Wild! RCT3 and Graphics - Printable Version +- RCTgo Forums (https://forums.rctgo.com) +-- Forum: The Games (https://forums.rctgo.com/forum-8.html) +--- Forum: RollerCoaster Tycoon Discussion (https://forums.rctgo.com/forum-9.html) +--- Thread: Wild! RCT3 and Graphics (/thread-9520.html)

RCT3 and Graphics - FlightToAtlantis - Jun 15, 2013 During my recent screenshots of Marina Towers I noticed some of the shots didn’t come out as expected. I believe these shots may give some insight as to how the game engine works. This is one of the shots I took of a corner of my park. You may remember the shot looking like this: However, one of the shots of the exact same location taken either a moment before or a moment after the shot above, that shot came out in this way: A shot like this makes me wish I was using shadows in my park at the time because I would have liked to see if they would have rendered in the second image. As you see the supports show in the second image and the only ride showing is Dolphin show. The shadowing in the sea and the flickering surface highlights on the sea bottom are also in the irregular image. The entire pool structure, food court/arcade/promenade and pool shops structures are missing as is the user interface and the surface of the sea. I notice the Gondoletta track is there but the Gondoletta boats are not. Here are a few other shots to show the difference between what we see and what it is the game apparently renders behind the scenes. Following is the shot of the shark that I took: And this is the irregular version of the same image, same exact location taken either just prior or just after the photo above: It seems the RCT3 game engine renders what we see in layers, and that in the irregular images some of the layers aren't showing. For simplicity and convenience I’ve divided the images into what I call regular and irregular images but I’m sure the game engine uses more than two layers to show us what we see while we’re playing the game. Again, here is my regular birds shot: And the irregular one I found among all the images. I see the clouds show in both images. If it was an overcast day I wonder if the drizzly haze would have shown in the irregular image: Except for the final image ("06 Irregular") I haven’t lightened or darkened any of these images so it appears the time of day lighting is also on a layer. I wish they were some night shots so we could see if the lights and reflections show in the irregular images. Here’s the Kiddie Korner image that I took: And the irregular image of Kiddie Korner, again exact same location and a second before or after the shot above: While getting my shots ready for my park update I can remember observing one of the images I took from above Lazy river and noticing that the portion of that ride’s supports that were underwater were the only supports showing in that image. I can’t locate that image, though – I know I’ve deleted it and I’ve been through the recycle bin so I think I deleted it permanently Again, view over Pool Complex showing Inverurie Inlet: And same view, irregular version below: It’s interesting to see that in this image some of the vegetation is showing and some isn’t. The pool ladder is there but the Jacuzzis, lounges and pool lights are not. I’m also not sure what has happened below the small square fountains I placed in the Pool Complex. I placed the pool then placed all the scenery as a single structure without myself adjusting any of the landscape but it appears there’s space between the bottom of the fountain and the level of the ground (in the middle of the image just off to the right). Now this following pair of images is particularly interesting for two reasons. One being I took two shots from approximately the same area to check later for the better one. One of the shots, this one following, all of the takes were regular: This upcoming shot, with the camera moved slightly, all of the takes were irregular: What’s also interesting about this shot is it shows that beneath some of the pool out of sight where nobody can see it the teak cladding is being rendered. This shot was originally darker and I have deliberately lightened this image so we can see the cladding more clearly. But in image “05 Irregular” above we can’t see any cladding. Mind you, image “06” is a raised pool and the decking for the raised section is at the top of the Atlantis cladding that can be seen in “06 Irregular” so if the ground were transparent in image “05 Irregular” we might see all the cladding for pool complex under the ground. What a waste of polys! RE: RCT3 and Graphics - FlightToAtlantis - Jul 27, 2013 Here are the last of the images showing what appears to be the different layers in which the game is rendered. By the way, both in this post and the previous post I’ve taken all these regular/irregular images with a third party screen shot application. This one is an out-take that I didn’t post with my park update. This is the same shot as before but it came out in this irregular way. I’ve just noticed that the text I’ve added has hidden what’s in the top left corner but there’s not much to see in this image anyway. Here’s the second irregular shot now showing supports and the entire aquarium including the seaweeds but excluding the glass and the water inside. In the shark tank at the right of the image I’ve taken a close look at the original image and that might be a shark in the centre section of the tank all the way over to the right side. I really can’t tell for sure but if it’s a shark there it’s not appearing in the next image below. The bottom right of these images reminds me that I still need to go around my park and match all the terrain tile edges. Here’s another out-take, this time of the north-west corner of my park. In this irregular image the rides and water are completely gone while the ripple effect of the water remains on the bottom of what used to be the sea. Here’s another image that I think I used in my park update, or perhaps I used another one