Project CARS 3 Launches to Mixed Reviews

2. They simplified (by a large margin) the underlying physics, and removing tire wear and damage they should have reduced a lot the computational power required.

They made it easier to drive, but that doesn’t mean that the physics model is simpler and easier to calculate.

It’s also worth keeping in mind that graphics is generally handled by the GPU, while physics is handled by the CPU, so it’s not that easy to transfer resources between physics and graphics.
 
I don't know if it's Bandai Namco or SMS who made PCARS 3 what it is title and tone wise, but whoever signed off on it being the next mainline title will probably be losing their jobs. This is a title that very much should be a spinoff, therefore minimizing the cross-stream between the categories, and not muddying up the brand that you've built PCARS up to be. That wasn't done, and now the brand is at a crossroads where it doesn't need to be.

PC3 was the last standalone title from SMS/BN. New/next titles will be under the Codemasters label. My assumption is that Codemasters gives **** all about the Project Cars brand and was more interested in the simulation/livetrack/devs and will integrate/leverage all of those in a new or existing title. So no crossroads, but a dead end.
 
I still think this was rushed out to make to best of the large user base that is PS4, before the PS5 launches. Lets face it, your not going to launch a new game on PS4 AFTER the PS5 is out, and theres no point in releasing a new game 2 months before a new console. So its now or never.
SO.... rush out PC3, dumb it down because theres no time. And work on PC4 as a proper sim for PS5 in two years time.
There is no sence in not keeping PC3 a true successor to PC2, its crazy.
Its like making Far Cry 7 a platform game with cute fluffy bears, you WOULDNT call it far cry 7.
Thats what i think has happened, its a panic rush job, to cash in.

I second this fully.

They made it easier to drive, but that doesn’t mean that the physics model is simpler and easier to calculate.

It’s also worth keeping in mind that graphics is generally handled by the GPU, while physics is handled by the CPU, so it’s not that easy to transfer resources between physics and graphics.

You are neglecting the removed tire and fuel wear, and damage.
If you don't have simpler physics without tire wear and fuel wear, well mate I don't know what to tell you.

We enter in the realm of "I neglect the evidence", so I am not replying this further.

In the meanwhile, I progressed a bit on Road D, fully completing the first part.
And yes, this game graphics-wise was probably in need of at least other two/three months of optimization.
I had some races (under full daylight) which were simply amazing to look (Caldwell Park, Donington National and Lydenn). Very detailed, spot on lighting, no jaggies, 60 fps solid.
Then one race in the rain (Catalunya National) which was not well done, stuttering, screen tearing, low resolution, and the rain effects were poorly done.
This game could really shine if they can achieve the same level of quality and detail throughout all tracks/times/weather.
Hopefully they will optimize that.

I am still struggling to get the audio right.
The car sounds feel a bit dumb.
 
Last edited by a moderator:
You are neglecting the removed tire and fuel wear, and damage.
If you don't have simpler physics without tire wear and fuel wear, well mate I don't know what to tell you.
Setting parameters to zero does not automatically mean all the physics code has been removed as well.

Your assumption is wrong. :)
Happy to be proved wrong on that and also *very* interested in what Codemasters is going to do with this. 👍 (Though you probably can't share yet?)
 
I doubt the change in direction was done purely out of time constraints. We knew the direction was shifting as early as 2018.

They probably realized making a full fledged sequel to PC2 on the same hardware was unnecessary, and tried to make a mass market game with PC underpinnings to capture the more casual audience of console generation before they move on to the new consoles. They just didn’t market/title it properly
 
(Though you probably can't share yet?)
If something isn’t public knowledge, it’s not something I should say :)

For what it’s worth I don’t really know anything about future plans, outside of the DiRT Rally series which is my primary role. I’m only hanging out in this forum because I genuinely like playing Project CARS 3 and think it’s good fun.
 
Disclaimer: The statements below are pure speculation from my part.

The state of the game in consoles right now probably has to do with a sequence of events, starting with Codemasters' acquisition of SMS, which was probably discussed during a good portion of 2019 and maybe even before that. I'm fully aware Codemasters has 0 input in PC3's game design choices technical development or marketing, they are just buying an asset, namely the Madness engine and the know-how about it.

