Animated planet files

Development of artwork, requests, suggestions, samples, or if you have artwork to offer. Primarily for the artists.
Message
Author
User avatar
pd
Graphics Lead Emeritus
Posts: 1924
Joined: Mon Mar 08, 2004 6:17 pm
Location: 52°16'N 10°31'E

#46 Post by pd »

geoff wrote:pd's theory about the angle matching the position of the star on the sidepanel doesn't really work because the angle is the same for all planets in a system, but the line from the planet to the star is different for each planet and changes as the planets are scrolled anyway.
That's not important. It just has to look as the sun at the top could illuminate the planets - it doesn't has to fit 100%. Humen are used to seeing planets like we've shown them on the sidepanel. I tried out how they look lit from above and it just looks odd to me.

As for the planet rings. They should be fine as they are now. They will be mapped on a plane, which is is then placed at the same position as ring should be. I'm looking forward how this will look in-game, especially if the rings are slowly rotating.

The city lights are a good start. I would try to integrate some really big structures made of dots - like overlapping rings or other geometric patterns.

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 13587
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#47 Post by Geoff the Medio »

pd wrote:
geoff wrote:pd's theory about the angle matching the position of the star...
That's not important. It just has to look as the sun at the top could illuminate the planets - it doesn't has to fit 100%. Humen are used to seeing planets like we've shown them on the sidepanel.
I'd think most people are used to seeing the moon wax and wane mostly left-right, not diagonally. Like these images.
I tried out how they look lit from above and it just looks odd to me.
I'm not suggesting lighting them from above; I want to light them from the side. A zero degree rotation axis tilt would then be oriented vertically, aligned with the shadow.
Image
As for the planet rings. They should be fine as they are now. They will be mapped on a plane, which is is then placed at the same position as ring should be.
I figured doing depth-sorting would be difficult if you intend the ring to be displayed as just a tilted sprite. The planet needs to be behind the rings in front of it, and in front of the rings behind it... Since AFAIK depth-sorting is usually done per-polygon, I figured it'd have to be done with separate polygons making up the ring shapme, rather than one polygon that slices through the planet model.
Image

User avatar
eleazar
Design & Graphics Lead Emeritus
Posts: 3858
Joined: Sat Sep 23, 2006 7:09 pm
Location: USA — midwest

#48 Post by eleazar »

Geoff the Medio wrote:I'd think most people are used to seeing the moon wax and wane mostly left-right, not diagonally. Like these images.
I tried out how they look lit from above and it just looks odd to me.
I'm not suggesting lighting them from above; I want to light them from the side. A zero degree rotation axis tilt would then be oriented vertically, aligned with the shadow.
It doesn't matter much to me, though having the shadow exactly at 90% feels slightly artificial.

tzlaine
Programming Lead Emeritus
Posts: 1092
Joined: Thu Jun 26, 2003 1:33 pm

#49 Post by tzlaine »

As for the planet rings. They should be fine as they are now. They will be mapped on a plane, which is is then placed at the same position as ring should be.
I figured doing depth-sorting would be difficult if you intend the ring to be displayed as just a tilted sprite. The planet needs to be behind the rings in front of it, and in front of the rings behind it... Since AFAIK depth-sorting is usually done per-polygon, I figured it'd have to be done with separate polygons making up the ring shapme, rather than one polygon that slices through the planet model.
Image

The ring textures are fine. In case you're curious, here's how depth is determined in this situation:

1) The planet shpere is rendered into a scene. When it is rendered over a certain pixel P in the buffer, this also causes a depth value to be written into the corresponding pixel P in the depth buffer.
2) The ring texture is rendered into the scene. Say we're rendering it to pixel P0, at which pixel it has a depth of Dt. If P0's depth value Db (read from the depth buffer) is greater (farther away) than Dt, the ring is rendered on top of P0. Otherwise, the old value at P0 (the color value for the planet sphere) stays, unaltered.

