blackrack

I will need your kopernicus and scatterer configs for this, I can not know like this.

You can run this without EVE/SVE and with the ocean shaders turned off to have a much smaller impact on performance.

He means from the mod main menu. Anyway, for this bug I thought I found a new way to fix it without having the alternate shader toggle in the main menu, guess it's not working for everyone, I will put everything back the way it was for next release. Sorry I can't really see what's going on, perhaps more pictures and description would help. You're right, it's centered on one point in the middle of the sun, it's the target if the raycast that checks wether or not there is something in front of the sun. Anyway, the problem is I'd have to do a bunch of raycasts at different points in the sun and average the result to get that effect, or I would have to find a way to render the sun surface with a replacement shader then downscale it somehow to a single pixel which gives the "occlusion" ratio. neither of these solutions seem very elegant to me so i just left it as it is. Personally I don't like stock's fade out animation as usually with timewarp on I can see the flare going well inside occluding objects.

Please don't keep reporting this issue with the new version, I haven't updated the mod for it yet. Opengl or dx11?

Btw there's a test version in the scatterer thread with some new water effects.

btw @Sigma88 I just remembered, my original idea for the shader was to have a different front face and backface texture for the ring, so as to be able to have this effect where a faint portion of the rings is only visible from the back and where some sections look different from the back: front: back: The idea was to have the shader read the texture in one direction (eg horizontal) and depending on whether it's the front face or the back face use one half or the other of the texture (eg top or bottom), to store both textures in the same file. In fact I think I put this in the shader code but the alpha value was still being read from the front face texture only so it's like it's not there.

Looks like someone modified the ring scale but didn't pass the right parameters to the ring shader. Looks like for the inner radius it's correct but not the outer one. I will take a look when I can. Edited: I think @Thomas P. found the problem and did a commit which fixes it.

Like I said, it's caused by the scatterer shadow distance setting and how unity handles shadows.

I've noticed this with the small parachute as well, will fix when I can. Do you have some actual issues in the game apart from 2 nullrefs showing up in the log? I don't know, this is too vague, try removing some of your mods until you find the issue. 1. Remove all installs of EVE, SVE, etc 2. Install this http://forum.kerbalspaceprogram.com/index.php?/topic/149733-122-environmentalvisualenhancements-12-2/& 3. Install latest scatterer and enable "EVE cloud integration" from main menu 4. Done The issue here is NOT caused by the near plane but by how the unity shadow system handles the different cameras and the total distance covered by each camera. In the stock game the shadow distance is set to a really low value so you can see it, in scatterer the shadow distance is set to 100 Km or so and that "dilutes" the shadows on the far camera, I'm not sure what the exact issue here is though. If you lower the scatterer shadow distace (it's somewhere in the config file, the original value is something like 5 Km (5000 in the .cfg)) it should about look the same but you will lose the ability to see terrain shadows from afar (which imo add much more to the look of the game than consistent shadows up-close).

http://s1.webmshare.com/GyaEb.webm Although it might look here like it works, what I tried was a hacky solution, it basically moved the water level height for every part between frames. The issue with this is that it doesn't interpolate the forces correctly between frames, so that for every part, the water level doesn't appear to go up and down smoothly so that they bob along with it, instead it disappears instantly and reappears at a different level, parts just fall down and slam into the water surface and ships break. It was unstable and very prone to kraken attacks and crashes, and didn't look realistic once you tried to move the ship over the water surface. My conclusion from this experiment is you need to write your own buoyancy engine, something like FAR but for water physics. Personally I haven't researched this much and I have no knowledge about this kind of thing but it doesn't seem like a simple task. Perhaps @ferram4 would know where to start. I personally think KSP is the right platform for this, maybe not immediately (and I would prefer to see a new terrain system and a few updates to the graphics before) but I definitely think more attention to planetary bodies and the planetary exploration aspect of the game would help it a lot.

I will need to look at an example when I get home as I have no idea about most the issues you're describing. Is there any config in particular that illustrates these problems?

