Another preview - cloud shadows

[selden]

This shot shows some lines that are along 180 west. Are they known? They disappear in a different rendering mode with shadows off.

I see them with medres moving clouds. They seem to be fixed relative to the surface texture and don't move with the clouds. They're also visible when cloud rendering is disabled. My guess would be that they're actually related to the new specularity feature -- perhaps an artifact due to a problem with how the edges of the surface texture and/or specularity maps are handled. Boundary effects are always a pain.

[cartrite]

Concerning the lines at 180 west,

Chris wrote:

Oops! That was a silly oversight. I just checked in a fix for this problem. If you update the to latest CVS build, the cloud shadows will go away when cloud maps are disabled.

The cvs version I'm using is about an hour and a half old now. Chris stated earlier that he commited a fix for shadows that didn't disappear after clouds were toggled off. You may have the older version were the shadows remained? My version turns the lines off when the clouds are toggled off even in OpenGL. Or the rendering mode changes and the clouds are on.

[cartrite]

I just turned on the cloudspeed. And your right Selden. The lines are in a fixed position relative to the cloud movement. But there is a shift to the left or right (depending on time forward or backward) as the specular highlight pases over the area. If that means anything to anyone.

cartrite

[chris]

This shot shows some lines that are along 180 west. Are they known? They disappear in a different rendering mode with shadows off.

I see them with medres moving clouds. They seem to be fixed relative to the surface texture and don't move with the clouds. They're also visible when cloud rendering is disabled. My guess would be that they're actually related to the new specularity feature -- perhaps an artifact due to a problem with how the edges of the surface texture and/or specularity maps are handled. Boundary effects are always a pain.

It's related to cloud shadows, not the new per pixel specular lighting. Cloud shadows involve rectangular to spherical coordinate transformation, which means that the vertex shader uses an arctangent function. The discontinuity in arctangent results in the line along the 180 meridian. I'm working on a fix.

--Chris

[Chuft-Captain]

I should have put that on the list too. I recently fixed one depth sorting bug that Selden had alerted me about. But, if you're talking about depth sorting of transparent objects, that's a ways off.

I'm happy to wait for as long as it takes if it's still high on your list of priorities. (Actually I've no choice in the matter as I'm not in a position to fix these bugs myself )
JOOI: What's the technical hurdle that makes the transparent depth-sorting more difficult to fix than Selden's bug?

It's a fundamental limitation of all currently shipping graphics hardware. In order for translucent objects to be rendered correctly, the triangles must be sorted so that the ones furthest from the viewer are rendered first. You can do the sorting on the CPU, but for a moderately detailed model this will be very slow. There are no practical workarounds that completely solve the problem. Only partial fixes exist.

OK, that makes sense. Thanks for the explanation.
Damn!

Secondly, to which category do the problems I reported to you earlier in the year belong to? (Do you remember those?)
Do you think those problems will be resolved by the bug-fix you mention above, or are they in the category of transparent object depth sorting bugs?

I'll have to look at your bug report again; there's a good chance that if it didn't involve translucent models, your problem may have been fixed by my change.

I think this was in a PM to you. (I can resend if you can't find it). My problems DID involve models with transluscent parts, however the DS problems occurred between opaque surfaces of the model.

BTW. Do you have any rough gut feeling for the likely "performance hit" of solving the DS problems as compared to the performance hit of the rendering improvements in your recent threads? Is it likely to be more or less of a performance drag than these other improvements?

It's a more complicated discussion than I have time for right now But I do have a very good understanding of all the performance implications.

That's OK. I'm sure your time is better spent coding, than discussing. I would assume that it's not orders of magnitude more costly than the other changes anyway.

PM me if you need me to resend that bug report.
Cheers

[fsgregs]

Chris:

Just a short note. The cloud shadows look great.

Edited: I just found the answer to my question originally posted here about including shadows being cast on spacecraft by planets below. Thanks for planning to include it in the next pre-release.



Frank

[Cham]

How do I turn ON cloud shadows, with Celestia 1.5.0 ? And what is the SSC commands for Cloud bump maps ?

In the atmosphere code, I tried "CloudShadows true", "CloudShadow true", "Cloudshadows true" and "Cloudshadow true", none worked.

[selden]

Cloudshadows are supposed to be always on. There is no SSC control over them.

Unfortunately, it looks like they've stopped working.

[Cham]

Selden,

are the shadows "broken" on your system too ?

and what about bumps ?

