Page 5 of 7

Posted: Sun Sep 24, 2006 4:39 am
by tzlaine
I have to say, I can't really even tell a difference. If other people can, though, I can easily make this change in the official data.zip. Thoughts, pd?

Posted: Sun Sep 24, 2006 12:29 pm
by eleazar
Also, as has been mentioned previously, it can be easy to loose the explorable stars against the nebula backgrounds. I've tweaked the nebula graphics to be less domineering. I also removed the obvious stars from the nebulae, since these could be confused with the explorable stars.

The image is an animated GIF, for better compare/contrasting:

EDIT: animated Gif removed, becuase it wasn't helpful, and blinked annoyingly. :shock: Try the files yourself.

The current version is slightly prettier, but i believe my modification will be easier to play with.

you can get the files from here:

Ultimately i'd recommend that distinctive nebulae (like the horsehead) should appear only once per galaxy, if at all. A simple XML file in which the probability and repeatability of each nebula graphic could be set would do the trick.

Also the paralax effect would work better if the nebulae scrolled slightly slower than the explorable stars, (i.e. were slightly behind). This would help sustain the illusion of depth in regions were the nebulae are thick.


Where does one find the official "data.zip"?

Posted: Sun Sep 24, 2006 12:56 pm
by Geoff the Medio
http://freeorion.sf.net/data.zip

The whole nebula generation system needs some tweaking. They are currently much too large, especially on small galaxies.

I don't follow why you want the nebulae to scroll separately from the stars though... They presumably part of the same galaxy, so would be on the same plane and would move together.

But regardless, IMO the whole parallax scrolling thing is a bad idea to start with... It makes it harder to see what's going on with "real" stars, it makes no sense astronomically, and it doesn't even look very good.

Posted: Sun Sep 24, 2006 1:31 pm
by eleazar
Geoff the Medio wrote:http://freeorion.sf.net/data.zip
Thanks
Geoff the Medio wrote:...I don't follow why you want the nebulae to scroll separately from the stars though... They presumably part of the same galaxy, so would be on the same plane and would move together.

But regardless, IMO the whole parallax scrolling thing is a bad idea to start with... It makes it harder to see what's going on with "real" stars, it makes no sense astronomically, and it doesn't even look very good.
I have to disagree. IMHO, Paralax make it easier to distinguish the foreground, interactable galaxy from the eyecandy. If there's anything in the background, besides pure black, the paralax makes it easier to differentiate. I do agree that more could be done to distinguish the background stars from the explorable ones. I also think paralax looks cool, 8) especially if you switch the graphics as previously suggested.

A usable galaxy map is compelled to make "no sense astronomically". It's esentually 2D after all. I wish to put the nebulae on different layer primarily because that will make it easier to distinguish the gameplay content. I'm considering the nebulae as eye-candy, and an aid to help you orient yourself in the galaxy. Therefore from the perspecitve of gameplay, it is background. It can be more intuitively recognized as background if it seems futher away.

Posted: Sun Sep 24, 2006 2:37 pm
by pd
I like the nebulae to be as intense as they are now. It's a nice colorful galaxy this way, and not a dead boring part of open space. I think theay should be part of the star layer. In future they could even have effects on the stars they cover. Nebulae could also work as 'fog of war'.

I also like the paralax effect a lot, and it looks great in my opinion. It could need some tweaking and use some bigger images to prevent the tiling effect. I would even add some layers(like 10). To me, it adds a lot of depth to the galaxy map.

Posted: Sun Sep 24, 2006 9:28 pm
by Geoff the Medio
eleazar wrote:IMHO, Paralax make it easier to distinguish the foreground, interactable galaxy from the eyecandy. If there's anything in the background, besides pure black, the paralax makes it easier to differentiate.
I want to remove the background entirely, not make it non-parallaxed.
A usable galaxy map is compelled to make "no sense astronomically". It's esentually 2D after all.
To me, a flat galaxy is a reasonable approximation that is self-consistent. The infintely-large planes of "background stars" don't make any sense, however. Are these stars part of the same galaxy as the forground? If so, why do they extend infintely far in every direction on the sides? Are these real stars, or something different? How are they different from the forground stars and why don't they affect gameplay?

A perhaps-better alternative would be to have a few layers, in front and behind the layer of gameplay stars, of either diffuse semi-transparent gas or faint star shapes, which move around such that they 3D shape of the galaxy is consistent (though it would still play as a 2D galaxy). Instead of infinite planes of stars in all directions, there'd be stars that move around consistent with them all being part of the same galaxy. Off to the sides outside the galaxy, there'd be no parallax stars, and in the centre there'd be several layers that move at different speeds to convey their depth to the player, but which remain self-consistently located relative to eachother. The speeds of the different layers would have to adjust for the map's zoom level to remain consistent... so it would end up being like doing an axis-aligned 3D rendering of several overlapping partly transparent textures with differnet depths. Alternatively, if it's just faint stars and not diffuse gasses, they needn't be confined to discrete layers, and could be randomly distributed throughout the depth of the galaxy.
pd wrote:In future [nebulae] could even have effects on the stars they cover. Nebulae could also work as 'fog of war'.
Right; even though they're presently just eye candy, eventually nebulae could be strategically important features of the game map.