At that time, 2 SMS products are on the pipeline (F&F Crossroads and PC3) to be distributed by Bandai Namco, so their resources are divided.
They finally come to an agreement and the acquisition is announced in November of 2019, and that includes setting (in stone) the launch dates of the games to get out of the distribution deal with Bandai Namco. At this time, Covid is just remote news/rumors from China and the world is unaware of what is about to happen this year.

For the first/second quarters of 2020, Covid wrecks havoc on the world and strains the resources of everybody, but the release dates cannot be moved (contractually) so they have to make hard decisions about what to optimize so they can finish BOTH games in time. Since Crossroads comes first, more resources go to that. For PC3 they concentrate in finishing all the game features/content they want to include, and as far as graphics go, they concentrate on PC optimization first.

Summer of 2020: F&F Crossroads releases with little promotion or fanfare and the game receives poor reviews, but it's complete and certainly looks as polished as it can be in the presentation/graphics/sound department.
PC3 manages to launch in the date announced three months before, and is also complete in its game contents/features and functional, but graphics optimization (due to resources/time/covid constrains) suffers in base and even premier consoles (PS4 Pro and One X). In the end, they have met their launch dates and are now free from any more obligations in that respect to Bandai Namco.

What now? They (SMS and Codemasters) must be aware that the no. 1 complaint right now is the graphic presentation in consoles. Will they devote time (and resources) to improve them? The game has suffered already from the controversy about its change in direction and focus, but leaving the game in the state it is in consoles could be a fatal blow to the series (as enjoyable as the game really is) and then Codemasters will have to make some decisions.

End of my crazy morning ramblings. Oh, and I finished Road E series and did the Caterham daily rivals, I'm loving this mess of a game.
 
Last edited:
I’m currently on an ADV motorcycle tour up the eastern seaboard of the US, so I have not had the opportunity to try PC3 on my rig. We did however have an overnight stop last night at a good friends house who is an avid sim racer. (thank you Jesse for putting up with 8 bikers on short notice) I played PC3 for a few hours on PC using his Fanatec DD1 and Heusinkveld sim pedals sprint.

The gameplay itself was entertaining. I like the slow progression using money and XP, the challenges and rivals are fun.
I do not like the driving model the cars feel bland and quite often break traction far too easy. If I could give a terrible analogy, it would be like I just purchased a 2020 Ferrari SF90 Stradale that was delivered with 1960s era bias ply tires on it. I want to purchase when I get home thinking maybe some set up work could help with the oversteer.

Initial impressions are I can clearly see the fun factor with hopes the ffb can be fine tuned with a custom file.
 
Last edited:
You are neglecting the removed tire and fuel wear, and damage.

Not at all. Tire wear and fuel consumption can be done with one line of code each (four lines if you count all four tires of the car). A damage model can be complex, but you rarely have all components failing simultaneously and extremely rarely at a constant rate of 60 failures per second.

And we still have the fact that graphics are handled by the GPU while the physics are handled by the CPU. The GPU is designed to work in parallel and process multiple pixels simultaneously, while the CPU is designed to process data from memory, one item at the time. The CPU is much slower than the GPU when it comes to processing graphics, so the few CPU kiloflops you might save by building a simpler physics engine won't be of much use in the shading computation. In fact it would probably make it worse.
 
Having spent a few hours online + carrer I must say this is not very good. Sometimes it's fun, barely.
In Europe, on PS4, evening or weekend lobbies hardly hit 80 concurrent players
These are roughly the numbers of Grid Autosport on Switch, a digital only game, a polished one.

PC3 looks unfinished, from UI to ingame, there are bugs or missing parts. I guess nobody really QA'ed it.
Looking at the Rivals mode rankings... they probably sold around 4k copies at best
I really wonder what their sales projections looked like given the PS4 userbase
 
And we still have the fact that graphics are handled by the GPU while the physics are handled by the CPU.

I'm afraid its not a fact but a common misconception. The CPU is very involved in graphics. rFactor2 for example runs 2 threads - one for the graphics and one for the physics. And it is intended that most of the CPU power is used for graphics. This is described by the developer Terence Groening

Additionally, it might help you to understand the graphs with the knowledge that the graphics and physics are running on separate threads. The graphics is usually expected to run as fast as possible, while the physics is expected to get its work done in the allocated time. So the fat graphics bar is usually full, meaning it's just using all the CPU it can. But the fat physics bar should be less than full (preferably less than 80% or so), meaning it has completed it's work on time.

https://forum.studio-397.com/index.php?threads/cpu-time-graph-what-does-it-show.28856/#post-471825
 
Having spent a few hours online + carrer I must say this is not very good. Sometimes it's fun, barely.
In Europe, on PS4, evening or weekend lobbies hardly hit 80 concurrent players
These are roughly the numbers of Grid Autosport on Switch, a digital only game, a polished one.

