v1.3.2 final - Display Problems

[Seb]

Chris, Good work on the new version. Everything seems fine, and I dont seem to be experiencing the problems I've seen written in the forum, and previous bugs I had have gone, also general performance seems very good..

With one exception.. When viewing earth from Mars, there seems to be triangle holes appearing in the clouds.. Thought it might be because im using quite hi-res textures, but it does it on original textures also.




My system:
CPU:3.1ghz 512meg
GPU:nVidia MX440

[selden]

Yup.

It looks to me like holes are appearing in the Clouds texture. Presumably it's related to the other transparency problems.

Start Celestia
Let it goto Earth
Track Earth
GoTo Mars
Increase magnification until you can see continents and water.

Enable clouds

Observe lots of glitches come and go in the image.

Also, the position of the Earth isn't stable: tracking seems to be having problems holding the image steady.

System:
512MB 2.4GHz P4, WinXP Pro SP2
128MB FX 5200, Forceware v61.76
Celestia v1.3.2 final

[Seb]

Just noticed something else which was there before but didn’t realise it was a bug until I noticed how it should look. When viewing models, the textures are not shaded correctly most of the time, but when viewing certain models only from a specific angle it does display correctly.

Correct..


Incorrect..


My System:
CPU: 3.1Ghz 512m
GPU nVidia MX440
(latest drivers)

[selden]

seb,

This is a known bug related to MX graphics cards. It isn't clear if Chris can fix it in Celestia or if it'll require updated drivers from Nvidia. I suspect the latter.

Sorry.

[RND]

On mine the IIS solar panels shimmer. They did that in 1.3.1 also.

also i think the ring shadows on saturn are wrong. They seem way too big for me. for example:



If you type saturn in Googles image bank, youll notice how narrow the ring shadow is.

[selden]

Rnd,

Chris has posted about the ISS panel shimmer several times: they're due to a defect in the model that is included with Celestia. He's hoping to repair it in the not too distant future.

Since you have "Ambient Light" turned on, the ring shadows are drawn as if they wrapped around the planet. They don't really do that, of course. Turn off Ambient Light to see them as they should be.

The vertical widths of the shadows are exactly right, though. They vary as Saturn revolves around the sun. Remember that Saturn's rotational axis is tilted. As it travels around the sun, during Saturn's summer and winter one or the other of Saturn's poles is tilted toward the sun: that hemisphere is experiencing summer. At those times of year, the rings cast broad shadows on the hemisphere that's experiencing winter. During Saturn's spring and fall, the poles are tilted toward the side and the rings' shadows are almost invisibly narrow across the equator.

[RND]

i turned ambient off, but theyre still there

[selden]

When ambient light is off, the dark side of Saturn is black and the ring shadows cannot be seen there. If you can see ring shadows on the dark side of the planet, then ambient light is not off.

The ring shadows should go right up to the border of the planet's dark side, of course, so they look like half of a circle.

[John Van Vliet]

the problem with modles not rendering right with opengl is still there but when using only the nivida setings the models render fine

[selden]

the problem with modles not rendering right with opengl is still there but when using only the nivida setings the models render fine

Which OpenGL rendering problem do you mean?

Don't forget that there is a bug in recent versions of the video driver for Nvidia MX cards. I suspect that this is not something Chris will be able to do anything about in Celestia. It'll need a fix in the drivers from Nvidia.

[johnvanvliet]

yes it is the mx bug and it has been around for a year think ,oh well

[Seb]

Thanks John, I forgot I could change the render path. Seems to do the trick, though there is still some flickering on a couple panels on the IIS.

On other point, Celestia seems to occasionally crash when hidden behind other windows.

I just noticed that Celestia says Venus is 230k, is it not hotter than that?

Going of the subject even further, how do we know that our Galaxy (Milky way) looks like this? Looking at the positioning of all the stars they dont look like they would form part of this shape.

[selden]

Thanks John, I forgot I could change the render path. Seems to do the trick, though there is still some flickering on a couple panels on the IIS.

this has been found to be due to a problem in the model of the ISS that comes with Celestia. It'll probably be fixed in the next major release, especially if someone with some modelling skills helps. In the meantime, you might want to consider downloading one of Andrew Farnaby's ISS Addons. They're hosted on bh's site at http://www.berasan.com/celestia/iss.html.

On other point, Celestia seems to occasionally crash when hidden behind other windows.

I haven't had this problem. This suggests to me that it's most likely caused by bugs in the graphics drivers, not Celestia itself. If you haven't already, you should upgrade to the most recent drivers available from the manufacturer of your graphics card.

I just noticed that Celestia says Venus is 230k, is it not hotter than that?

You're probably thinking of Venus' surface temperature, which is extremely high. Don't forget that it's enhanced by the "greenhouse effect" caused by the deep atmosphere of mostly carbon dioxide. Celestia shows the "blackbody temperature." It's essentially the temperature that an object would have if it had no atmosphere.

Going of the subject even further, how do we know that our Galaxy (Milky way) looks like this? Looking at the positioning of all the stars they dont look like they would form part of this shape.

The distribution of many bright stars near the Sun does indeed seem to be slightly tilted relative to the plane of the real Milky Way galaxy. This is known as "Gould's Belt". See, for example, http://ottawa.rasc.ca/observers/1999/an9912p8.html

The overall band of the milky way drawn by Celestia matches fairly well the band that can be seen on the sky. This is because the orientation of Celestia's 3D model was adjusted to agree closely with the plane of the galactic coordinate system used by astronomers.

The overall 3D shape of the model isn't particularly realistic, though. For example, it only has two arms, while it's been established that the Milky Way has at least four arms. (This same shape is used for all of the spiral galaxies that Celestia draws. It'd be nice if there were some way to modify its shape with more arms, etc) Also, the local irregularities that Celestia draws are random and have nothing to do with the irregularities present in the actual Milky Way.

Does this help clarify things somewhat?

[Seb]

Does this help clarify things somewhat?

-yep. Thank you.