Posted: Mon Sep 25, 2006 12:53 pm
by eleazar
pd wrote:I like the nebulae to be as intense as they are now. It's a nice colorful galaxy this way, and not a dead boring part of open space. I think theay should be part of the star layer. In future they could even have effects on the stars they cover. Nebulae could also work as 'fog of war'.
Please try my nebulae in-game. After the fact, i don't think my preview is very good.

I have trouble thinking of an effect nebulae could have on this type of game. Yet, if it were of strategic importance, the current type of graphics are lousy indicators, since they fade out so gradually, it will be unreasonably hard to say if many planets are "in" or "out".
pd wrote:I also like the paralax effect a lot, and it looks great in my opinion. It could need some tweaking and use some bigger images to prevent the tiling effect. I would even add some layers(like 10). To me, it adds a lot of depth to the galaxy map.
Rather than larger images, i think it would be better to have a few randomly mixed star-images per layer. That would better fight the uniformity. I'd also limit all background star sizes to 1 pixel, and indicate distance, by having the futher layers generally dimmer.

Geoff: It sounds like what you want is a much more complicated, less homogeneous paralax view.

Posted: Tue Sep 26, 2006 12:02 am
by Geoff the Medio
I have trouble thinking of an effect nebulae could have on this type of game.
- Limiting travel speed (differently from normal) or preventing travel until appropriate tech is researched
- Preventing use of certain weapons / defenses / misc. ship parts
- Hiding fleets within them
- Damaging ships that spend time in them
- Nest for space monsters
- Being a source of a resource for players... eg. set up a Nebular Gas Refinery in a system inside the nebula, get supply of Verteron Particles or Metaphasic Anti-Omega Radiation used to unlock / speed production of a particular building / ship part
Yet, if it were of strategic importance, the current type of graphics are lousy indicators, since they fade out so gradually, it will be unreasonably hard to say if many planets are "in" or "out".
Then they'll have to be replaced. Maybe some volumetric fog could be generated instead.

Edit: If the nebulae were generated procedurally, then we could do some fancy stuff like lighting them, and having them cast shadows, by the star lights near them. Then if we were really bored, we could have nebulae generated and exapanding visible as the turns progress after supernovae, or new stars be created within existing nebulae... /Edit
eleazar wrote:Geoff: It sounds like what you want is a much more complicated, less homogeneous paralax view.
More complicated, yes. I'm not sure what you mean by "homogenous", though in a sense yes, as they're confined to the volume of the galaxy and react properly to zooming, rather than spreading out infinitely in all directions for no logical reason.

Really I'd prefer no cluttering and confusing background stars at all, as the map is much cleaner without them at all zoom levels, though if there have to be some, they might as well make some sense.

Posted: Wed Sep 27, 2006 3:53 am
by eleazar
It's been previous noted that the game stars can be hard to distinguish from the background stars when zoomed out. Not that things were horrible, but they could be a little better. I've tweaked the starfield graphics to look less like the game stars. As you scroll around the repeating patterns are less obvious with these files. These pixel sized stars make the screen look sharper, IMHO, and the starfield deeper.

Image
comparison at maximum zoom out

get them here.


Earlier i read (somewhere i can't find now) someone suggesting that the game star graphics be separated into 2, a small core that doesn't scale, and a halo that does scale. I feel that might look a little better, but it probably less important than a lot of other things. However i could provide some graphics if tzlaine wants to mess with it.

Edit: By the way, tzlaine, and option in the preferences to "reload graphic files" would be a great aid in producing/improving the graphics. Or possibly better to reload all the assets, though changing the rules of the game in the middle might cause a crash.

Posted: Wed Sep 27, 2006 5:29 am
by Geoff the Medio
eleazar wrote:I've tweaked the starfield graphics to look less like the game stars. As you scroll around the repeating patterns are less obvious with these files. These pixel sized stars make the screen look sharper, IMHO, and the starfield deeper.
Definitely fixes the confusion issue, though there are still a few obvious repeating pattern bright stars, particularly on the front (slowest scrolling) layer. I suggest making the front layer less densely packed with stars than the middle or back layers. It might help a bit... and look a bit better, as a realistic 3D depth projection of a uniform volume of stars would appear to have them packed more densely per unit screen area on the more distant planes.
Earlier i read (somewhere i can't find now) someone suggesting that the game star graphics be separated into 2, a small core that doesn't scale, and a halo that does scale. I feel that might look a little better, but it probably less important than a lot of other things. However i could provide some graphics if tzlaine wants to mess with it.
Someone posted an example of multiple independently-rotating layers for stars. I think a bit of animation would look pretty good... I'd suggest trying that before the core / halo distinction... or perhaps try to do both.