PC3 looks unfinished, from UI to ingame, there are bugs or missing parts. I guess nobody really QA'ed it.
Looking at the Rivals mode rankings... they probably sold around 4k copies at best
I really wonder what their sales projections looked like given the PS4 userbase


It didn't crack the uk top 10 and its nowhere to be seen in the us amazon top 100 best sellers in video games.

While pcars 3 is not the game i wanted i don't wish it sells bad cause it maybe the end of the pcars series and that's not something i want i wanna see pcars 4 and hope that its a successor to pcars 2.
 
I'm afraid its not a fact but a common misconception. The CPU is very involved in graphics. rFactor2 for example runs 2 threads - one for the graphics and one for the physics. And it is intended that most of the CPU power is used for graphics. This is described by the developer Terence Groening

Additionally, it might help you to understand the graphs with the knowledge that the graphics and physics are running on separate threads. The graphics is usually expected to run as fast as possible, while the physics is expected to get its work done in the allocated time. So the fat graphics bar is usually full, meaning it's just using all the CPU it can. But the fat physics bar should be less than full (preferably less than 80% or so), meaning it has completed it's work on time.

https://forum.studio-397.com/index.php?threads/cpu-time-graph-what-does-it-show.28856/#post-471825

That’s not what he is saying though. He explains why it’s okay for the graphics thread to be busy while the physics thread should always have some overhead.

He is also not specifying what the CPU is doing in the graphics thread. Typically the CPU is keeping track of the stuff that should be rendered and telling the GPU to render them. Theoretically it can compute shading as well, it just takes a lot more time than having the GPU doing it.

And it doesn’t change the fact that you can’t remove fuel consumption from the physics engine and expect any noticeable graphics boost in return. That’s simply not how it works.
 
At this point I just hope there is a PC4. Whether it goes back to its roots of 1/2 or stays a simcade. They could do some great things next gen, but this game sounds like a forgettable mess from what I am reading.
 
Digital Foundry's review/comparison is just in:

Ouch...

Sounds like on the upgraded consoles that if “Prioritise Framerate” locked to 1080p instead of trying to go above it, it may get a locked 60. It’s not far off that 60fps when I play on Xbox One X, but the screen tearing is notable on a number of tracks.
 
Setting parameters to zero does not automatically mean all the physics code has been removed as well.


Happy to be proved wrong on that and also *very* interested in what Codemasters is going to do with this. 👍 (Though you probably can't share yet?)

Who said physics have been removed? Where is written?
Guys, really, if you want to keep misreadding words or twisting arbitarly the meaning of a sentence, this is not going to be a discussiom.

Did I say that tire / fuel wear, and damage have been removed. Is this a proven fact? YES. Do we need SMS to confirm this to you guys one more time? I hope not.

Did I say that physics have been removed? NO.

Does tire wear have impact on the physics, as a variable to be calculated and influencing the car physics? YES

Does fuel wear have impact on the physics, as a variable to be calculated and influencing the car physics? YES

Does actual (not visual) damage have impact on the physics, as a variable to be calculated and influencing the car physics? YES

Take out these variables and their effects on the car, and you have simpler physics, which by the way is evident while playing.

End of the story.

And it doesn’t change the fact that you can’t remove fuel consumption from the physics engine and expect any noticeable graphics boost in return. That’s simply not how it works.

Again, please don't twist my words. This is not going to be a discussion if you twist what I have clearly written.

Read my posts above, and then tell me where I have written that I expect graphics in PC3 to be better than PC2.

I have posed a question why in certain instances (nightime / rain conditions) the performance graphics-wise on the same platform is worst (lower resolution, jaggies, probably less AA, lower frame rate, etc etc) than PC2.

Never mentioned about expectations to be better, so maybe you are mixing with someone else's post.

You guys are mixing facts with assumptions.

Fact: removal of tire wear / fuel wear / damage.

Fact: under certain conditions the game looks worst than PC2 (at least on One X).

