Cool Thing of the Day
With the Unity II engine underway and our last mediocre success at lighting and weather we've been looking for an off the shelf solution a la SpeedTree for our sky. We have some pretty unique requirements for an MMO such as realism, vast zoneless land mass, time and season progression and oh yeah, you can fly through it! We'll I'll be daqmn4ed if this thing doesn't look HAWT.Labels: wwiiol
28 Comments:
Hawt.
You're linking to the tiny image though. Take the _tn off the end of the link. ;)
How much do you need?
Dunno yet really. Just starting looking at these. We've seen a few environment systems before but this one really knocks me out. Another contender would be Silver Lining.
Man if that would work without destroying my computer... Looks good. I hope you gusy can giver a try and it runs well.
So as a developer do they let you get kinda like a trial of the software to see if it will fit what you are looking for? Or do they require you to buy it outright before hand regardless if it is going to be a solution for you or not.
You can generally get a trial version of the API to see if it can be made to fit. in addition most of these places will help with a custom solution if you have the money and ultimately we'd probably be best served to hire a ronin coder to come in and help with the installation of some of these more exotic middle ware solutions to ensure best performance.
WOW.
I think the Hangar would almost fund that on their own!
Interestingly I see it includes a wind effect. Would that be something you could tie into the physics engine of the game?
That windwardmark stuff is very nice.
Probably not on the wind and physics. I can't really see us being able to but more so, I don't think players really want cross wind. Sure some flight sim guys want it but do gamers as a whole? Doubt it. Kinda like force feedback. Its great until you PvP with it and it gets you killed.
The more complex aspects of weather and the environment might not be a happy deliverable for the air-to-air game. Things look different from the point of view of the ground/sea players that are targets for air attack. They want aircraft target alignment to be realistically challenging.
Wind will be an important component of smoke/fog/low-altitude cloud movement at sea, as well.
Maybe the game needs wind that fades out above 300 meters.
I wouldn't assume that wind that affects clods has to affect physics on planes or bullets.
There is a much bigger issue though. Any of these weather solutions are not going to be designed to draw over 1000km without a seam or some serious computational hurt. They are going to have to be regionalized just like our terrain loading is done with super cells. That means a boundary. That means you can't have stuff blowing across that boundary. Well at least not without more muscle than anything I've yet seen.
At some level I know this can be done because Ive flown a LONG way in MSFSX. If we can buy that is another question.
I guess there where a couple of reasons I was asking about wind, and strangely neither where about flying!
I was thinking about wind effecting things like long range rifle fire, especialy sniping.
And of course I guess the real reason I was asking was with the hope that if we eventualy get a hydro model with the new graphics engine that the wind might effect the seas state.
One can but dream. :-)
As far as I can tell, Silverlining costs 250 USD.
Thats ..... affordable.
Along the lines of similar broad-scale physics-based environmental systems:
Does anyone offer a conceptually suitable system for ocean waves and hydrodynamics, designed to provide dynamic interactions with "floating" objects in a complex-shaped virtual volume and to interact with a virtual weather system?
This enterprise comes with two products fully integrated and working together:
Nimble (3D Cloud System)
WindLight (Lightning Engine with HDR features)
http://www.windwardmark.net/index.php
It's OpenGL 2.0, shaders in Cg, and a API in C++ with lots of features.
It even has a level of detail method for Terrain rendering, and self-shadowing of that terrain.
This thing seems to be born for Unity 2 :)
Their release date is early 2007.
Don't worry about the money... remember that this thing isn't of Unity I quality, it's a Unity II feature to place us ahead of the simulation engines, since Oleg's has done the same kind of clouds for their next release (Storm of War: Battle of Britain):
http://www.youtube.com/watch?v=Y9OWQ55n8ig&NR
I'm a builder gold, at the time where SpeedTree showed up, so i think that a lot of Builders gold will be interested in making it to happen for Unity II. Terrain II should be next gen in all the features, and a good sky put into pair with "Storm of War: Battle of Britain" it's a mandatory feature for 2008 :)
Ask how many people will become builder if you integrate this into Unity II, and you will be surprised.
Nimble\WindLight; It's almost ready to release, to drop them an email, asking for an API beta... doesn't hurt.
