improving nebulae and background stars

[eleazar]

Yeah the option to turn this stuff on and off seems reasonable. You might also tie it in to stuff like the star halos, parallax background, and nebulae (unless they are given a game effect)

I've uploaded some new gaseous graphics back to 245x256.
screenshots would be appreciated, since i'm working blind. I've changed some stuff since the last ones that were this size.

[Geoff the Medio]

See attached...

Edit: Note that these are with 500 stars. FPS are higher with smaller galaxies. /Edit

[Geoff the Medio]

In response to SVN revision 2437...

- The tiny black holes seem a bit overdone
- The blueness of blue stars could be more consistent
- Non-tiny blue and white stars are hard to distinguish
- Red, orage and yellow can be hard to distinguish in general

[eleazar]

Yeah for skinny starlanes

In response to SVN revision 2437...

I presume you mean that these comments are on the same subject as that commit— not that the commit created these issues? Well, except for the black hole, which used to be too hard to see in the wrong situation, but i did overcompensate on that last commit.

- The tiny black holes seem a bit overdone
- The blueness of blue stars could be more consistent
- Non-tiny blue and white stars are hard to distinguish
- Red, orage and yellow can be hard to distinguish in general

In general the level of saturation needed for the tiny sizes would be obnoxious at closer zooms. So while we can tweak things to make it work better, there's going to be a certain amount of noticeable disconnect between some of the zoom steps as long as:
1) we only use two image sizes, and
2) the stars look good at min and max zoom

Solutions:
• I've been staring at the halo star graphics long enough that it probably time to make another revision. The stuff i have in mind would tend to help.

• Whatever zoom level is shown on the left there should use the tiny stars rather than the shrunken full-size graphics. I.E. the tiny graphics should be used for one additional zoom level. They have ended up a little bigger than my original mock-up.

• If you want to put in some extra effort to make the transition smooth, i'd be happy to make graphics of intermediate saturation to go to work in the low-middle zoom range.


P.S. the red star in your screen-shop has a fully opaque pixel in the lower right corner, but that is not in SVN can you try getting it again from the trunk?

[Geoff the Medio]

I presume you mean that these comments are on the same subject as that commit— not that the commit created these issues?

The comments are on what is seen in game as of that commit.

Regarding the blue stars looking white, the issue seems to be that the blue star disc is rather unsaturated and large. The more-saturated blue halos don't show up around the larger disc until closer-in zooms, while for smaller discs, the halo appears around them more prominently at further-out zooms.

P.S. the red star in your screen-shop has a fully opaque pixel in the lower right corner, but that is not in SVN can you try getting it again from the trunk?

That's some weird quirk with the rendering, not the image itself. There apppears to be one red star that doesn't have it, and none of them do at the next higher zoom in. It's probably specific to my system, although hopefully someone else can confirm that.

[eleazar]

Regarding the blue stars looking white, the issue seems to be that the blue star disc is rather unsaturated and large. The more-saturated blue halos don't show up around the larger disc until closer-in zooms, while for smaller discs, the halo appears around them more prominently at further-out zooms.

Yeah, that's what's happening. Unfortunately it's also why the blue stars look better IMHO than the others when zoomed in.

[eleazar]

See attached...

Edit: Note that these are with 500 stars. FPS are higher with smaller galaxies. /Edit

Regarding the zoomed-in screenshots, did you change the interpolation method? Or were the screenshots taken as JPGs before you saved them as PNGs?

The squarish artifacts that i saw previously is no longer present in these screenshots.

[Geoff the Medio]

Regarding the zoomed-in screenshots, did you change the interpolation method? Or were the screenshots taken as JPGs before you saved them as PNGs?

The squarish artifacts that i saw previously is no longer present in these screenshots.

Saved direct to PNG from screen capture. I didn't change any interpolation, but when did you see these artifacts and what had them...?

[pd]

This one has some artifacts. They seem to be gone on the latest screenshots. It's most notable in the upper part of the picture.


I second, that the skinny starlanes look great.

[eleazar]

OK, i did some relatively thorough fake testing, and created a new set of gaseous graphics. These should look a bit more chaotically natural, and not present such a homogeneously even surface where the stars are dense.

As always a screenshot would be appreciated.

Please note: Previous images were delivered at full opacity which required the transparency to be increased for the screenshots shown above. I don't know if Geoff did that with a graphics program or in the code, but these images should not need any such adjustment.

[Geoff the Medio]

My graphics programs are pretty much limited to MS Paint, which doesn't have any capacity to do transparency.

Screencap with SVN 2452 attached. Code (now) uses unaltered transparency of input image; previously alpha was adjusted by code as described in previous post.

[Redcap]

My graphics programs are pretty much limited to MS Paint, which doesn't have any capacity to do transparency.

Check out http://www.getpaint.net

[eleazar]

Screencap with SVN 2452 attached

I'm quite pleased.

There's more fine-tuning i could do, but that would probably be more profitably done once i get a release with this feature in my hands.

The last thing i'd like to see is the size of each gaseous sprite decreased about 25%. I think that should help give a smoother gradient between the dense and diffuse parts of the galaxy... and of course help minimize whatever jaggies lurk in the close zooms.

P.S. After trying to use some other pictures as models, i ended up using this galaxy picture as inspiration.

[Geoff the Medio]

The last thing i'd like to see is the size of each gaseous sprite decreased about 25%.

attached

[eleazar]

The last thing i'd like to see is the size of each gaseous sprite decreased about 25%.

attached

Yeah, i think the tighter definition is good.
just committed revised files to bring the over-all opacity back up, and to bump up the color variation.