2 will improve on this.
 

Mascot

Member
From the PS Pro Beast Mode thread:

I'm going to fire up Project Cars now on my Youtube Stream to see how it is (I'll enable 60FPS mode). If you want to watch it's youtube.com/padinn813.

I have played this game like twice, so prepare for terrible.

For anyone who missed this it should be up on Padinn's channel later. It's well worth a watch. Thanks Padinn, and for handling all my requests.

From what I saw the game looks locked to 60fps with zero tearing in daylight, 36 cars on track.

I then asked Padinn to run a night race at Monza in a thunderstorm, again 36 cars. It looked amazing. It's harder to tell with all the spray but performance was superb, orders of magnitude above vanilla PS4. Not a sixty lock in the storm, but bloody good. There was minor tearing in this race but it was very slight and you had to be really looking for it. Monza's first chicane behind a full pack is a real stress test even on a clear day, but it was pretty smooth here at night, in a storm.

Very, very impressed indeed.

Thanks again Padinn. Very much appreciated.

Edit: the full video is here:

https://www.youtube.com/watch?v=cMIAYEld8oA
 
I then asked Padinn to run a night race at Monza in a thunderstorm, again 36 cars. It looked amazing. It's harder to tell with all the spray but performance was superb, orders of magnitude above vanilla PS4.
Normally the number of opponents is more CPU stressing and look at benchmarks how CPU bound Project CARS is...
CPU_01.png

below a 3.2GHz Core i5, you're really wasting your GPU potential.
only slightly over 20% (edit: ok, theoretically it could be 33%) of Steam users are there yet (from Jan 2017):
cpuspeedsteamf0u0s.png




that it's better on a Pro with many opponents is weird.
 
Closed platform development?

I don't know...
the opponents are still on the same tire model and AI as in the PC version, the physics engine is at the same framerate... you can't really do those calculations on a GPU and there aren't any special features that the PS4 Pro CPU has you could optimize for.

Sunny weather (with Crepuscular Rays turned on) might be more stressing than a thunderstorm, that's what I remember from pCARS on the XB1. Could that be the thing? Maybe the original PS4 handles the opponents just as fine/smooth in the thunderstorm weather setting.

I mean, I can believe that it runs better, but numbers would be interesting.
 

Mascot

Member
I wouldn't have expected a big improvement with the pro either because of how CPU bound it is on PC. That's interesting.

Don't forget, vanilla PS4 performance post-patches was very impressive already. This Pro Beast Mode beta seems to do that little bit extra needed to eliminate the minor frame drops and tearing in the dry, and boost performance in the wet to a more stable framerate (not 60fps at night in a thunderstorm, but much better).

In the dry the game seems to run faultlessly now, even with 36 cars.
 
Mascot, you're right, weather-condition is a big difference and clear skies give a much better performance.

In Sunny weather, most frametimes are about 8ms (=120fps), in a thunderstorm I get framedrops (everything over 16.7ms).
pcars1dkuic.png


My GPU performance is a tiny little bit better than a PS4 Pro, but my CPU is way better... still framedrops, but it did look smooth to me.

To be really scientific I should do the same with my CPU downclocked, but I'm not gonna do that.


...and I didn't have fun driving either, trail-braking in pCARS is garbage even in GT3 cars with ABS on. I spun out in the AMG and the RUF. Tried brake bias all the way to the front to get understeer... nope, spinning. This game looks fantastic and runs great on my PC now, but the physics... man!

Is there an App that saves the telemetry and lets you export it to MoTec Pro or even just an excel table or so?
 

Mascot

Member
Yes it looks good, not new console purchase good, but decent. Usually i run with 16-20 cars, framerate has been pretty good in the base PS4. https://www.youtube.com/watch?v=VaL7_H06xv0

I've been thinking of upgrading for a while in prep for GTS so this was just the nudge I needed. As you say, vanilla PS4 performance is pretty good and frame drops and tearing are minor in dry weather with sensible grids. Beast mode on PSPro just looks effortless though. Seemingly rock-solid. And the increase in wet weather performance might make me actually use those clouds for once.
 

VVV Mars VG

Member
I have to admit I've not been too impressed with anything on the PS4 Pro yet, for PCars 2 I'm more interested in the Scorpio at present. I hope the new more powerful consoles don't overshadow work on the base systems.
 