[selden]

And there are no cloud bumpmaps, only
CloudNormalMap

[Cham]

Where can I find a CloudnormalMap file, so I can test it ? I'm unable to create one yet.

Please, be precise : are the shadows broken for you too ?

[t00fri]

How do I turn ON cloud shadows, with Celestia 1.5.0 ? And what is the SSC commands for Cloud bump maps ?

In the atmosphere code, I tried "CloudShadows true", "CloudShadow true", "Cloudshadows true" and "Cloudshadow true", none worked.

You first got to assign a key via a celx script like so

Code: Select all

-- start.celx

function slowgoto()
    sel = celestia:getselection()
    obs = celestia:getobserver()
    if sel and obs then
        obs:goto(sel, 15)
    end
end

function togglecloudshadows()
    t = celestia:getrenderflags()
    t2 = {}
    t2.cloudshadows = not t.cloudshadows;
    celestia:setrenderflags(t2);
end

keyhandlers =
{
   G = slowgoto,
   V = togglecloudshadows,
}

function handlekey(k)
    handler = keyhandlers[k.char]
    if (handler ~= nil) then
        handler()
        return true
    else
        return false
    end
end

celestia:registereventhandler("key", handlekey)

-- end


Once you run that script (e.g. by making start.celx the default startup script in celestia.cfg, then you toggle the cloud shadows by pushing SHIFT+V.

There is also a slowgoto binding which you may try out.
It's simple, isn't it

Oh yes: note cloudshadows ONLY work in the OpenGL 2.0 path!


Bye Fridger

PS: Note the cloudshadows still produce artefacts (bands in a certain distance...)
So don't report these. We know that already

[selden]

Correct: cloud shadows aren't visible on my system. Some recent code change must have broken them.

[selden]

Fridger,

Then something has changed.
It was the case that cloud shadows were on by default. I've never had to toggle them before.

*sigh*

It'll take me a while to test your script.

s.

[t00fri]

Fridger,

Then something has changed.
It was the case that cloud shadows were on by default. I've never had to toggle them before.

*sigh*

It'll take me a while to test your script.

s.

Here is what Chris wrote already in march:
(Dev list)

This morning I checked in a change that adds a new render flag for
cloud shadows. This flag is off by default, so you'll no longer see
cloud shadows when you start up Celestia. At the moment, there's no
way to turn them on via the UI. I think we've exhausted all the
keyboard bindings, and I'm not sure it that toggling cloud shadows
deserves it's own key binding anyhow. You can turn on cloud shadows
with a celx script, however.

Bye Fridger

[Cham]

I just tried the script. The slowGo option works (I'm not sure it will be usefull, however).

And the cloudshadow doesn't work properly. I get some color garbages around the screen, the ocean becomes deep blue, and no shadows.

[selden]

Fridger,

Thanks for the reminder. I now vaguely do recall that posting. I guess my assumption was that the toggle would show up soon in the GUI menu (it hasn't) , so I just forgot about it

Cham,

Fridger's CELX script works fine. Copy it to a .CELX file, double click on its icon, and then cloud shadows will toggle on and off when you type a Capital-V.

[t00fri]

Chris:

Just a short note. The cloud shadows look great.

Edited: I just found the answer to my question originally posted here about including shadows being cast on spacecraft by planets below. Thanks for planning to include it in the next pre-release.



Frank

Sorry, Frank,

I am not Chris, but I'll answer nevertheless

Personally, I am a bit surprised that you find the present cloud shadows so nice. There are SERIOUS artefacts (that Chris is aware of) repeating in a certain distance. They get increasingly disturbing the higher the resolution.

Didn't you spot these?

Next, I think with the /present/ cloud textures, the cloud shadows look pretty unnatural! So did you make new, appropriate cloud textures already? Let me know...

Bye Fridger

[Cham]

Fridger's CELX script works fine. Copy it to a .CELX file, double click on its icon, and then cloud shadows will toggle on and off when you type a Capital-V.

I know it works, I've tried it (the slowGo option works). But the capital-V gives me some colorfull garbage around the window, the oceans becomes dark ("deep blue"), and no shadows. And Celestia becomes very slooooow !

[t00fri]

I just tried the script. The slowGo option works (I'm not sure it will be usefull, however).

And the cloudshadow doesn't work properly. I get some color garbages around the screen, the ocean becomes deep blue, and no shadows.

Did you switch on the OpenGL 2.0 render path???

The script certainly works for me since JULY 2006!

Bye Fridger