Cid you rock. You'll notice thought hat if you click on the title of the post it takes you there =]
Silver Lining seems to be a better fit for Unity I and while not nearlt as fully featured or as good looking as Nimble, it looks like a great deal at the price.
gophur, is it possible to group super cells? if you could some how sew a few of them together you could then start thinking about localized weather patterns that 'flow' across an area. create another cell entity called 'weather' and have it be the parent to super cell children then hook it into the map/strat/STO system so it gets translated to all the playes.
this way vlissingen could have a completely different weather pattern as say cambrai etc.
I think the single biggest need for wind would be its effect on smoke.
Sure, smoke should move but that isn't necessarily "wind".
For instance if a game has cloth billowing in the wind, and trees swaying in the wind, does that represent an actual physical force or is it merely an effect?
Wind to me is a physical affect that would produce a force on an object. making smoke move is like having our flags wave. Merely an effect, not a force.
right localized shader effect vs ... mmm .. meterorlogical.climate effect on a global scale.
but specifically regarding cloud patterns IE weather patterns storms etc moving across the map seemlesly can this be done across super cells at all? we are just talking about moving the seeds correct? this cant be done across super cells? or just a physical wind effect must be contained within a super cell?
just do it.
it is time.
Well that really depend on the middle ware we use. Silver Lining for example doesn't like to be more than 100km x 100km so we'd have to have several instances running to cover the whole game world. The problem is then that these instances will have boundaries that exist in the world so clouds moving across the edge of one would just disappear.
interesting. 100x100km eh? heh that sounds like so much room but it's really not on our map.
In the real world, low and medium altitude cloud patterns often are discontinuous at the boundaries of large bodies of water. Would the game-map boundaries of program-instances necessarily be rectilinear? If not, their boundaries along the Channel and North Sea might be able to be masked by means of coincidence with that realistic meteorological behavior.
Of course, that assumes the European weather mode that moves from the general Greenland direction, so that discontinuities normal to the general flow-direction would be visually sensible. When the weather source-direction is from the south Atlantic or the Sahara, instancing would be harder to hide using available features.
Is it necessarily the case, though, that a seed "movement" pattern that's synchronized across an instance-boundary and programmatically always limits cloud generation to discrete types that can be shrunken/broken up/made wispy out of existence in one instance as they approach a boundary, and expanded into existence as they originate at the corresponding boundary in the next instance, would look "bad"? As I imagine the effect that might result, I'd think the look could be acceptable.
Not real sure about the boundaries and to be honest, don't really care all that much as long as it isn't jarringly ugly oer immersion breaking. I think bang to buck starts dropping pretty fast every time we spend effort becoming expert weather simulators.
P: Hey can I fly through those clouds?
D: Yes.
P: Ahh cool! Nice sunset BTW.
D: Thanks. Did you see the rain storm yesterday?
P: Yeah, whatever, cool, I guess, at least I could fly over it.
D: Yeah. Did you notice that the clouds moved in the wind?
P: Huh, I guess. Hey how come the rain didn't make my helmet shiny and make the ground all slippery? You know what would be cool....................
..."bang to buck starts dropping pretty fast every time we spend effort becoming expert weather simulators".
Heh. Well, another perspective would be that the game's natural environment communicates very effectively regarding its immersive realism or lack thereof, because we're inherently attuned to it...and weather is a key element of the natural environment. Perhaps the dominant element, for air and naval players.
Meh... Plane guys want clear blue skies 24/7/365. maybe a few puffies to hide in or a spare high level deck to play cat and mouse in, that's about it.
Navy is different but arguably, infantry and ground is where weather has the most impact...ok you know, that's not even really arguable.
But modeling snowflakes is way more work than just having a rain sprite and compared to the amount of work is of almost no benefit. Lightning looks cool. If I spend the time to model its exact characteristics and also make sure that it fries circuits or brains depending on which type of object it contacts and of course we need to model the correct type of strike parameter so that it has preference to higher grounded targets.... bullshit.
I am 100% sure that if you started a "realistic 3D volumetric cloud system with weather support" funds program, a LOT of people would join that program and support it with money...
Just something you should give a thought!
Post a Comment
Links to this post:
Create a Link
<< Home