User avatar
pd
Graphics Lead Emeritus
Posts: 1924
Joined: Mon Mar 08, 2004 6:17 pm
Location: 52°16'N 10°31'E

#50 Post by pd »

So, I guess we need a way to mask out the atmosphere images. Do you have any ideas?

tzlaine
Programming Lead Emeritus
Posts: 1092
Joined: Thu Jun 26, 2003 1:33 pm

#51 Post by tzlaine »

pd wrote:So, I guess we need a way to mask out the atmosphere images. Do you have any ideas?
This is no problem. You just render the planet sphere as normal, then turn off writes to the depth buffer, render the atmosphere texture, then turn wrte to the depth buffer back on, then render the rings. This means that the effective depth of every pixel covered by the sphere is unaffected by the atmosphere.

User avatar
pd
Graphics Lead Emeritus
Posts: 1924
Joined: Mon Mar 08, 2004 6:17 pm
Location: 52°16'N 10°31'E

#52 Post by pd »

Sounds nice! I would love to see a screenshot of this :)

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 13587
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#53 Post by Geoff the Medio »

Regarding rings:
- What about shadows of the ring on the planet and vice-versa? Is this doable with plane-texture rings?
- Remember that they should tilt to align with planet rotation axis

Another nifty thing:
- Is there any chance we could get normal maps / bump maps for planet textures, instead of pre-rendered shadows on the textures? It would be excessively cool to have the shadows of mountains or craters on planet surfaces adjust and rotate and lengthen as the planet rotates...

Regarding terminator tilt:
eleazar wrote:...having the shadow exactly at 90% feels slightly artificial.
In what sense...? The idea is that the observer is seeing each planet while standing on the ecliptic plane, looking at the planet from a point not on the ray from the sun through the planet... perhaps on the planet's orbit somewhat ahead of it, looking back at it. The terminator would appear vertical. With the tilted terminator, it's not really clear where the viewer is... some arbitrary angle is chosen that has nothing to do with the ecliptic or the planets' rotation axes...

tzlaine
Programming Lead Emeritus
Posts: 1092
Joined: Thu Jun 26, 2003 1:33 pm

#54 Post by tzlaine »

Geoff the Medio wrote:Regarding rings:
- What about shadows of the ring on the planet and vice-versa? Is this doable with plane-texture rings?
Yes.
- Remember that they should tilt to align with planet rotation axis
That's the plan.
- Is there any chance we could get normal maps / bump maps for planet textures, instead of pre-rendered shadows on the textures? It would be excessively cool to have the shadows of mountains or craters on planet surfaces adjust and rotate and lengthen as the planet rotates...
I don't want to have planets that look like the cartoon planets in Spore (I like a more realistic look), and that's the sort of relief you'd need to actually see shadows from surface features. That said, if someone provides me with the bump maps, I can try to get this working, and we can decide then if we like it.
eleazar wrote:...having the shadow exactly at 90% feels slightly artificial.
In what sense...? The idea is that the observer is seeing each planet while standing on the ecliptic plane, looking at the planet from a point not on the ray from the sun through the planet... perhaps on the planet's orbit somewhat ahead of it, looking back at it. The terminator would appear vertical. With the tilted terminator, it's not really clear where the viewer is... some arbitrary angle is chosen that has nothing to do with the ecliptic or the planets' rotation axes...
I agree that it looks artificial. If you were in a system, on the ecliptic, with the ecliptic lined up horizontally in your view, you'd see the planets all tilted at different angles, not all lined up at 90 degrees to the ecliptic.

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 13587
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#55 Post by Geoff the Medio »

tzlaine wrote:
...normal maps / bump maps for planet textures, instead of pre-rendered shadows on the textures?
I don't want to have planets that look like the cartoon planets in Spore (I like a more realistic look), and that's the sort of relief you'd need to actually see shadows from surface features.
I'm thinking of this mostly for the barren planets, which already have visible shadows on craters. This wouldn't cast long rotating dramatic shadows, but would just cause craters to have dark and light sides consistent with the sun locaction and the planet's rotation.
I agree that it looks artificial. If you were in a system, on the ecliptic, with the ecliptic lined up horizontally in your view, you'd see the planets all tilted at different angles, not all lined up at 90 degrees to the ecliptic.
The planet would be tilted differently, but the shadows would all be aligned vertically. The shadow angle is my point. I think you, and maybe others, are confusing rotation axial tilt and shadow angle, or at least what I'm talking about when refering to them.