Mascot

Member
I have to admit I've not been too impressed with anything on the PS4 Pro yet, for PCars 2 I'm more interested in the Scorpio at present. I hope the new more powerful consoles don't overshadow work on the base systems.

I really only looking for that little boost that makes annoying dips and tears a thing of memory. I'll still be in 1080p. It's a shame devs feel they need to stress the hardware so much that performance is compromised. PS Pro seems to take up that slack.
 
Is there an App that saves the telemetry and lets you export it to MoTec Pro or even just an excel table or so?
Yes there is,"pCARS Telemetry".

I gave it a spin and exported data from my Assetto Corsa and Project CARS with the BMW 1-Series M Coupe @Brands Hatch GP which is laserscanned in both games.

Filled both cars with 30 liters of fuel and did 12 laps each(went out of fuel on lap 13 in pCARS, AC's fuel would have laste ~15-16 laps).
The ambient temperature was 24°C in both games and track temperature 32°C in AC, 30-29.3°C in pCARS.

ABS set to ON in both games, Traction Control and Stability Management OFF.

The tire used in pCARS was the "medium street" tire (slower than the "track" tire), which I wanted to set against AC's "Street" tire, but after I was done lapping in pCARS and did my first AC lap, my lap time was SO much slower, that I picked the faster "Semi-Slick" tire.
I was still a lot slower, so I did one last thing and set the track surface condition in AC to "optimum", even though it's said to be unrealistic and the "fast"-track condition gives more realistic results and more natural driving feel.

So, here is the data from my best laps each:

in BLUE: Project CARS BMW 1-Series M Coupe @Brands Hatch GP [lap #12, lap time: 1:38.464]
in ORANGE: Assetto Corsa BMW 1-Series M Coupe @Brands Hatch GP[lap #9, lap time 1:45.399]
(the x-axis is "distanace in m" for comparison's sake, it's not time in seconds!)

What I see in that data:
- pCARS accelerates a little bit faster (longitudinal g graph)
- pCARS has less drag (speed graph: slope decreases earlier in AC)
- both games brake similarly well, pCARS a little better (longitudinal g graph)
- the tires in pCARS can take more lateral g's (lateral g graph)
- AC has more understeer (front tire slip angle graphs: higher spikes)

- not directly in the data, but I could just put the power down more early on pCARS, the more rear tire wheel spin (in the RL rps and RR rps graphs and the RL & RR Slip graphs) didn't slow down and didn't send me spinning from backbite/snap after countersteering. Which ...this less snap should be accurate for "Street" tires, compared to "Semi Slicks" in AC, but that the street tire then also grips better is making it unrealistic again.


In which game did I have more fun? In AC I think, but I was much more consistent in pCARS, which probably makes for more fun races if the AI plays along.

Which time is more realistic? Well, I don't have a BMW road car lap time for Brands Hatch GP, but in the "Brands Hatch 2 Hours 2016", an Aston Martin Vantage GT4 (slick tires) qualified with a 1:39.6 and did a best-lap-overall time of 1:38.94.
The pCARS time is almost as fast as my best BMW M4 Coupe (430hp compared to 340hp in the 1-series M Coupe) time in Gran Turismo 6, which is usually said to be faster and more forgiving than in the real world. pCARS definitely feels like a sim, but this car is just much too fast.

Why did I not pick a GT car? Didn't want people saying that it's the setup, especially with different aero the results could be extremely different. A production car should have been a level playing field.
 
I love that I have telemetry apps for almost all racing games now. Can directly point to the bullshit



Ford Escort @Zolder.
Mostly over the last year with more PC sim experience, I've trained myself to automatically let off the brake a little, start turning in ...and then leave off the brake more with more turn-in to make the most of the friction circle freedom and the higher front tire-load. You know, just like this "string from foot to steering wheel" technique illustrates.

Doesn't work in pCARS. As I've described before with the AMG GT3, there I got oversteer when turning and still braking, here with the Escort it's the other way round:

I get understeer while still braking, like the brakes are stabilizing the car immensely (usually you'd expect more weight on the front should let you turn in more or even too much) ...and mind you all without the tires locking up at any time(!) - you can see that in the two graphs on the bottom, front and rear tires keep rotating during the whole maneuver [don't know why it's negative rotation, probably some x,y,z-axis choice by the devs and doesn't matter]).
Look at the yaw-angle graph, I turn in with still some brake-power on, yaw rate almost doesn't care, understeer, then with the foot off the brake entirely, BOOM: yaw-angle explosion, sudden oversteer.

