ROAM Planet rendering

[DavlexDesign]

G'day again,

Just checking stuff, and here is this shot AGAIN !!! running under OpenGL this time....

CloudsNew_08.jpg

It runs a little slower than DX, but at least it still looks the same, so Linux and Mac users won't fall short on using this thing.

Alex

[ultramedia]

Well that's considerate

For the linux users, you may want to use a larger size for the text and maybe even comic sans for the font.

And not too many big words.

[Jabberwocky]

Really digging the latest screenshots.
The atmosphere and fog look great on the landscape. And that planet shot from space is really cool.

After you've finished tweaking your atmosphere, water, and clouds, are you considering tacking foliage? In my opinion, that'd have the biggest impact visually of anything you could do next. If so, are you thinking of using PagedGeometry, SpeedTree, or your own custom solution?

[DavlexDesign]

G'day Jabberwocky,

Thanks for the nice comments mate, much appreciated.

are you considering tacking foliage? In my opinion, that'd have the biggest impact visually of anything you could do next. If so, are you thinking of using PagedGeometry, SpeedTree, or your own custom solution?

Yeah, that's my next big reveal mate, Paged geometry is nice, and so is SpeedTree (albeit a bit expensive), I already have a custom placement system in place that will easily integrate with the two solutions you mentioned, but I'm going to have a go at something I've been toying around with first, If it doesn't pan out and ends up in the junk drawer, then I'll look at Speedtree I think, the only thing I'm worried about with outside solutions, is that the procedural nature of this thing will end up going by the way side. I really want to keep to the original plan of procedural content, considering when I have a Galaxy to populate, different looking foliage and props will end up being way too much to store. I've kept the program down to 300kb at the moment, with about 30 mb for textures and stuff, even though I'm not using them all, I will break it all down to the bare bones when I upload the demo, then we'll see just how big this thing will be.

The foliage thing I'm toying with is a sort of proceduraly modifiable building block system, so I have all these little primitives that I put together to build a whole depending on characteristics of the celestial body that they reside. They will be fully controllable by artists too, I think that is a main priority when it comes to game content, I have the basics in place, but the full blown version of this you will see later, for the time being, I will just populate it with geometry first, a few dozen different types of trees and bushes and grasses, just to get the mechanics down pat for the placement and rendering of vast expanses of this stuff, all the while, keeping the RAM overhead to a bare minimum, well, that's the plan anyhow.

I'm actually going to go for a Pandora look for the interim, see if I can pull that off first, fully populate a planet with dense foliage all over, then from that, if it can handle that, I can back it off to a more natural look, open expanses with not much, and denser foliage in gullies and stuff like that, because after having it running fully and densely populated at first, means that it will handle just about anything you guys can throw at it, and know it won't fall over in a big heap, if you try to set it to a jungle planet.

Alex

[ultramedia]

Hi Alex,

Isn't the license for speedtree integration complicated?

[DavlexDesign]

G'day ultramedia,

Isn't the license for speedtree integration complicated?

I don't really know, But I know Steve (Sinbad) had made an integration library for Speedtree and Ogre3D, I haven't looked at it yet, as I was going to take a look when I had got a little further with this thing, apparently you only get a month to trial it before you have to buy it, both Steves' and the SpeedTree runtime library, so when I knew I could handle the placements correctly, then it would be just a matter of hooking it all up, so I could thoroughly test it for a good portion of the month trial, to see if it would work the way I need it to.

This still may be an option, and I like to keep my options open, I know sovaka is really keen on using SpeedTree, but we'll see how my setup works first, like I said before, I need to deal with allot of this stuff proceduraly, tree / foliage and prop generation, otherwise populating even a planet the size I have (I think I wrote the figures down in a previous post, the square kilometer area of this planet) to generate quite a diverse number of different foliage and prop geometries, and storing those for static loading without procedural generation techniques would be a ridiculous burden on the end users' storage devices.

I am working on a caching system, where the style of foliage will be stored in script form, and compressed that way, and when it's not needed, that data will be flushed from the cache, sort of like a web browser does for it's data, set a limit for cache size, so then when you re visit an area, that data may be still in the cache, or may not be, and have to be re generated. It's all in the process of being developed here, just another one of those distractions I use when I get tired of working on something else, or get stuck in a never ending loop, trying to solve some obscure problem that creeps in every now and then, you know, one of those little annoying bugs that you just can't seem to nail down, nothing critical, just some little thing that niggles at you, a jitter here, a crack in the geometry there, a click in the physics or animation, that sort of thing. I find it useful to just switch out to something else completely different, and take your mind off of it for a while, and usually (not always) a solution to the niggling problem rises up and presents itself.

