Problems and issues with Celestia 1.2.4 under Windows XP Pro SP1
- Matrox G400 Max (latest stable WHQL drivers)
- Bumpmap is not working or is not noticeable (perhaps correct behavior?)
- Moon seems to be sitting sometimes next to its orbits
- ATI Radeon 9700 Pro (latest drivers)
- Saturn's texture is misplaced: clouds bands are perpendicular to the plane of the rings instead of being parallel.
- Saturn's texture exhibits odd behavior when zooming in and out near on of the poles.
- Same thing sometimes happens with Neptune's texture
- It's impossible to active pixel shader
- It's impossible to activate vertex shader
- Bumpmap is not working or is not noticeable (perhaps correct behavior?)
- Moon seems to be sitting sometimes next to its orbits
- General issues
- Though they are supposed to be far less noticeable than Saturn's rings, there are also rings around Jupiter and Uranus, there are partial rings around Neptune too.
- When you zoom out too much, the Milky Way does not stay at the center of the screen, but keep appearing around the corners of the display. This is probably a number approximation problem.
- Milky Way is barely seen on screen: not enough brightness (perhaps correct behavior?).
- On 1280x1024, 1280x960 and 1152x864 at least (upper and lower resolutions not tested) if Celestia's window is maximized and then the program is closed, at next startup the window is no longer maximized though it's size is nearly the size of the display. There is a small empty margin between the right border of Celestia window and the border of the display.
- Even if we do not have direct surface picture of the surface (though we do have a sonar scanner of Venus thanks to Magellan), shouldn't the textures of Venus, Jupiter, Saturn, Uranus and Neptune be clouds textures instead of surface textures?
- Remarks
- It would be interesting to put in both Pioneer and both Voyager drones (and their trajectories as well if possible).
Celestia 1.2.4, Win XP Pro SP1, G400 Max, 9700Pro
-
Topic authorbouye
-
Don. Edwards
- Posts: 1510
- Joined: 07.09.2002
- Age: 59
- With us: 22 years 2 months
- Location: Albany, Oregon
Ah!
The joys of using none Nvidia display cards. As it is quite well known by most people here in the forum I will pass the bad news to you. All Radeon cards have problems of one kind or another in Celestia. Yours however are exceptional. Please give us some screenshots of Saturn! As for the Matrox card that also is going to be a problem. I dought Chris will do any serious work for that now outdated card. He is working on making Celestia work much more friendlier with the ATI Radeon cards though. He said something about being around version 1.2.6 or 7 or somewhere in there. Sorry for the vagnous of the info but this is a very complex program and trying to support every single video card would be next to imposible for Chris to implement. Computer games don't even support every card on the market. You must make your purchase on what you read and what word of mouth you here and make the best choice you can. You made the coice of the ATI Readeon over the NVidia GeForce family of cards and now you can see were a big choice can put you in the position you can't access all the cards capabilities under certain software programs. You might see if you can't find at least a cheap Nvidia GeForce2 based card(they can be had for as low as $40 US) to see bumpmapping and Spectural lighting effects. Of couse a GeForce3 or better will give you the most outstanding results. Sorry for the bad news.
The joys of using none Nvidia display cards. As it is quite well known by most people here in the forum I will pass the bad news to you. All Radeon cards have problems of one kind or another in Celestia. Yours however are exceptional. Please give us some screenshots of Saturn! As for the Matrox card that also is going to be a problem. I dought Chris will do any serious work for that now outdated card. He is working on making Celestia work much more friendlier with the ATI Radeon cards though. He said something about being around version 1.2.6 or 7 or somewhere in there. Sorry for the vagnous of the info but this is a very complex program and trying to support every single video card would be next to imposible for Chris to implement. Computer games don't even support every card on the market. You must make your purchase on what you read and what word of mouth you here and make the best choice you can. You made the coice of the ATI Readeon over the NVidia GeForce family of cards and now you can see were a big choice can put you in the position you can't access all the cards capabilities under certain software programs. You might see if you can't find at least a cheap Nvidia GeForce2 based card(they can be had for as low as $40 US) to see bumpmapping and Spectural lighting effects. Of couse a GeForce3 or better will give you the most outstanding results. Sorry for the bad news.
Bouye,
One of the advantages of Celestia is that you can use whatever texture files you want for most of its objects. I seem to recall reading a thread discussing using appropriate cloud and surface maps for Venus. The forum's "search" page should be able to find it.
The lack of rings on some of the planets is not due to problems in the hardware vendor's OpenGL library that Don mentioned, but due to a lack of information provided in Celestia's solar system definition file (solarsys.ssc in the data directory) It's a plain-text file.
No rings are defined for Jupiter or Neptune.
However, if you look *very* closely, you will see that Uranus does indeed have rings. They're just extremely faint, as they should be.
I'm sure if someone were to provide appropriate ring declarations and texture files Chris would be glad to include them.
My understanding is that the trajectories for the Pioneer and Voyager probes will be available as add-ons at the time of Celestia's next major release.
