Improve Planet Coaster 2 compared to Planet Coaster 1:

[Ramm]

Hello Frontier,

The following suggestions are to make Planet Coaster 2 better compared to Planet Coaster 1. They are suggestions that make the game better for die-hards that like to have full control, but also make the game easier for novice who just doesn't want to spend time placing every single streetlight and tree and so on.

General suggestions:

• Build for performance with optimization in mind so the game can actually be future-proof and run big parks with a lot of rides, coasters and guests. (No more max 2000 guests so you can get over 20fps and stuff like that but have 10k+ guests and 20 coasters and 30 rides and still run at 60fps please.)
• Max parksize at least as big but preferably bigger than Planco1 scenario editor.
• Steam Workshop support
• Mod support (Also steam workshop)
• Everything that Planet Coaster 1 had but better&more.

Coaster suggestions:

• Better Coaster Smoothing Tool CONFIRMED
• Continuous roll/heartline
• Smooth transition in inversions (so you don't have to go straight into a corkcrew etc)
• Save custom elements or track pieces
• Coaster switch tracks CONFIRMED?
• Manual control a coaster (No Limits 2 style)
• Multiple stations on a coaster so you could use a coaster as a transport ride like in RCT1/2.
• Coaster with dual stations for fast loading/unloading
• Better catwalks and netting for coasters that go over a path.
• Test coasters from a certain point in the track (so it doesn't have to start at the station every single time while building it)

Path/Blueprint Suggestions:

• Place paths with curb but also choose if you want to have lighting, benches, bins automatic deployed while making a path. This will just make things easier for big parks. Of course you can also choose to still manually deploy lights, benches and bins on paths too if you want. CONFIRMED?
• Better Pathing tool in general so much more responsive less fiddly. CONFIRMED
• Save paths with blueprints CONFIRMED
• Save ride/coaster entrance/exits with blueprints. CONFIRMED

Other suggestions:

• Spline based fence tool
• Spline-based tool for other objects such as rocks/trees/lights
• Easier object brush tool to plant trees, bushes, rocks (randomize and rotate them while painting the terrain with this object brush tool). Just select what trees you want and you can fill up big parts of the park at once to make a forest etc. CONFIRMED
• Weather, rain and umbrella shops CONFIRMED
• Paint more than 8 terrain materials CONFIRMED
• More realistic scenery objects to make things like offices, backstage area and so on.
• More see-through glass windows, roof pieces etc so it's easier to make indoor sections.

And this was about it what I could currently think of, maybe I will have some more suggestions over time but I think these are must-haves.

Thanks!

 

[WingardiumLevicoaster]

The only thing with wanting perfect future proof optimisation is there is a limit to what is possible. You can't have endless building without some computers slowing down at some point. It will always slow down for someone, no matter how much optimisation there is. Also optimisation can often be a trade off on quality of graphics or other limits. It heavily depends on the spec of the PC. It will always be the case to an extent. Optimisation isn't an endless magic thing. I know my current PC isn't going to cut it even if planet coaster 2 is better optimised then the original as it is also better graphically. My PC is 8 years old and struggles with latest release games. But played Planet Coaster fairly well. I will need a new computer no matter what, it isn't going to reach the expected minimum specs I'm sure. It doesn't for a lot of new games.

 

[DisneylandFreak]

It seems from the Demo some of the Planco Youtubers saw most of your suggestions regarding performance and pathing are being adressed. I like your list

 

[Ramm]

The only thing with wanting perfect future proof optimisation is there is a limit to what is possible. You can't have endless building without some computers slowing down at some point. It will always slow down for someone, no matter how much optimisation there is. Also optimisation can often be a trade off on quality of graphics or other limits. It heavily depends on the spec of the PC. It will always be the case to an extent. Optimisation isn't an endless magic thing. I know my current PC isn't going to cut it even if planet coaster 2 is better optimised then the original as it is also better graphically. My PC is 8 years old and struggles with latest release games. But played Planet Coaster fairly well. I will need a new computer no matter what, it isn't going to reach the expected minimum specs I'm sure. It doesn't for a lot of new games.

This is not completely true since Planet Coaster 1 is a game from 2016 and still runs bad on todays latest hardware. So it is not future proof.

The problem is with engine limitations or something. The directx version if my memory serves me right.

Obviously the game would run bad on slow PC's that just meet the bare minimum specs. That's why future proofing means that even if it runs bad on your PC now, in 3 years when you upgrade it will run better.

I don't understand how your PC can struggle with latest titles but planet coaster runs 'fairly well' because it is a well known fact the optimization in planet coaster is bad. Unless you only have small parks, big parks are still impossible even on the latest hardware.

It seems from the Demo some of the Planco Youtubers saw most of your suggestions regarding performance and pathing are being adressed. I like your list

thank you

 

[rct-uwe1]

I would like to have a light frame (Lichtraumprofil) or however it's called (this will follow the track while construction to avoid any collisions). And for PC2 this thing could have a mounted camera and a manual movement on the track. Plus attached terrain tool and scenery placement. This would allow especially darkride constructions in a new dimension.