I'm back on the foliage now, because I have an issue with the volumetrics of the clouds, hopefully I will come up with a solution to the volumetrics shortly.

Alex

[DavlexDesign]

G'day guys,

Here's the newest incarnation of the volumetric clouds planet wide ....

CloudsVolume_01.jpg

It's not very optimized just yet, but it is beginning to take shape.

Alex

[Zonder]

they don't look right but I can't put my finger on why which isn't very help full :S

maybe need to see the simulation running

[aguru]

Compared to http://www.ogre3d.org/forums/download/f ... &mode=view this looks much better in my opinion, yes there might be some details missing, but one no longer sees the "layeredness" of the approach. This was my main concern in the pics back then and I'm really happy you are tackling this now Great work, as usual ..

[aguru]

Oh and you might like these.

"The satellite, known as Elektro-L No.1, took an image from its stationary point over 35,000 kilometers above the Indian Ocean. This is the most detailed image of the Earth yet available, capturing the Earth in a single shot with 121-megapixels. NASA satellites use a collection of pictures from multiple flybys stitched together. The detail in the pic is just amazing."

[youtube]3hdyRh60R-Q[/youtube]

[vitefalcon]

Oh and you might like these.

"The satellite, known as Elektro-L No.1, took an image from its stationary point over 35,000 kilometers above the Indian Ocean. This is the most detailed image of the Earth yet available, capturing the Earth in a single shot with 121-megapixels. NASA satellites use a collection of pictures from multiple flybys stitched together. The detail in the pic is just amazing."

You really think that's real? I would call that particular day "The day when the earth stood still!". With no clouds moving in a whole day that's really weird claim don't you think? Maybe I'm missing something here...

[zootlewurdle]

Look harder. There is cloud movement. And consider the scale. Although I agree it doesn't look real.

[DavlexDesign]

G'day guys,

That video is interesting, didn't expect that kind of coloring on the planets land masses, makes me think I'm being too conservative with mine.
As for the clouds not looking real, they shouldn't at the moment, All that picture was meant to show, is that the system can produce volume planet wide, that is only the base volume, and I will be adding in the detail to this base volume with noise and other things, so don't be dis heartened looking at that pic.

All I was doing was experimenting with an idea on how to get full planetary volume clouds without too much layering being visible, I have yet to optimize it, so the frame hit you see in that shot will be allot less in the end (I HOPE). I'm still experimenting with this, Hopefully by the end of this experiment, I will have pretty nice looking, light scattered, self shadowing animated clouds.

That's the plan anyway.

The clouds you see are approximately 6km in height, and that will be tweakable via a shader parameter, because this data is still just a single grey scale height field, I'm trying to create a really nice algorithm to make these things look real, keeping to a very simple data source, which will make it easier to use a flow simulation to control the whole thing.

It is looking promising though.

Alex

[Zonder]

I think I know why they are looking odd to me its they aren't transparent enough in places and also don't have defined areas. I am thinking when I have been in planes and they seem that sort of height you can see lots or round masses and plumes. I think a picture was posted a while back.

I actually woke up at 3am with that thought lol

[Zonder]

Found a pic

[DavlexDesign]

G'day zonder,

That's what I'm trying to add in now (what's in that pic), Like I said before, the volumes in the pics prior are just the base volumes, and now I will try to add in the details to make them look like they should.
Thank's for the reference image, it will come in really handy.

Alex

[palaslet]

Alex,

About the vegetation... SpeedTree will definitively not work for populating a galaxy of planets (I doubt we'll ever see a pine or a spruce on an alien planet). You probably need a flora generator that can be preset with environmental parameters defined by the biome, from those parameters it can procedurally grow some trees and other flora stuff.

Any thoughts on that?

[LordBaal]

Oh and you might like these.

"The satellite, known as Elektro-L No.1, took an image from its stationary point over 35,000 kilometers above the Indian Ocean. This is the most detailed image of the Earth yet available, capturing the Earth in a single shot with 121-megapixels. NASA satellites use a collection of pictures from multiple flybys stitched together. The detail in the pic is just amazing."

You really think that's real? I would call that particular day "The day when the earth stood still!". With no clouds moving in a whole day that's really weird claim don't you think? Maybe I'm missing something here...

"..NASA satellites use a collection of pictures from multiple flybys stitched together..."

just saying...

About the foliage, I think mr. Palaset is right about the need of procedural generation of vegetable life on planets depending on it's characteristics.

