Virtual Disney (w/screenshots)

[DigitalDisney]

Virtual Disney (w/screenshots) - Upd 6/8/03

As some of you may remember, I wanted to create a virtual version of Shrunken Neds Junior Jungle Boats. That would have eventually lead to a complete Virtual WDW.

Well, I chose to temporarily skip the Shrunken Ned Simulator and jump straight ahead to the Virtual WDW simulator.

Screenshots from 6/8/03:

A test shot of the newly implemented light system. Of course, more lights will be added to the scene.

Here are the first screenshots (6/7/03):
- http://www.progoth.com/~darkhelmet/wdw/shots/060703/poly-day-smaller.jpg
Looking across (what will be) the marina at the Poly

- http://www.progoth.com/~darkhelmet/wdw/shots/060703/poly-contemp-sunset-smaller.jpg
Looking at the contemp from the Poly at sunset

- http://www.progoth.com/~darkhelmet/wdw/shots/060703/poly-night-smaller.jpg
Looking across the Poly at night (there are stars in the pic, but you can't see them)

There will be a LOT more where these came from

Currently, the engine is very flexible. Very little is hard-coded into it.

It loads objects and scenes from XML files and renders them on the fly

It also supports real-time changes to the sky. At night, you can see a blue-ish horizon with stars overhead. Of course, they disappear at sunrise and reappear at sunset.

 

[SirNim]

Awesome! :sohappy:

 

[Brett]

All I can say is Wow! And good luck! I definently want to see more of this!!

 

[dylan0511]

That is sooo cool looking.
Good job!

:sohappy:

 

[MicBat]

Wow!!... VERY nice!

 

[Tocpe]

It looks like we're both on the same track (A virtual WDW) . I've been working on what sounds like the same thing. I've been looking into the Unreal engine to power my project. I'd love to hear the background on your decision to go down the path you chose.

I'm still in the process of researching the engine, but maybe we can combine efforts? I started a Yahoo Group a little while to provide a home base for the project:
http://groups.yahoo.com/group/VWDW_Group/

I'd love to hear your thoughts on maybe joining forces.

Tocpe

 

[DigitalDisney]

For now I am writing my own custom engine in C++ with OpenGL (vid), GLUT (windowing / input) and fMod (audio). Although I am strongly considering changing to Crystal Space because it can dynamically generate lightmaps on the fly, import many different model types, handle particles, etc., all of which would save me a lot of work and time.

The reason I'm using a custom engine is because this is a highly custom project. It needs to seemlessly combine indoor and outdoor environments at any time of day (the time and lighting need to adjust in real time). I also want a high degree of interactivity and dynamics in my world.

I also want weather, to some degree. It just wouldn't be a trip to WDW without one of those 5-6pm thunderstorms, would it?

Of course, this project is still in its infancy, and there is a long way to go.

Ideally, I want to include almost every aspect of WDW into this simulator. This includes (but is in no way limited to):
- waking up in a hotel in the morning and going to the theme park of your choice (by car, boat, foot, or bus)
- driving your car around WDW
- driving a water sprite down the Sassagoula or the Seven Seas Lagoon
- Watching parades or shows such as the Electrical Water Pageant
- Historical attractions could be included. Each attraction at WDW could have a timeline, and you could go back to any time you wanted to. So, you could experience the Skyway or World of Motion whenver you wanted to.

Do you have any screenshots of your project?

While UT2003 is a reasonable choice for game engines, the one big problem is that you are limiting yourself to people who own the game. Unless you pay the thousands of dollars for a license, you can only distribute levels and not the game engine itself.

If you want to get crazy, you can work with the Quake2 engine because it's now open source. However, that engine is only designed for indoor levels (this is natural behavior of a BSP engine).

 

[NowInc]

Id say use the doom3 engine or halflife 2 engine..but yeah..licencing issues...and since this is a HOBBY..and not a JOB...its not worth the cash youd have to spend..

Impressive none the less..keep up the good work.

 

[Tocpe]

Originally posted by DigitalDisney
For now I am writing my own custom engine in C++ with OpenGL (vid), GLUT (windowing / input) and fMod (audio). Although I am strongly considering changing to Crystal Space because it can dynamically generate lightmaps on the fly, import many different model types, handle particles, etc., all of which would save me a lot of work and time.