 

[rct-uwe1]

What I always hated and still do is the layout of the pirate ship ride. I hate to see those ugly stairs on the platform. It should be flat with slopes to the sides to match row entries of the ship but everything underneath should be in ground.

And please adjust the ride bases. The actual tea cup ride base is far too big. There's no need for as much space around the rotating platform.

 

[DistortAMG]

This is not completely true since Planet Coaster 1 is a game from 2016 and still runs bad on todays latest hardware. So it is not future proof.

The problem is with engine limitations or something. The directx version if my memory serves me right.

Not entirely true. Today's latest hardware, an RTX 4090 and a 13th or 14th Gen i9 run the game very, very well. With huge, huge parks and tens of thousands of guests (+20,000) you are talking well, well above 30fps, some cases closer to 60, in 4k, running ultra. Even a much slower 4 year old i9-10850k and RTX 2080 run the game very well with say. 15000 guests. It runs much, much better now on todays hardware.

Source: https://youtu.be/9SegehLNxdI?si=U3sxSIso5WwnZyEi


Could they handle the peep simulation better? Sure, I am sure there have been many advancements in that area since 2016. That is a huge bottleneck in performance and will tank FPS, the peeps are probably the biggest bottleneck in the game, all of that is done on the CPU, pathfinding and what not. Nothing to do with DirectX, except for drawing the peeps themselves but that is fairly cheap.

At the end of the day, you are playing a game where you can build as much as you want. Every single computer on this planet has a finite amount of processing power and capability. The more you build, the more resources you use up in your computers pre determined capability. You are ALWAYS going to at one point or another hit the performance threshold and tank the FPS. That is the cost you pay for essentially having the capability to build unlimited, there is no work around if you want to build unlimited things you will always eventually, run into bad performance. The more powerful your PC, the longer it will take but you will always hit it. DirectX 12 in this game isn't going to be some magical thing that solves this issue, it will still exist in PC2 as long as we can build as much as we want.

With the guests being the biggest bottleneck to performance I hope we see improvements there. I am sure Planet Zoo does some pre calculations caching of maybe path finding upon loading a park, regarding this to save on CPU tasks when in game, so hopefully we will see this improved on for PC2.

But just to manage expectations, DX12 will not solve the issues people seem to think it will solve. The DX11 limitations were not the main limiting factor in PC1 and it won't be in any game where you can essentially build as much as you want, putting more and more strain on your hardware. DX 11 was still a multithreaded renderer, DX 12 does handle things better yes, but draw calls were not the main issue. Guest were.

Mythica - Land of Myths and Legends
Take the above park. Running this extremely dense and scenery heavy park on my average Ryzen 5700x, 32GB 3400 RAM and RTX 3070 system, I get between 35 to 40FPS on an empty park. CPU utilisation is about 30%, GPU utilisation is 100%. This shows a very, very clear bottleneck on the GPU. Not DirectX. The GPU is running as fast as it possibly can, it is being fed all the information it can handle. If put a new more powerful GPU in my PC, my FPS would increase in the game in this setting.

When you open the park and the tens of thousands of guests arrive, the FPS tanks to about 10 to 15 and the CPU now becomes the bottleneck. You can observe GPU usage dropping to about 50%, and CPU usage rising hugely, to around 70 to 80% in some cases. This GPU is now bottlenecked by the CPU. This isn't because of the DX11 draw calls I do not think, it more due to what I assume is the heavily multi threaded path finding algorithms taxing all 16 threads of the CPU. There is still room for DX11 to run just fine and issue it's draw calls as the CPU isn't at 100% and I've just explained that an empty, scenery heavy park DX11 isn't bottlenecking anything. The CPU isn't totally used up so DX12 and the seemingly assumption of mutithreaded draw calls won't do a single thing to improve performance here. What is going on though is everything else the CPU has to do to render each frame apart from DX draw calls and path finding is now being held up due to the heavy calculations required for the guests, holding everything up, lowering FPS. Things such as physics, collision detection, audio to name a few all need to be calculated per frame and everything needs to be ready at the same time. Changing to DX12 won't change ANY of this as it is not the bottleneck. Rendering the guests is seemingly very cheap. In this scenario, putting in a new GPU won't change the FPS here, putting in a new CPU will though as that is the bottleneck but not quite due to DX.

Having 12 will of course help streamline things and give access to many new shiny features such as DLSS and ray tracing for example. I don't think ray tracking should be added to a game like this due to the heavy CPU usage it requires and that for sure will be needed elsewhere. Besides, you can get lighting and shadows looking almost as good, including darkness inside buildings without ray tracing at HUGE cost saving to performance. No Limits 2 coaster simulator does this with buildings very very well with some nice features in their deffered renderer.

Resizable bar /Smart Access Memory which is switched on by default on most modern PCs now will have a much bigger impact on improving performance than DX12.