[DavlexDesign]

G'day palaslet,

I thought I explained this before, about the foliage I mean, I will be proceduraly generating all of this stuff pretty much, building the foliage from base elements and primitives specifically designed for foliage generation. I mentioned speed tree for the fact that it may be an option if the system I'm developing ends up in the junk draw, I iterate, may be an option.

(I doubt we'll ever see a pine or a spruce on an alien planet)

Are you sure about that, I think that if the climate is right and the gravity and stuff, those trees would be there in some sort of form, the formation of all things is based on patterns. Those patterns are seen everywhere, under the sea, on the land, even out in space, patterns everywhere, and we are all made from the same stuff really, when it all boils down to it, so I can't see any reason that trees as we know them, would not be anywhere else in the universe. Granted, they may not be exactly the same, but still, they would probably be pretty similar. Take a look at some of the satellite images of Mars, looks like petrified forests on there, albeit much larger growths than on Earth, but then, less gravity would allow them to get that big. Even if they aren't trees, they look awfully close to what we have here.

Alex

[LordBaal]

Hi mr Alex

I think palaslet meant on another kind of planets. Earth like planets (same atmosphere, gravity and such) certainly could have similar vegetation, but a planet of ice with chloride atmosphere would be another history completely.

I'm wondering about your procedural method. Which parameters do you use? Gravity for the height and atmosphere color/composition for foliage color are obvious options (I think), but what else do you use on it. Humidity is another one that comes to mind.

[palaslet]

I think palaslet meant on another kind of planets. Earth like planets (same atmosphere, gravity and such) certainly could have similar vegetation, but a planet of ice with chloride atmosphere would be another history completely.

My thoughts exactly. Also the lighting condition might be different in other systems making green less optimal for photosynthesis.

Didn't mean to be a negative critic, just probing for more detailed thoughts about the fauna when populating different planets and how to accomplish a "alien-looking" fauna when the need arises.

Great work on the clouds by the way.

[Sovaka]

Actually it's quite scientific how it all boils down, so I won't even pretend I know the complete subject matter.

Photosynthesis isn't based off the colour of the leaves. Photosynthesis works from Chlorophyll which is what makes the green pigment in most leaves. It's best at using the blue to red electromagnetic spectrum's of our light. The poor absorption of the green spectrum is what gives Chlorophyll it's colour in plants.

Since scientifically that most atmospheres that support flora, fauna and liquid water would have an atmosphere like ours, most of those planets would also have near identical plant types.
I remember this being explained in some science journal when the question of varied validity of planet plant life in the Star Trek series was correct or not.

Since as we know it, plants only absorb carbon dioxide and expel oxygen. Maybe some other trace gases possible.
So on other planets that don't have a oxygen/nitrogen rich atmosphere, you'd find that there would be plant life at all like we know it. They would be more like sea sponges, coral and other types of organic structures.
You might even have silicon crystal type "plants" (I used the word loosely), serving the purpose of taking in oxygen and expelling helium or some other gas.

For us, we are going to try and make it as correct as possible, so you won't see any recognizable plants or plant structures on alien planets that do not support an oxygen/nitrogen atmosphere.

That being said tho, there isn't anything wrong with us taking certain liberties with that by changing the wave length of colour that filters through an oxygen/nitrogen atmosphere. Which would certainly change the flora and fauna's natural colouring

[LordBaal]

Really nice to know this....

I can't wait to put my greasy, candy filled hands on it!!! I'll use a napkin, I promise! Jajajaja....

[DavlexDesign]

G'day guys,

A bunch of stuff

Well said mate ... I second that !

As for building this stuff in geometry, well, I will be using primitives, as in shape descriptors for all different scenarios, and base material descriptors for the look of the skin / bark, also the leaves (if there are any) stuff like that. I will have an atlas of this stuff that I can pull from, and run a Fibonacci sequence over it, as well as a bunch of parameters based on makeup of the celestial body in question, to actually build the base geometries for each species of flora and fauna on a celestial body (if there is any flora or fauna on its surface).

I have a ton of stuff I'm working on in the back draws of my HDD, all distractions at the moment, while I work on the principal Celestial body make up, I really hope to nail the planet side of things shortly, then I can start off from a sound foundation, with all the integration of the physics and game play, nothing worse than building on top of shaky ground, things are bound to fall over in a big heap.

Well that's what I think anyway.

Alex

[UT2007]

Hi,

Your webpage has been suspended, "Error loading page. Please contact the support team. " and link redirect to "http://www.titaniumarts.com.au/cgi-sys/suspendedpage...."

I am sure community can contribute somehow, could you please share the codes please ? Or do you plan it to be a christmas gift

Regards