The braking physics in this game are completely beyond me.
 
So I just started playing this game and slowly going through the career mode. When I first saw the Monaco race, I was like coool, I am looking forward to this. I am playing in the highest division and around the start, the cars are scraping the sides and then the first two curves the cars get stuck and block everything. Can't believe the developers jut left it like that. So buggy!!
 

ManeKast

Member
Cheap as chips on current psn sale. The dlc is on sale too. Couldn't help it and picked it up. Hopefully its a serious racer I can sink into.
 
So... Now it's been a while since I've played a decent racing game, I've decided to spend a good chunk of the long weekend playing this again.

How has it held up and anything I need to know update wise as a controller user?
 
So... Now it's been a while since I've played a decent racing game, I've decided to spend a good chunk of the long weekend playing this again.

How has it held up and anything I need to know update wise as a controller user?
You need to know that, as a controller user, it still sucks.
 

danowat

Banned
So... Now it's been a while since I've played a decent racing game, I've decided to spend a good chunk of the long weekend playing this again.

How has it held up and anything I need to know update wise as a controller user?

Use these

Steering deadzone 6
Steering sensitivity 0
Throttle deadzone 10
Throttle sensitivity 30
Brake deadzone 10
Brake sensitivity 30
Speed sensitivity 55
Controller filtering 40
Advanced on
Soft steering damper on

Use real life assists
 
So I've been playing this game quite a bit on the Xbox. I always come back to it after a few weeks and forget if I had changed my controller settings for the kart racing segments or not. I found that whatever settings I had made the steering very slow on the tighter corners for the karting tracks. I started that campaign from the bottom but it has no problems throwing me invitations for other disciplines in between kart races but that just adds more issues.

Do you guys normally feel the need to alter control setups depending on the series you are doing?
 

Md Ray

Member
It'd be really helpful if someone can tell PS4 equivalent settings in pCARS for the following visual settings:

Reflection
Track Detail
Enhanced Mirror
Motion Blur
Particle Level

I know most other settings like for e.g. AF is equivalent to 4x, AA is EQAA and Shadow quality is between high-med, Environmental Map is equivalent to PC's Low setting etc. It's just the aforementioned settings I'm not sure of. I read DF's article but it's still unclear.

Thanks.
 

Gestault

Member
I just grabbed the new 387.92 Nvidia drivers and I'm killing time during a big download, so I decided to run some single-race events while I wait. My wheel is connected, but not plugged into the power strip, so I decide to just use the XB1 controller on my PC. I've barely played the PC version with anything but the wheel.

I don't have words for how poor the controller implementation still is in the up-to-date build. What a terrible experience. It's worse than I had remembered, and it's frustrating even with an AWD car at moderate speed. I don't want to slip into hyperbole, but this is unplayable. I was under the impression this had been adjusted over time, and I had a passable experience on the PS4 and XB1 versions some time ago with a handful of different tweaks for different classes of car. I literally don't understand how SMS managed to get standard controls so fundementally bad.
 
I just grabbed the new 387.92 Nvidia drivers and I'm killing time during a big download, so I decided to run some single-race events while I wait. My wheel is connected, but not plugged into the power strip, so I decide to just use the XB1 controller on my PC. I've barely played the PC version with anything but the wheel.

I don't have words for how poor the controller implementation still is in the up-to-date build. What a terrible experience. It's worse than I had remembered, and it's frustrating even with an AWD car at moderate speed. I don't want to slip into hyperbole, but this is unplayable. I was under the impression this had been adjusted over time, and I had a passable experience on the PS4 and XB1 versions some time ago with a handful of different tweaks for different classes of car. I literally don't understand how SMS managed to get standard controls so fundementally bad.
I don't want to defend the default setup, not even pCARS1 physics and tire model, but you have to acknowledge that the approach was to make a car simulation first - unlike GT and Forza who make a game first. If you move the steering wheel of a race car on hard springs, set up for being fast with balance and a slight tendency for oversteer, really quickly then a LOT will happen, if you move it quickly in a really soft car almost nothing will happen.