I'm sure there will be many new optimisations and improvements to the back end and engine that will be added to mitigate the performance issues as much as possible. But that won't get rid of what is essentially an almost impossible problem to get rid of. You cannot cheat maths and physics.

To put it simply, DX12 is not the magic API that is going to solve everyone's performance issues, as has been the common consensus around here since 2016.

 

[Ramm]

^Thanks for the detailed reply. Unfortunately I haven't switched over to 14th gen but I would very much like to experience the difference and load up my old parks that I never came to finishing due to the fps becoming too low. I'm really curious if it will run better.

I'm not sure about those youtube videos though, because I think they record it in 4 x less speed than in post processing they do it 4x faster to make it run at 60fps or something

 

[DistortAMG]

^Thanks for the detailed reply. Unfortunately I haven't switched over to 14th gen but I would very much like to experience the difference and load up my old parks that I never came to finishing due to the fps becoming too low. I'm really curious if it will run better.

I'm not sure about those youtube videos though, because I think they record it in 4 x less speed than in post processing they do it 4x faster to make it run at 60fps or something

Channel 5 gaming is a good indication of real world performance with 14th Gen and a 4090. It's very good. I would expect so too with that sort of hardware.

I've seen the performance with my own eyes on a friends PC. You can get well above 30fps with 20,000+ guests on ultra in 4k.

 

[Ramm]

Channel 5 gaming is a good indication of real world performance with 14th Gen and a 4090. It's very good. I would expect so too with that sort of hardware.

I've seen the performance with my own eyes on a friends PC. You can get well above 30fps with 20,000+ guests on ultra in 4k.

But how many rides and shops and so on?

I also watched Channel 5 gaming and he clearly states that the performance is bad and he gets 15fps and he records at a quarter speed, then later speeds it up 4x in post processing to make the videos 60fps.

So what you see on youtube is not the actual game performance

 

[DistortAMG]

But how many rides and shops and so on?

I also watched Channel 5 gaming and he clearly states that the performance is bad and he gets 15fps and he records at a quarter speed, then later speeds it up 4x in post processing to make the videos 60fps.

So what you see on youtube is not the actual game performance

You are watching old videos and basing performance off that then.

Look I've seen it with my own eyes, huge parks, lots of scenery, shops and rides, u mean loads. Performance is good. I would expect so for £3000.

 

[Ramm]

You are watching old videos and basing performance off that then.

Look I've seen it with my own eyes, huge parks, lots of scenery, shops and rides, u mean loads. Performance is good. I would expect so for £3000.

Channel 5 gaming uploaded 1 day ago stating from 1 minute and 50 seconds in that he recorded at negative 400% speed and later speed it up to get a smooth framerate:

Source: https://youtu.be/-Dthk2NDveo?t=108

I wish I could be optimistic but i'm gonna be careful. For me having a big park with good performance is a must-have. I know big is subjective, and that park from channel5 doesn't even have many rides just a lot of scenery. My parks are always very basic scenery but lots of rides and the game has always struggled with it. For me when you build your dream park I guess you should have at least 20 coasters right and 30 rides or something but preferably more of course and still have 15k guests and some scenery and lights and I still be able to run 60fps on 4k. That is just what I expect in 2024 and I don't wanna settle for less or defend it when it doesn't meet this very basic standard like people did when Cities Skylines 2 was released or with the performance of Planet Coaster 1 on release.

And yes I am aware it's a sandbox and you can do infinite things on finite resources etc but it is just my opinion that the game should be both optimized and future proof-ed way more than PC1.

 

[DisneylandFreak]

From what Rudi said the park they showed of in the demo is the one from the trailer. We never see the whole park but it seems like a small to middle sized one. According to Rudi it has stable 60FPS. If I open up a similar sized park with about the same detail in planco 1 (maybe a little more detailed) I get about 45-50 fps. I have decent hardware,not the newest. What I think we can infer from this is that they at least slightly tweaked the performance for this sequel. I also realise there is only so much they can do if they want parks to be rendered in full constantly,especially if they are big and very detailed.

 

[Ramm]

From what Rudi said the park they showed of in the demo is the one from the trailer. We never see the whole park but it seems like a small to middle sized one. According to Rudi it has stable 60FPS. If I open up a similar sized park with about the same detail in planco 1 (maybe a little more detailed) I get about 45-50 fps. I have decent hardware,not the newest. What I think we can infer from this is that they at least slightly tweaked the performance for this sequel. I also realise there is only so much they can do if they want parks to be rendered in full constantly,especially if they are big and very detailed.

I guess people who play Planet Zoo big parks on max settings maybe could tell us if the game runs good if you build crazy big?

 

[Ramm]

Edited what is confirmed. So nice to see you can save paths and entrances and have an object brush tool and that you can place paths with lights and bins (although I dont think we have seen it yet, I think they confirmed it). Same goes for switch tracks, heard talks about it but have we seen it?