Don't change the main radius (Rg) in thr config tool, what matters are the ratios or Rt/Rg and Rl/Rt, generate your atmo and once you get ingame it will rescale it to fit the planet, basically just use the default radius, all the other settings are fine to change in the config tool.

If anyone knows how to consistently reproduce this bug it would help a lot, I can never reproduce it when I want to which makes it hard to debug.

If I remember correctly, I used the nvidia gfexperience overlay once and it was "intercepting" the F10 F11 button presses. Anyway, if you reinstall the mod make sure you overwrite the config file as well, do it manually to be sure, I have no experience with CKAN and can't comment on it.

I didn't even realize the old shader read the diagonal, I'll take care of it.

Is this in the preview version? I had the feeling it fixed it.

Well, in dx9, but there might still be tricks. Anyway, shadows should be possible technically but unity doesn't like the projected grid and it makes the shadows flicker on all the other objects when I enable them on the ocean. Not sure why atm.

Is the stock atmo visible in this screenshot? I can't really tell. In any case I'm pretty much done with what I wanted to add to the ocean and I'm going to dedicate some time to fix bugs and refactor the code. Btw the ocean shader is at the limit of what I can fit in a dx9 shader atm. It might be difficult in the future to make it work with planetshine without breaking refractions. I will probably get back to you when I get around to bug fixing.

Yes, same reason. You could try removing AA and downsampling instead but I haven't tried that in KSP in a while. I should probably make the depth-based transparency togglable for this reason. No they're not supposed to be visible, all good. I did a small cosmetic tweak on the extinction tint but don't remember consciously changing mieG, thanks for testing.

No hurry though, whenever you have time

So I'm in need of some testers for the new effects. Things to keep in mind for this (preview) version: -New effects include: Water refraction and depth-based transparency, contact foam and underwater surface and "fog", there are some new parameters in the ocean tab to control them -If you disable refraction, underwater effects may not work, this will be fixed. -Ocean lights compatibility will break all the new effects, disable it for now. This will be fixed. -Sunflares still render underwater, this will be fixed. -From orbits around 70-160km, when you have refractions on, you might see outlines around the shore, this will be fixed. -I haven't made configs for laythe or EVE oceans so they'll get the default values. -The setting for alternate SQRT is removed, it should not be needed anymore, please report if you still get issues with the "leaking halo" around planets. -There used to be a very obvious line around 300m in front of the camera where the postprocessing effect starts, this is fixed. -I haven't tested with EVE or any other mods, I expect the clouds might be visible through the underwater surface, please tell me if this is the case. -NearClipPlane parameter is back in the config file, I don't recommend messing with it. -When the camera is just at the water surface level there may be some artifacting, a bit lower or higher everything should work well. Report any new issues and let me know what you think of the new features. Have fun https://mega.nz/#!KBgjhQgY!A3ccBD_YyURyanNrmu9i2DTNXuM3-KKQ8-Wu11Z7PwM

That's just me forgetting to update the version number. Although i might have released a bugfix or two which are bundled in that SVT version (can't remember what they were frankly).

Sorry I thought you had the white ball not the purple ball? Can I see a screen and a log? Also try starting in dx11 or opengl if you haven't already. It's nothing that elaborate, at the moment it's just checking the depth buffer to see how close the surface is behind it. ie it doesn't leave a trail after a ship has moved, so no such a thing as decay or collider checking or anything. You can see it here around the base of the plane: https://vid.me/zP0x If the plane moves there appears to be a bit of foam around where it touches the water but it will not leave a trail, that's a bit more complicated.

i haven't tried the newest version of EVE yet (in fact didn't know a new one was released). The fixed files you see above are for a very old release (september), you don't need them for the newer versions. The current scatterer cloud integration is meant for the version of EVE released in october (second one in the downloads above). Maybe get that one or disable the cloud integration from scatterer and see if it helps. You don't need the fixed dlls

Thanks, I've learned a lot in the two years I've been working on this mod. The best part now is playing other games and being able to see why certain technical choices were made or how a certain effect is achieved. All in due time.