I always loved what Dan Greenawalt said, that at Turn 10 the physics simulation always comes first, then they build driving and controller aids on top of that to make it enjoyable for everyone. Problem is, it's not true and no game/sim has nailed this concept so far.
 

Gestault

Member
I don't want to defend the default setup, not even pCARS1 physics and tire model, but you have to acknowledge that the approach was to make a car simulation first - unlike GT and Forza who make a game first. If you move the steering wheel of a race car on hard springs, set up for being fast with balance and a slight tendency for oversteer, really quickly then a LOT will happen, if you move it quickly in a really soft car almost nothing will happen.

I always loved what Dan Greenawalt said, that at Turn 10 the physics simulation always comes first, then they build driving and controller aids on top of that to make it enjoyable for everyone. Problem is, it's not true and no game/sim has nailed this concept so far.

I know from (years of) experience that you can dial something in that works with one or two different styles of car (I ended up doing some enjoyable laps in a Subaru last night), it's just eye-popping when you play the game "normally" and assume it will even resemble a well-made experience. This isn't poo-pooing the wheel implementation at all, that's been fine, but I wonder if any serious testing was ever put in on controller support.
 

kraspkibble

Permabanned.
i finally bought this game a couple weeks ago for just £6. i was going to get Forza 7 this year but bullet dodged there and i'm not interested in GTS. the reason i left it this long to play Project Cars is because it's more of a simulation game and everyone says how awful it is with a controller.

i wish i bought it sooner because i'm having a load of fun with it even playing with a 360 controller. this is the first racing game i've used manual transmission and i'm using less assists.

the only thing i'm not happy about is the variety of cars. i find myself using the same 2-3 cars all the time. i'll play this for a while and then pick up PC2.
 

Gestault

Member
i finally bought this game a couple weeks ago for just £6. i was going to get Forza 7 this year but bullet dodged there and i'm not interested in GTS. the reason i left it this long to play Project Cars is because it's more of a simulation game and everyone says how awful it is with a controller.

i wish i bought it sooner because i'm having a load of fun with it even playing with a 360 controller. this is the first racing game i've used manual transmission and i'm using less assists.

the only thing i'm not happy about is the variety of cars. i find myself using the same 2-3 cars all the time. i'll play this for a while and then pick up PC2.

Yeah, the car selection is a bit thin, but the track selection still stands as excellent, even compared to newer releases from this year. Wheel or no, the fewer assists you end up using (assuming you've got a feel for it), the more nuance you feel in the handling. Even silly stuff like the dusty braking sound on some cars is handled beautifully, and gives an extra layer of feedback between you and the game.
 

Mascot

Member
I don't have words for how poor the controller implementation still is in the up-to-date build. What a terrible experience. It's worse than I had remembered, and it's frustrating even with an AWD car at moderate speed. I don't want to slip into hyperbole, but this is unplayable. I was under the impression this had been adjusted over time, and I had a passable experience on the PS4 and XB1 versions some time ago with a handful of different tweaks for different classes of car. I literally don't understand how SMS managed to get standard controls so fundementally bad.

Oh, it was an utter fucking disgrace at launch. I tried the pad for a painfully short while and found it literally pointless, like the car was a puppet that i was trying to manipulate with loose elastic bands.

I know a whole bunch of people that noped out in one day, never to return. And they are car game nuts who'd been looking forward to the game and following development for several years.
 
You know, in Assetto Corsa where the assists are much closer to reality, I'd disagree(except fir STM in AC, which is not a real simulated aid in that game).
About controller testing: I think some people in the WMD community actively wanted the controller to be impossible to drive with zero accomodations in steering aids and as close to 1:1 shittyness as possible. Because on a controller it should not be possible to be faster than a wheel because "sim" and "realism".

Blly155, can I ask you, what your 2-3 favorite cars are in pCARS1?
 

Gestault

Member
About controller testing: I think some people in the WMD community actively wanted the controller to be impossible to drive with zero accomodations in steering aids and as close to 1:1 shittyness as possible. Because on a controller it should not be possible to be faster than a wheel because "sim" and "realism".

This possibility came to mind when I started it up with default controller settings, but I didn't want to assume. That anecdote sure feels believable.
 
Top Bottom