similar to this. Again this one came out regular. And here’s one of the irregular versions. Oddly, in this one only some of the user interface is displaying. Now remember in image “02” above, how deep the supports are under the sea below that path in the foreground. In image “03” here this is the same path in the same area but notice how much of the supports have been rendered in the following image. I couldn’t tell you if the game renders them like that so the entire support won’t show through the water during gameplay or if this is part of the irregularity. In that image most of the pool scenery is gone while some of the pool plantings appear to be still there. The pool path up by the Egyptian pillars is gone but the fence for that path is still in place, and lazy river supports are there but there’s no river. A-a-a-nd finally on this topic, here’s the last regular shot. And the irregular one. You can see the differences for yourself. __________________________________________________________ At the end of my previous post on this topic I mentioned an issue with swimming pools, teak cladding and polys. Getting back to this topic I opened a workbench and began a pool, dragging out one that was somewhere between 40x40 to maybe 50x50. For clarity I deliberately made the pool deck this high. Right away we can see the number of polys that are suggested in the ghost image. Remember this is only pool decking. And here is the same pool deck with pool placed and one of the LOD options turned down. I thought a close-up of this hypnotic image would be in order to show all the polys that are rendered beneath the pool decking: Let’s see what would happen if we put a pool into this deck: And played with the LOD’s again: I’m not sure I understand why there’s an alteration to about four tiles in width on the left side and top edge of the pool in an area where there is no pool water. However, here’s another close up: There don’t seem to be any unnecessary polys below the pool where there is water in which the guests can swim. However, if you look closely at the bottom of the previous image you will see water ripple effects are not only on the surface of the pool but on the ground at the pool base – they might be easier to see if I hadn’t made the deck quite so tall. Let’s turn up that LOD again. Same pool in same view, different LOD. But now we know that there are thousands of unseen polys being rendered in all their splendid teak glory in an area of the game where no one can see them. __________________________________________________________ You’ll be able to tell from the following images that I’ve been continuing work on Marina Towers Park and Spa Resort. Here it will be observed I’ve dropped that park’s pool complex into a workbench to experiment. While playing around with the pool, I noticed two ‘bald’ guests. You may recall I mentioned bald guests in this previous post http://forums.rctgo.com/thread-7326.html. In this workbench I’ve used different peeps generated by Peep Factory as I did from the previous bald post. You may recall I decided the next time I saw a bald guest I’d zoom in and take some additional shots. Let’s example Nicholas Rios, first with two side images: And now a full rear image. In these images the colour of the skull is more skin-toned at the top of his head, while further down behind the skull it gets a bit redder and is looking more hair-toned. I don’t believe these guests are bald. I think they have dark skins with very light hair, in addition to which this peep sub-type has receding hair which would suggest the remaining hair is thin. But check this out. Let’s get closer to Nicholas and see what happens. In the image above the camera is actually looking inside the guest’s head, showing that guest polys can be seen even from inside the guest. We should be seeing nothing but the distant trees in that space, not the inside of the guest’s face. In the next image the shoulder has disappeared and you can see the rear side of the poly that would ordinarily show Nicholas’s chest. Then the far side of his neck can be seen and the entire ear all the way down to its ‘root.’ At least we only see the guest’s eyes on his outside – that’s with this particular guest type, anyway. These images would indicate that for every guest we see in our park the game renders two guests, the seldom seen guest that’s inside out, and the guests we all know that are right side around. So if we have a park running with 2,500 guests, with RCT3 actually running these guests in duplicate the game engine is compiling polys for 5,000 guests. If Nicholas Rios has a camcorder in his hand, has eaten a burger, drank a milkshake and ridden the woodie, in view of the fact the game renders the guest polys in duplicate if we assume what I've written at the start of this sentence represents four sets of data do you think the game engine compiles four sets of data or eight sets? __________________________________________________________ Which brings to mind a question: is this the type of thing the .ovlCleaner will remove? Will it take hidden polys out of the game? Could someone share with us what it is that the .ovlCleaner removes? RE: RCT3 and Graphics - sparkler1500 - Jul 27, 2013 Could your computer maybe be doing some of this stuff instead of it being in the game? (I'm programming illiterate so please excuse the stupidity) RE: RCT3 and Graphics - FlightToAtlantis - Jul 30, 2013 I’d considered the possibility. 1. Regular/irregular images: There may be a slim possibility my machine might be responsible. However as I only see these irregular images in RCT3 screenshots and never during RCT3game play, never in videos recorded from RCT3 and never in screenshots taken from any other application I highly doubt it’s my machine. 2. Swimming pools and teak cladding polys: No. 3. Guest polys: No. RE: RCT3 and Graphics - sparkler1500 - Jul 30, 2013 So it's only when you take screenshots?!?! That's kinda cool that it reveals some of the game design (and design flaws)!