One of the advantages of Celestia is that you can use whatever texture files you want for most of its objects. I seem to recall reading a thread discussing using appropriate cloud and surface maps for Venus. The forum's "search" page should be able to find it.
The lack of rings on some of the planets is not due to problems in the hardware vendor's OpenGL library that Don mentioned, but due to a lack of information provided in Celestia's solar system definition file (solarsys.ssc in the data directory) It's a plain-text file.
No rings are defined for Jupiter or Neptune.
However, if you look *very* closely, you will see that Uranus does indeed have rings. They're just extremely faint, as they should be.
I'm sure if someone were to provide appropriate ring declarations and texture files Chris would be glad to include them.
My understanding is that the trajectories for the Pioneer and Voyager probes will be available as add-ons at the time of Celestia's next major release.
Selden
-
Topic authorbouye
After trying various ATI stable and beta drivers without success, it appears that the Saturn texture bug is fixed when using DirectX 9 RC0 with current ATI bet DirectX 9 compatible drivers. Saturn is now correctly display and does not effect other planets rendering anymore*.
*Note side-effects of this bug : since I had some free time to travel in the solar system, so I quickly found the odd side-effect this bug has on other celestial bodies.
Initially all other planet/moon than saturn are displayed correctly.
If you go to Saturn, the texture mis-alignement problem appears as previouly discribed.
But then if you go to any other planet you'll see that the textures are not correctly mapped on the surface:
- texture is always facing the carema (texture does not rotate with planet).
- when you zoom in and out the planet the texture exhibits some oddbehavior as it is zoomed in and out as well.
- no more clouds or nightlights.
Trying to explain that better now:
- go to earth, earth is displayed correctly, and behaves normally if you zoom in/ou or rotate around
- go to Saturn, which is not corectly displayed.
- go back to Earth, you are now facing let's say Antartica, no matter how many degrees and in which direction you rotate around the planet, you are always facing the Antartica texture from the same point of view.
- zoom in, you are still facing Antartica but the texture has been zoomed in as well.
If you zoom too much, you'll arrive to the point where one texture pixel fills the entire half-sphere/hemisphere you're seeing as if you were looking at it through a lens.
- similar thing when zooming out.
All other planets except Saturn behaves like this.
- go back to Saturn, the planet still has it's own texture bug...
*Note side-effects of this bug : since I had some free time to travel in the solar system, so I quickly found the odd side-effect this bug has on other celestial bodies.
Initially all other planet/moon than saturn are displayed correctly.
If you go to Saturn, the texture mis-alignement problem appears as previouly discribed.
But then if you go to any other planet you'll see that the textures are not correctly mapped on the surface:
- texture is always facing the carema (texture does not rotate with planet).
- when you zoom in and out the planet the texture exhibits some oddbehavior as it is zoomed in and out as well.
- no more clouds or nightlights.
Trying to explain that better now:
- go to earth, earth is displayed correctly, and behaves normally if you zoom in/ou or rotate around
- go to Saturn, which is not corectly displayed.
- go back to Earth, you are now facing let's say Antartica, no matter how many degrees and in which direction you rotate around the planet, you are always facing the Antartica texture from the same point of view.
- zoom in, you are still facing Antartica but the texture has been zoomed in as well.
If you zoom too much, you'll arrive to the point where one texture pixel fills the entire half-sphere/hemisphere you're seeing as if you were looking at it through a lens.
- similar thing when zooming out.
All other planets except Saturn behaves like this.
- go back to Saturn, the planet still has it's own texture bug...
Bouye,
It's good to know that they've gotten rid of a lot of bugs. From what you write, it sounds to me like the low-level drivers that they're updating for DirectX are the same ones that are used by their OpenGL library, so fixing bugs for one API fixes bugs for the other.
Do I also understand you to mean that the texture bugs you've so carefully described are now gone?
It's good to know that they've gotten rid of a lot of bugs. From what you write, it sounds to me like the low-level drivers that they're updating for DirectX are the same ones that are used by their OpenGL library, so fixing bugs for one API fixes bugs for the other.
Do I also understand you to mean that the texture bugs you've so carefully described are now gone?
Selden
-
Topic authorbouye
Yes they are gone with DX9 RC0 and DX9 beta drivers from ATI.
I'll report here again if the bug appears with final DX9 and official drivers, and well, with newest version of Celestia if it does appear again.
Though not perhaps not as clear as I meant it to be (english is not my native language), I described the effects that occurs using DX 8.1 and various flavours of regular/supported and beta/unsupported ATI drivers, "just in case* to leave a record".
*It's better to report a pb than saying nothing.
I'll report here again if the bug appears with final DX9 and official drivers, and well, with newest version of Celestia if it does appear again.
Though not perhaps not as clear as I meant it to be (english is not my native language), I described the effects that occurs using DX 8.1 and various flavours of regular/supported and beta/unsupported ATI drivers, "just in case* to leave a record".
*It's better to report a pb than saying nothing.