My hat's off to you for tackling something like that! :sohappy:

I had thought about something like that but I quickly realized that I'm not a programmer. I know there are tons of advantages to having your own custom engine. Many of which you've presented. The reason I started looking into off-the-shelf engines like the Unreal engine was the base of pre-existing level designers (some of who might be Disney fans...) and innerworking with existing modeling software like GMax and 3D Max studio (more on that later).

The reason I'm using a custom engine is because this is a highly custom project. It needs to seemlessly combine indoor and outdoor environments at any time of day (the time and lighting need to adjust in real time). I also want a high degree of interactivity and dynamics in my world.

Oh I totally agree. That's one of the major drawbacks to the game engines that I've found. They hate huge open spaces. Time of day and weather...pretty non-existant for most FPS. I have found some tutorials on some workarounds for the openspaces...but I'm not totally convinced yet...

Sound is important too. Have you figured out a way to put in background music in the areas of the park. For instance, the background music played around Tomorrowland through the speakers in bushes and stuff?

I also want weather, to some degree. It just wouldn't be a trip to WDW without one of those 5-6pm thunderstorms, would it?

Nope not at all.

Of course, this project is still in its infancy, and there is a long way to go.

Are you open to having help? I can provide modelling and texturing help along with layout and park attraction and layout accuracy.

Ideally, I want to include almost every aspect of WDW into this simulator. This includes (but is in no way limited to):
- waking up in a hotel in the morning and going to the theme park of your choice (by car, boat, foot, or bus)
- driving your car around WDW
- driving a water sprite down the Sassagoula or the Seven Seas Lagoon
- Watching parades or shows such as the Electrical Water Pageant
- Historical attractions could be included. Each attraction at WDW could have a timeline, and you could go back to any time you wanted to. So, you could experience the Skyway or World of Motion whenver you wanted to.

Wow! That sounds like a huge project. I was outlining a much smaller one:

Here's some of what I had worked out so far:
Guidelines and plans for the Virtual Walt Disney World (WDW) Project.
5/28/03
This is a rough outline for the project's scope and plans. It will be modified and updated
as needed.

Pre-Production
1. Research suitable engine for rendering in real time an immersive WDW experience.

The engine should be capable of the following (in no particular order):
a. Realistic objects at a decent to high resolution and scalable to fit different users system requirements.
b. The ability to handle detailed texture maps and shaders such as bump maps to compensate for mesh density.
c. Sound - 3D if possible. Compact and high quailty.
d. Supports more than one user at a time i.e. multiplayer. This would allow a group to explore together.

2. Locate low to no cost model software to allow for the community to contribute to modelmaking. This will facilitate build of an area/park/etc faster.

3. Finish collecting primary researh materials and plans for initial park: EPCOT. We will start with Epcot due to it's geometrically shaped buildings. See the section below "Production" for more info.

4. We will begin with one park and model onlt the exteriors and environments first. This will keep us from getting lost in the details. After completeing the first park we will move onto the next until all 4 of the parks are complete. Then we have two choices:

Begin modelling individual attractions (assined to individual teams) or complete the modelling of the rest of the property and non-park attractions (hotels, water parks etc.)

5. There is no wayy all of this can be stored in one model/file what-have-you so we will break each area up to as many sub-levels as needed. For example, Epcot may consist of two main levels: Futureworld and World Showcase. Each attraction in each of those levels will be another "level" unto the main level of FW or WS.

Production

Production will follow this path:

1 EPCOT
a. Future World
b. World Showcase (place holders for countries)
c. World Showcase individual exterior pavilion details

*****More to be written here*****


Another reason I wanted to starte with Epcot is because I already have about 75-80% of it built with about a 85% accuracy as close as I can figure.

Do you have any screenshots of your project?

Yep

While UT2003 is a reasonable choice for game engines, the one big problem is that you are limiting yourself to people who own the game. Unless you pay the thousands of dollars for a license, you can only distribute levels and not the game engine itself.