Also, even if tzlaine doesn't feel like doing something, I might give it a try (though possibly after some other higher priority items on the TO DO list...)

Posted: Wed Sep 27, 2006 1:34 pm
by eleazar
Geoff the Medio wrote:
eleazar wrote:I've tweaked the starfield graphics to look less like the game stars. As you scroll around the repeating patterns are less obvious with these files. These pixel sized stars make the screen look sharper, IMHO, and the starfield deeper.
Definitely fixes the confusion issue, though there are still a few obvious repeating pattern bright stars, particularly on the front (slowest scrolling) layer. I suggest making the front layer less densely packed with stars than the middle or back layers. It might help a bit... and look a bit better, as a realistic 3D depth projection of a uniform volume of stars would appear to have them packed more densely per unit screen area on the more distant planes.
True, but a realistic 3D depth projection would also take into account that more distant stars are less likely to be bright enough to be seen. And i agree with the project's stance that realism is not a game virtue. However i added more dim stars to the back layers, and it looks a little better. The previous link now contains the new images.
Geoff the Medio wrote:Someone posted an example of multiple independently-rotating layers for stars. I think a bit of animation would look pretty good... I'd suggest trying that before the core / halo distinction... or perhaps try to do both...
Can you link me, or are the images gone?

Posted: Wed Sep 27, 2006 2:25 pm
by Geoff the Medio
eleazar wrote:True, but a realistic 3D depth projection would also take into account that more distant stars are less likely to be bright enough to be seen. And i agree with the project's stance that realism is not a game virtue.
The point isn't to realistically simulate a three-plane scrolling starfield... Rather, it's that a better sense of depth and reduced obviousness of the repeated starfield pattern might be / is achieved with more densely packed stars in the back rows. Most people don't actually scroll around space fast enough to see differential star parallax, though they do drive next to fields of crops and see how the closer ones move faster and are less densely packed per solid angle than the slower-moving far away ones...
However i added more dim stars to the back layers, and it looks a little better.
I concur, though now it seems like there's just too many stars... they sort of blur together into "noise". That said, it looks a bit more like the title splash screen starfield now as well... so maybe that's good...? Or maybe not... as I'd rather err on the side of less noise / cleaner in-game presentation (ie. fewer discrete stars).
Geoff the Medio wrote:...multiple independently-rotating layers for stars.
Can you link me, or are the images gone?
Check the link at the bottom of this post. And actually, much of the thread is probably relevant and/or otherwise worth your reading.

Posted: Fri Sep 29, 2006 2:28 am
by eleazar
Geoff the Medio wrote:
Eleazar wrote:
Geoff the Medio wrote:...multiple independently-rotating layers for stars.
Can you link me, or are the images gone?
Check the link at the bottom of this post. And actually, much of the thread is probably relevant and/or otherwise worth your reading.
I've created a new thread to discuss improving the playable stars here.
Geoff, give these a shot if you want.

Posted: Sat Sep 30, 2006 7:17 pm
by LithiumMongoose
Geoff the Medio wrote:
I have trouble thinking of an effect nebulae could have on this type of game.
- Limiting travel speed (differently from normal) or preventing travel until appropriate tech is researched
- Preventing use of certain weapons / defenses / misc. ship parts
- Hiding fleets within them
- Damaging ships that spend time in them
- Nest for space monsters
- Being a source of a resource for players... eg. set up a Nebular Gas Refinery in a system inside the nebula, get supply of Verteron Particles or Metaphasic Anti-Omega Radiation used to unlock / speed production of a particular building / ship part
In MOO2 (and possibly MOO3, I don't remember) nebulae prevented shields from working in combat. MOO2's "Hard Shields" ship component allowed you to bypass this restriction (along with other benefits). Star Trek consistently has nebulae both making shields not work and greatly hampering sensors.

Yes I know we're in no way shape or form trying to base our content on existing games/franchises, but I like to point this stuff out when possible since I personally have great respect for these games/franchises. :) That being said I think everything in Geoff's list is good.

Posted: Sun Oct 01, 2006 3:43 pm
by eleazar
How does the game engine use nebulae anyway? I can't find a text file with the graphic's names. Does it use any graphics with the name "nebula#.png" ? Does it take into account the size of the nebula graphics when figuring out how many a galaxy needs? I.E. would smaller nebula graphics cause the engine to place more in the galaxy and visa versa?

This isn't a brainstorming thread, but i will point out that most of those ideas IMHO don't mesh with the game mechanics already laid down. For instance, ships are either in transit along a star-lane (presumably a different dimension) or in a star-system. There's no lurking off to the side hidden in a nebula.