Assumption: how SMS is allocating the power between CPU and GPU for resolving the physics and graphics loads. Until someone from SMS will come here explaining clearly how they do, this is unknown. Can be assumed based on other titles, but is not 100% guaranteed they are doing the same. So, it remains pure speculation.
 
Last edited:
Who said physics have been removed? Where is written?
You said physics were simpler, which implies a removal of physics code. But if parameters are not removed but set to zero there is no real simplification of the model and as such no performance gains. Weight calculation still has to take place, even if fuel usage is set to zero. Tyre physics still have to be calculated, even with wear set to zero. Damage, we don't know. What we do know: performance is not improved in a major way, it even degraded. So it's pretty safe to assume they didn't simplify anything in a meaningful way.

End of the story.
You don't decide that, especially when not understanding what others are trying to tell you.

You guys are mixing facts with assumptions.
Take a look in the mirror. End of story. /s
 
You said physics were simpler, which implies a removal of physics code. But if parameters are not removed but set to zero there is no real simplification of the model and as such no performance gains. Weight calculation still has to take place, even if fuel usage is set to zero. Tyre physics still have to be calculated, even with wear set to zero. Damage, we don't know. What we do know: performance is not improved in a major way, it even degraded. So it's pretty safe to assume they didn't simplify anything in a meaningful way.

So, you are not even considering the difference between dynamic and static conditions for the physics?

Hands off then.
 
Sounds like on the upgraded consoles that if “Prioritise Framerate” locked to 1080p instead of trying to go above it, it may get a locked 60. It’s not far off that 60fps when I play on Xbox One X, but the screen tearing is notable on a number of tracks.

I also have an X1X and fiddled a bit with the settings, neither is really satisfying. At some times this game can look pretty good, similarly to Project CARS 2. But on other situations it looks absolutely aweful, I clearly remember that Havana circuit race in the wet (Road C career I think) where just before you dive into the tunnel, the view into the distance above that tunnel entry looks absolutely bad, very pixelated. I’ve never seen anything as bad as that because I sit quite far away from my TV (65” 4K HDR).

Over the weekend I’ll try to dig a bit deeper into it, first make sure my settings are priority graphics and make some comparing screenshots.
 
So, you are not even considering the difference between dynamic and static conditions for the physics?

Hands off then.
For someone who claims others are misrepresenting him you do an awful lot of misrepresenting yourself. So yes, let's end this convo. 👍
 
For someone who claims others are misrepresenting him you do an awful lot of misrepresenting yourself. So yes, let's end this convo. 👍

For me is enough that everything stays recorded in the posts above.

All replies, and all questions (even the ones I asked you, and you preferred to go personal instead of replying keeping the conversation relevant to the game).
 
Well I backed the 1st game and made 3 times my money back and hosted most days until pc 2 came out then hosted that every night for over a year yet have zero interest in playing this so far. Its not that its arcade I been playing crew2 since release that's 2 years now I just see nothing new just more of the same if not less :(
 
Digital Foundry's review/comparison is just in:

Ouch...


In short:

PS4>Xbox One
Xbox One X>PS4 Pro

I dunno about ouch, its about on par with PC2. Obviously it has a few of its own issues to sort but like theynsay, it looks like a limit of the engine more than anything.
 
@David Wright you mentioned that PC3 sales were bad, are the stats out for sales? is it too early to tell?

My thanks to Chikane_GTR for giving some new info on thei above.

My initial statement was based on peak steam player numbers in the first few days of launch - 1,218 for PC3 compared to 6,775 for PC2 in the first few days from launch. At just 18% of PC2 numbers, sales are arguably catastrophic rather than simply bad, but obviously console sales are a big factor.

Console sales are not easy to judge at launch. While I appreciate most console sales will be digital, Amazon' do at least provide sales charts and Amazon.uk and Amazon.de indicated PC3 reached the bottom end of the top 20 at launch and have fallen away since (just checked Amazon uk now #87 on PS4 and #46 on XB). By comparison, PC2 reached #2 in the weekly UK sales chart. According to the Metro article, sales are 86% lower than PC2. I'm sure part of this fall reflects the move to digital sales (which are not included in the charts.)

https://metro.co.uk/2020/09/01/anim...new-releases-games-charts-29-august-13205555/
 
Last edited:
Back