Yeah, I know, another one of the drawbacks of using an off-the-shelf GE. But I had intended to on distribute the levels. Licensing is simply not an option. It's only a hobby. But as for people who only own the game, they can always go out to their local computer store and pick up a copy. Granted this doesn't apply to people in some countries and I know so this again falls into a mark against the FPS engines. :brick: :animwink:

If you want to get crazy, you can work with the Quake2 engine because it's now open source. However, that engine is only designed for indoor levels (this is natural behavior of a BSP engine).

I think I'll leave the engine stuff up to you.

 

[DigitalDisney]

I had thought about something like that but I quickly realized that I'm not a programmer. I know there are tons of advantages to having your own custom engine. Many of which you've presented. The reason I started looking into off-the-shelf engines like the Unreal engine was the base of pre-existing level designers (some of who might be Disney fans...) and innerworking with existing modeling software like GMax and 3D Max studio (more on that later).

This is perfectly reasonable, and using Unreal is the best option for a non-programmer. Heck, when I was crazy about Half-Life back in the day, I tried to recreate the Jurassic Park River Adventure at IOA with no luck.

I'm sure you're already aware of this, but many existing 3d engines can import other objects. I was planning on adding support or at least 3DS and MD2/MD3 into my engine.

Sound is important too. Have you figured out a way to put in background music in the areas of the park. For instance, the background music played around Tomorrowland through the speakers in bushes and stuff?

I was going to cheat a little bit on this one. My world will be divided into many sectors or zones. In these zones, you can specify ambient non-directional sounds such as birds, growls, etc. The music could even be started at a random point so it's not the same experience every single time. Of course, the ambient sounds would fade in and out instead of just starting and stopping when you enter different sectors.

One drawback that I see is that I obviously cannot include background music loops with the program. They would make it a rediculously large download, and there are copyright issues involved too. However, I can make a system where you can insert your own sounds into the game.

Are you open to having help? I can provide modelling and texturing help along with layout and park attraction and layout accuracy.

Definately. Once I have a solid foundation in terms of the program's structure, all that will essentially be left is modelling, which will be a huge task. Eventually, this will be open to the public so we can have everybody contributing to this massive project.

However, that point is still a ways away. I still need to work on my engine a lot more before I get to that stage. I need to add animation capabilities (say a spinning waterwheel), scripting (boats going across the Seven Seas Lagoon), some degree of interactiveness (Buzz Lightyear anyone?), and I need to work on making the world seamless so that you can walk around anywhere, and the system would manage resources effeciently without destroying the framerate or bogging down memory/CPU.

Your POV-Ray screens look good, and they even give me a little deja vu (I tried to do the exact same thing a long time ago with no luck). Your Epcot model should prove to be an excellent reference for the exteriors of Epcot buildings. Getting the interiors is a whole other ballpark

You should be able to use an existing game engine (HL, Quake3, Quake2, Unreal, etc) for any interior scene and some exteriors. For example, you could probably recreate Horizons with one of these systems with relative ease, and it would even be ridable. The only thing you couldn't pull off would be the Omnisphere, finale, character animations, and a couple other small things. Otherwise, you can basically recreate the whole thing. The same goes for any other indoor attraction I can think of.

 

[blm07]

I remember someone working on putting Epcot into the Serious Sam engine, but he hasn't updated for quite some time. I actually think that would be a good engine cuase it does real good with wide open areas (and it works good on low end machines).

 

[Tocpe]

Originally posted by DigitalDisney
This is perfectly reasonable, and using Unreal is the best option for a non-programmer. Heck, when I was crazy about Half-Life back in the day, I tried to recreate the Jurassic Park River Adventure at IOA with no luck.

I'm sure you're already aware of this, but many existing 3d engines can import other objects. I was planning on adding support or at least 3DS and MD2/MD3 into my engine.

3DS support would be good. Might I also suggest OBJ support too? I think between those two formats you'd have most of your major bases covered. How are you building the models now?

I was going to cheat a little bit on this one. My world will be divided into many sectors or zones. In these zones, you can specify ambient non-directional sounds such as birds, growls, etc. The music could even be started at a random point so it's not the same experience every single time. Of course, the ambient sounds would fade in and out instead of just starting and stopping when you enter different sectors.