To clarify: Shadow angle is the direction the shadow points. Axial tilt is the difference in angle between the normal to the ecliptic plane and the axis about which the planet rotates. Planets all have different axial tilts, but all have the same shadow angle if in the same ecliptic plane.

See image:
Image
What we have now is the middle. The axis (yellow line) and the shadow (cyan dot arc) angles are both tilted as seen by the viewer, and the axis angle is aligned to the shadow (the yellow line connects the ends of the cyan dot arc). The shadow arc indicates the plane of the ecliptic because the sun is also in the ecliptic plane, so the arc angle of the shadow follows the angle of the ecliptic relative to the viewer's "up" direction. If arc and axis are so-aligned, it means that all the planets physically have 0 degree axis tilt angles (relative to the eclicptic) but are just displayed tilted.

To the left, the axis angle and shadow angle are different. The viewer's camera is aligned to the axis angle, so it appears vertical and the shadow is tilted. If we did this (planet axis is vertical) consistently, then all planets would have different shadow angles. This would probably be universally disliked.

To the right, the axis angle is tilted, and the shadow is vertical. This is the one in which the viewer is standing on the ecliptic plane, looking at the planet, which has a rotation axis tilted with respect to the ecliptic plane. I like this one. It's easy to see the planet axial tilt (different for each planet) relative to the vertical shadow and screen "up" direction.

User avatar
pd
Graphics Lead Emeritus
Posts: 1924
Joined: Mon Mar 08, 2004 6:17 pm
Location: 52°16'N 10°31'E

#56 Post by pd »

I feel like talking to a wall... I told you why the shaddow is the way as it is now.

User avatar
eleazar
Design & Graphics Lead Emeritus
Posts: 3858
Joined: Sat Sep 23, 2006 7:09 pm
Location: USA — midwest

#57 Post by eleazar »

I just realized Celestia (also released under the GPL) has texture maps in the same format that ours are. I don't want to perfectly replicate Sol's members, but it's nice to have something real to work from. (Yeah i realize these basic files could be found on NASA, but searching NASA is a big pain.)

Edit: actually it looks like a lot of Celestia's source files can be found here.

They also have a few bump maps, which should be great for testing a bump map feature if that's desired. At our current sidebar magnification i don't think bumpmaps would make much difference, bit if the planets are ever rendered larger, it would be nice.


I don't see why so much energy should be spent discussing the angle of the night sides of planets.
• It has no effect on gameplay, or the development of future graphics.
• It's trivial to change to your preference by editing a text file.
• The difference between the current angle and Geoff's preference wouldn't be noticed by 95+% of players.
Last edited by eleazar on Wed Sep 27, 2006 9:59 pm, edited 1 time in total.

User avatar
Geoff the Medio
Programming, Design, Admin
Posts: 13587
Joined: Wed Oct 08, 2003 1:33 am
Location: Munich

#58 Post by Geoff the Medio »

eleazar wrote:I just realized Celestia (also released under the GPL) has texture maps in the same format that ours are.
FO art is under the CC-SA, not the GPL. Are these compatible?
At our current sidebar magnification i don't think bumpmaps would make much difference, bit if the planets are ever rendered larger, it would be nice.
Again, it's really only for barren, which *do* have visible craters with clear direction to the shadows...
Image
I don't see why so much energy should be spent discussing the angle of the night sides of planets.
I nitpick somewhat indiscriminately, and some things just bug me for some reason.

User avatar
eleazar
Design & Graphics Lead Emeritus
Posts: 3858
Joined: Sat Sep 23, 2006 7:09 pm
Location: USA — midwest

#59 Post by eleazar »

Geoff the Medio wrote:
eleazar wrote:I just realized Celestia (also released under the GPL) has texture maps in the same format that ours are.
FO art is under the CC-SA, not the GPL. Are these compatible?
Err, i'm used to working on an all GPL project. I don't know if they are compatable, legalese isn't my thing. Anyway The bump maps can still be used for testing.

User avatar
eleazar
Design & Graphics Lead Emeritus
Posts: 3858
Joined: Sat Sep 23, 2006 7:09 pm
Location: USA — midwest

#60 Post by eleazar »

pd wrote:The city lights are a good start. I would try to integrate some really big structures made of dots - like overlapping rings or other geometric patterns.
I agree, i just want to see them in-game so i'm not working blind.

Post Reply