[Release] System Features: Rings

[gsagostinho]

Version 2.5 is now available at the extension manager and it should fix the issue.

@RockDoctor could you please test it and report back here?

[RockDoctor]

@RockDoctor could you please test it and report back here?

Choices - dealing with a game issue, or remembering a guy who risked his life to save mine, and who now has a funeral.
DIE Thargoids !!!!!!!
(Damn - now got to go back and get the version. Which I've forgotten what I'm looking for. Sorry, tonight isn't the night.)

[RockDoctor]

Oh, right, updated version. I'll get it just now, but got to do some work now. Probably report back tonight.
Before your update, I was getting some very weird effects. https://app.box.com/s/50ndbpswqc5mi492lnliu6a7sra8ghak
(Is it possible to directly display images from Box? I can't see an obvious option.)

[RockDoctor]

Beats working.
Zip file of images - looks great (but I do see the Moire effects other people have mentioned.) https://app.box.com/s/8a4hqtlze3f6j7cavwvcxjmxd7scj4m1
I've updated your "rings", and a number of other OXPs.

[Norby]

Is it possible to directly display images from Box?

No, they do as much as possible to prevent this, even the url in the preview is a temporary one. Use postimages.org direct link instead.

[gsagostinho]

Zip file of images - looks great

Great to know it is working now!

but I do see the Moire effects other people have mentioned.

Well, you will have some Moire effect with this work, there is no way around it (at least with the algorithm I am using). You can take a look on the previous pages for the whole discussion, but notice that the Moire is worse the lower is your resolution (and from your screenshots it seems your screen is fairly small, isn't it?). But I really don't think you would have less Moire effect with the version 1.x on this same system, at least on my side the Moire patterns are less strong than before, even though they are still obvious. If anyone can come up with some specific suggestions (preferably some concrete lines of code) on how to improve this OXP, I would more than happily adopt it!

[RockDoctor]

Small screen? Well, it's a laptop, so it's got what it's got. OTOH, my first computer had a 640x480 graphics card but a 1024x768 monitor, and I do remember days of trialling different mode lines to get X working on Yggdrasil. I'm just glad I "acquired" a better graphic card from work (head office sent it out, but no one knew what to do with it, so I filed it into a drawer in the technicians workshop for several months before it got "lost". Beatles reference, "number nine", or something like that).
When did I last have a monitor? Pre-2010. Maybe pre-2003, the last time the burglars came visiting. I don't see that happening again. The monitor, not the burglars.

[gsagostinho]

Small screen? Well, it's a laptop, so it's got what it's got. OTOH, my first computer had a 640x480 graphics card but a 1024x768 monitor

I meant a screen with low resolution, as your images are 861 by 589 pixels long and I assumed they were close to your native resolution. Laptop screens can actually have very high resolution, I am also a laptop user (don't own a desktop) and I have 1920x1080, which is quite good. Anyway, my point is: the lower the resolution, the more Moire patterns. On my side it looks acceptable and way better than Rings v1.x, but on your computer it's your call what looks better.

[RockDoctor]

as your images are 861 by 589 pixels long and I assumed they were close to your native resolution.

Oh, sorry, of course.
No, that's just the size of the window I was running Oolite in. I'd have to figure out ... a process to find out what my actual screen size is. While [...]ing, I figured out. The Pause-F2 screen has a setting for whether to run the game in a window or full-screen, and it gets a figure from somewhere for what that screen size is. [Fires up] While that fires up, I'm trying to remember if the system I parked up in had any rings. Ohh, messages. Medicine run.
So, yeah, the maximum screen size that Oolite detects is 1366x768 (bloody widescreen fad ; it'll pass next time the TV industry wants to sell a new round of TVs). It's been a while since I tried setting up a virtual desktop - I could never bother remembering the keyboard mumbo jumbo to scroll the view pane left, right, whatever. Alt-tabbing between tools is more than adequate. But I could probably set up a virtual desktop if it would help testing.

I'll switch Oolite to full screen for a change, and find a ring.
OK - now I know why I never fullscreen Oolite. But yes, the fringing is much reduced at full ("native"?) resolution for the screen.

[gsagostinho]

that's just the size of the window I was running Oolite in.

Yes, if you are on windowed mode then of course it's the resolution of the window that dictates the severity of the Moire patterns.

So, yeah, the maximum screen size that Oolite detects is 1366x768 [...] the fringing is much reduced at full ("native"?) resolution for the screen.

Great to hear, I had it tested at 1366x768 before and the results look quite fine to me, nice to know it's also looking fine on your side. Cheers.

[gsagostinho]

Hi all, version 2.6 has just been added to the extension manager. This new version manly does two things:

- first, it attempts to solve the Moire pattern issues. I realised that I could simply "fade out" the higher frequencies of the colour gradient according to the distance, so that the further we are from the planet the less details we see (the high frequencies disappear), thus diminishing/solving the Moire pattern issues.
- secondly, I have added some shading to the granular structure when very close to the rings. This is the result:











I would really appreciate some feedback, in particular by people with low resolution screens and low end machines. I am particularly interested in feedback concerning 1) the Moire patterns or lack of them, 2) performance difference from previous version and 3) my selection of fading distances for each frequency of the radial colour gradient.

[another_commander]

Very nice work. My observations following a quick test:
- Performance is not affected compared to previous versions. It was never a problem for me with this OXP though in any case and despite my low-end machine.
- Moire: Vastly improved. There are still some hints of its presence, depending on viewing angle and distance (see pic and note that it was shot at the exact angle where the effect is more intense. The effect practically disappears as soon as I move slightly). Right now, I would say that it is not an issue of concern anymore; I can easily live with it as it is.

[gsagostinho]

Very nice work.

Thanks, and thanks for testing this!

My observations following a quick test:
- Performance is not affected compared to previous versions. It was never a problem for me with this OXP though in any case and despite my low-end machine.
- Moire: Vastly improved. There are still some hints of its presence, depending on viewing angle and distance (see pic and note that it was shot at the exact angle where the effect is more intense. The effect practically disappears as soon as I move slightly). Right now, I would say that it is not an issue of concern anymore; I can easily live with it as it is.

Great to know there was no performance hit, I was afraid that the modification added too many calculations per cycle. As for the Moire patterns, it's also very nice to know it is not an issue any longer! I will still see if I can think of some improvements to the algorithm so that it would take not only the distance but also the viewing angle in consideration for the fading of frequencies.

I also plan to improve the close up view of the rings, so that the brighter is a lane the denser in rocks it would be.

Thanks again!

[cbr]

I also plan to improve the close up view of the rings

Perhaps you could insert some orbiting rocks/asteroids moving in some of the lanes too...



some anaconda's moving over the rings.

[gsagostinho]

Perhaps you could insert some orbiting rocks/asteroids moving in some of the lanes too...

I actually wanted to do something like that as that would be nice, but the problem is that in order to have a nice effect I think we would need to add a lot of orbiting rocks. It might also be difficult to match their position to the rings, as the shader is what decides their size (as far as Oolite is concerned, the object behind the rings is simply a single gigantic plane cutting the planet in two).

some anaconda's moving over the rings.

Very nice