I think the UT2003 will allow something like that only with proxomity triggers. Fade in and out capabilities are provided automatically as well. Are you planning on implimenting a 3D sound system?

One drawback that I see is that I obviously cannot include background music loops with the program. They would make it a rediculously large download, and there are copyright issues involved too. However, I can make a system where you can insert your own sounds into the game.

oooh :hammer: Didn't even occur to me. Well the size problem did, but the other didn't. Yeah I think an inserting system would cover it there.

Definately. Once I have a solid foundation in terms of the program's structure, all that will essentially be left is modelling, which will be a huge task. Eventually, this will be open to the public so we can have everybody contributing to this massive project.

What's your estimate on when you'll be ready for modelers? 6 months, a year, two years?

You should be able to use an existing game engine (HL, Quake3, Quake2, Unreal, etc) for any interior scene and some exteriors. For example, you could probably recreate Horizons with one of these systems with relative ease, and it would even be ridable. The only thing you couldn't pull off would be the Omnisphere, finale, character animations, and a couple other small things. Otherwise, you can basically recreate the whole thing. The same goes for any other indoor attraction I can think of.

Yeah I'm confident that the game engine will do interiors. But do you think that with careful placement of "level" zones that you could fake the expansiveness of a park. Granted it wouldn't be seamless. i.e. If you walk across the bridge to WS in Epcot you hit a loading zone and WS loads. I haven't pushed the UT2003 engine to see how much it'll hold in open areas. Do you know how it does?

 

[DigitalDisney]

3DS support would be good. Might I also suggest OBJ support too? I think between those two formats you'd have most of your major bases covered. How are you building the models now?

Oh yes. I will have support for 3DS, OBJ, Lightwave, etc. If I can find the code to load a filetype, I will try to implement it in my program. I know I've seen code for 3DS, OBJ, and Lightwave. I'm sure I can find code for Maya, Truespace and other popular object formats as well.

Currently, objects are created in my proprietary XML-based format. It's very similar to OBJ format, except that it's structured nicely and includes support for textures and lightmaps.

Of course, this proprietary format won't work so well when creating custom objects, but it works nicely for the simple objects I'm creating now.

I think the UT2003 will allow something like that only with proxomity triggers. Fade in and out capabilities are provided automatically as well. Are you planning on implimenting a 3D sound system?

Definately. This will only add to the realism. Of course, the system will be scalable and mono and stereo will be supported as well. The goal is that this program won't require a 2Ghz CPU to run. I'm hoping that you'll just need a decent setup (a semi-modern 3D card and semi-modern audio card)

oooh Didn't even occur to me. Well the size problem did, but the other didn't. Yeah I think an inserting system would cover it there.

The same thing applies to videos, if I can somehow work videos into the engine (that would be something).

This is going to be a large project. Ultimately, it will take up hundreds of megs (if not over a gig) when music loops and such are thrown into the mix.

The important thing is to keep the project scalable. If a user only wants the MK, they can only download that portion. If somebody doesn't want resorts, they don't have to download them. This could even extend to things like store interiors and ride interiors. Ideally, this will be incredibly flexible. We'll see though

What's your estimate on when you'll be ready for modelers? 6 months, a year, two years?

Hopefully much sooner rather than later. I seem to be amking a decent amount of progress right now. My main concern is that I don't want people working on something, only to find out that they have to change their data because of a design change on my part.

Yeah I'm confident that the game engine will do interiors. But do you think that with careful placement of "level" zones that you could fake the expansiveness of a park. Granted it wouldn't be seamless. i.e. If you walk across the bridge to WS in Epcot you hit a loading zone and WS loads. I haven't pushed the UT2003 engine to see how much it'll hold in open areas. Do you know how it does?

I can't honestly say how an engine like that would do with open areas. I'm sure that engine technology has advanced quite a bit since I last toyed with them.

One option you might want to consider (I'm sure you've already thought of this) is using extremely low polygon models for these wide outdoor areas. You could simplify almost all WS pavilions to 10s of polygons when viewed from across the lagoon. The same could be applied to anything else. Once you get close to the pavilion / building, the higher-polygon model would replace the low-poly model.

