Oolite Bulletins

.....It's fulla stars!.....

No seriously, it looks lik an UFO.

LittleBear wrote:

Is this still controlled by the sky_blur values?

The generation of the sky hasn’t changed at all, only the texture and blending mode. So that shouldn’t be affected.

Commander McLane wrote:

Actually it's a picture of a star-cluster, not an elliptic galaxy (the original is here), and it's not trimmed at all.

Hmm… painting out some of the stars around the edges first to get a roughly elliptical shape, before blurring, might work better. I may play around with that one.

TGHC wrote:

Is this the sort of view you would get all the time, and is this suggesting what travel through "real" space would be like.

The screen shots are from worlds with relatively large numbers of blobs, with a large colour variation; they haven’t changed from the previous release. This is not in any way, shape or form a simulation of what space really looks like. :-)

Uncle Reno wrote:

Looks good, the only thing that I would say is that the colours are too bright IMO.

The colours haven’t changed either, although the blending mode has, so overlapping bits of different colour will tend towards white. I may experiment with desaturating the colours or limiting the palette somewhat.

*checks* Oh, wait, it is rather different… because of a bug in texture handling, the alpha channel for blobs (and anything else with an alpha channel) was being applied twice in earlier releases, at least under OS X. Hmm. Given this, reducing the brightness somewhat may be a good idea… Hmm. Actually, the new code was simply ignoring the alpha colour component (which is normally multiplied with the texture alpha component). Fixed.

Right. Having fixed the over-brightness, here’s the same nebula (in Leesti, one jump from Lave, pull back a little when you jump in) in 1.65:


…and in fixed 1.69 code:


…and in unfixed 1.69 code with one of Charlie’s textures (you’ve seen this one before):


I like the bright one. <:-( But the ability to tweak sky_blur_alpha in planetinfo.plist remains, and the whiter centres created by the new code will look better when it’s high. This is still more efficient, and will still benefit from texture variation, so keep ’em coming.

Just because I can, the same nebula with a 1.0 blob brightness in 1.65:


…and 1.69 using the default texture:


The planetinfo.plist used:

The bright ones look much better.

Will they remain constant though (generated acc. planetseed) or will they be generated de-novo, each time a system is populated?

...
Any chance of having a generic 'milkyway' backgroundhaze? (self.galatic plane)

I suggest keeping them the same - what planet to you know that has a sky that changes every week?

I know of only one such planet.....Raxxla.

Arexack_Heretic wrote:

Will they remain constant though (generated acc. planetseed) or will they be generated de-novo, each time a system is populated?

The fact that I've shown five screen shots of the same nebula taken in different versions of the app should be a hint. ;-) The geometry and colour generation has not changed at all, and is still constant for each system.

Arexack_Heretic wrote:

Any chance of having a generic 'milkyway' backgroundhaze? (self.galatic plane)

Possibly, but not until after 1.70.

Similar to one of Commander McLane’s, but based on this image which isn’t cut so close:


I’m not especially pleased with it, but it does demonstrate the point about not obviously being cut to a square.

@ Ahruman: Could you just post some screenshots of these starclusters (yours and mine, if you don't mind) in the game? In my taste they might work even better than the galaxies posted before, because they're not so defined.

You liked in-game-screenshot of my first attempt (thanks for that ), but I actually wasn't convinced so much, because it's so obvious that it's a picture of an elliptic galaxy which is just repeated. And why should you have nebulas that look like c&p'd galaxies?

But I guess the starclusters, if copied and pasted close to each other, will just remain what they are: starclusters, just bigger, which seems more logical to me. But I can't check it.

Perhaps--if it's that what gives you a headache about them--I could try to manipulate it so it looks less symmetrical and therefore less artificial or clinical.

Sure.




Yours nicely demonstrates the “squares” problem (and possibly gives an appreciation of how blobs are generated).

For reference, what I did was:

• Trim off some of the fat, leaving a generous border.
• Adjust levels so that the background was reasonably close to black.
• Paint out most of the background stars, producing a roughly circular pattern.
• Blur.
• Crop close to final area.
• Apply a new levels adjustment as an effect layer, with a radial-gradient mask, to ensure black edges.
• Crop away the black bits.

If I’d thought it through properly, I’d have avoided the first level adjustment.

I find this to give me a nice subtle look in 1.68, but very hard to capture on a Screenshot...

however im unsure how this will look when it gets brightend up...

I find it a particulary sort of feeling in my lower regions to avoid
repetiveness..

This texture should be a throw into texture directory and go fly and have a look.

Here's one originally from NASA:



Distribution under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 license and the GNU General Public License version 2 permitted.

The overlapping nebulae:

can they be only one texture, or can several different textures be overlayed?

If mixing of textures is possible, repetitiveness will be significantly reduced.
Except for very recognisable textures, offcourse.

The idea is for each instance to have a randomly-assigned texture. So yes, multiple textures will be used in a given blob.