I will be using the same technique with my engine (I'm not sure if the polygon rebuilding will occur automatically, I'm leaning towards them being provided by the designers)

A project of this calibre will undoubtedly require a lot of experimentation. As another poster mentioned, the Serious Sam engine may be a good choice because it can potentially handle outdoor scenes better than other engines. A lot of research and trial and error will be required to get something of this scale realized. It will be quite an interesting and hopefully very rewarding journey.

 

[Wilson]

Daggone, y'all are smart.

 

[Tocpe]

Originally posted by DigitalDisney
Oh yes. I will have support for 3DS, OBJ, Lightwave, etc. If I can find the code to load a filetype, I will try to implement it in my program. I know I've seen code for 3DS, OBJ, and Lightwave. I'm sure I can find code for Maya, Truespace and other popular object formats as well.

Excellent! The more formats, the more modelers that can be recruited.

Definately. This will only add to the realism. Of course, the system will be scalable and mono and stereo will be supported as well. The goal is that this program won't require a 2Ghz CPU to run. I'm hoping that you'll just need a decent setup (a semi-modern 3D card and semi-modern audio card)

That sounds great. It'd suck to have to have a super-computer to run this thing.

The same thing applies to videos, if I can somehow work videos into the engine (that would be something)

Have you thought about incorporating the videos into the lighting system or as part of the texturing. 3D max allows for both. You can project an image or video through the rays of light that a light produces. Or the diffusion map of a material displays the video. I don't know how the engine setup would differ between a Real-time and an engine like the one that runs 3D Max, but maybe the concepts could be shared?

The important thing is to keep the project scalable. If a user only wants the MK, they can only download that portion. If somebody doesn't want resorts, they don't have to download them. This could even extend to things like store interiors and ride interiors. Ideally, this will be incredibly flexible. We'll see though

Definately, and that would allow for periodic updates and releases. Have you thought about the possiblility of multiplayer experiences?

I can't honestly say how an engine like that would do with open areas. I'm sure that engine technology has advanced quite a bit since I last toyed with them.

I'm not sure myself, but since I've been talking with ya these last few days, I'm becoming less and less convinced that an-off the shelf engine can do it and do it well.

On a side note, do you have a max texture resolution spec yet? Just curious.

One option you might want to consider (I'm sure you've already thought of this) is using extremely low polygon models for these wide outdoor areas. You could simplify almost all WS pavilions to 10s of polygons when viewed from across the lagoon. The same could be applied to anything else. Once you get close to the pavilion / building, the higher-polygon model would replace the low-poly model.

Yeah I had thought about trying to impliment something like that in the project, but of course I can only do it if the engine supports it....

I will be using the same technique with my engine (I'm not sure if the polygon rebuilding will occur automatically, I'm leaning towards them being provided by the designers)

Excellent! That will have to be done in some form or fashion, especially when in Epcot or in one of the non-park areas too.

A project of this calibre will undoubtedly require a lot of experimentation.

No doubt. Well if you need a Beta tester, let me know, I'd love to get in on the ground floor of this project.

 

[Blizz]

Now forgive me if I am not up on all this stuff. I am still learning C++ and just finished with BASIC. But coudl you use what ever engine that powers the Sims or Dreamcatcher Entertainment's games like Dracula Reserection?

I might have no clue what I am talking about, but I know those games have the feel that I think you are looking for.

But I must say, I would love to play this when you finish! Good luck!

 

[isnet396]

WOW!

Can I be a beta tester?


hehe...I'm learning Java myself now, at harvard.....lemme know if i can help out.

 

[Blizz]

Originally posted by isnet396
WOW!

Can I be a beta tester?


hehe...I'm learning Java myself now, at harvard.....lemme know if i can help out.

Good luck with Java. I take that nextyear at my high school. I hear the first things you learn are a lot like C++.

PS - I also want to be a beta tester.

LOL

 

[Tocpe]

Well DigitalDisney looks like we've got some interest in your project Are you planning on setting up a site for the testers and project documentation when it's ready? Just wondering.

 

[isnet396]

Originally posted by Tocpe
Are you planning on setting up a site for the testers and project documentation when it's ready? Just wondering.

I can host it ......