New Celestia-1.4.0pre-FT1.2 Version for Download

[Boux]

Looks good here
Rendering is OK from all angles (for my own taste).
I will leave comments about the shape itself to the pros.

[Rich]

I think I'm either seeing a bug, or need to modify my add-ons for this version. Galaxies/galactic clusters add-ons appear over bright. Turning nebulae on reduces the problem for some reason, but even then things seem too bright.

Might this have to do with...
"
Suppressed apparent magnitude display for galaxy add-ons, lacking
specification of an absolute magnitude (absMag) input value.

Future DSO add-ons should preferrably specify an AbsMag value!
"
?

Me too:
If I use the "Celestia 1.4preFT1.2ish (11-16-05)" app only with my existing celestia stuff, I get the over-bright galaxies. They look really cool when zoomed in, but from Earth they're way over bright on my system (i.e., large pure white blobs)...
If I use the "Celestia 1.4preFT1.2ish (11-16-05)" discrete complete package, the glalaxies look 'normal'.

Is there supposed to be a 'happy medium"...?

-Rich

Latest System: Mac G4 Dual 1.25GHz / OSX 10.4.3 / / 1gbRAM / nVidia GeForce4 Ti 4600

[brokfn]

Hi,
The SBb template looks great, but I can't say wether it is scientifically accurate.

I've got an enlightment problem with some non spherical asteroids,
2003 EL 61 1 for example.
I set ambient light to none, then I twist the asteroids in all directions, and in some position, the shadow disappears entirely, leaving the asteroid brightly illuminated.
Has someone else the same problem ?

15 really ? You look quite younger !

Bye
Brokfn

[t00fri]

I think I'm either seeing a bug, or need to modify my add-ons for this version. Galaxies/galactic clusters add-ons appear over bright. Turning nebulae on reduces the problem for some reason, but even then things seem too bright.

Might this have to do with...
"
Suppressed apparent magnitude display for galaxy add-ons, lacking
specification of an absolute magnitude (absMag) input value.

Future DSO add-ons should preferrably specify an AbsMag value!
"
?

Me too:
If I use the "Celestia 1.4preFT1.2ish (11-16-05)" app only with my existing celestia stuff, I get the over-bright galaxies. They look really cool when zoomed in, but from Earth they're way over bright on my system (i.e., large pure white blobs)...
If I use the "Celestia 1.4preFT1.2ish (11-16-05)" discrete complete package, the glalaxies look 'normal'.

Is there supposed to be a 'happy medium"...?

-Rich

Latest System: Mac G4 Dual 1.25GHz / OSX 10.4.3 / / 1gbRAM / nVidia GeForce4 Ti 4600

Please, Rich,

I know that in the US "stuff" is a popular word for many things, but in our context of hunting potential bugs it is absolutely useless!

Instead of saying that

with my existing celestia stuff, I get the over-bright galaxies.

we NEED the name of one concrete add-on and its URL (Motherlode?), for which in your system the overbrightness happens! How should we otherwise locate what you denote with "stuff"?? It might actually make a big difference whether we talk about clusters or galaxies for instance. Many little details may be crucial for locating the bug! Please try a little harder...
What costs you only 1 or 2 precisely formulated report sentences may cost me 3 hours (!) LESS searching....

Same @Kolano!

Galaxies/galactic clusters add-ons appear over bright.

WHICH one?? I have to reproduce the effect before I can fix it. So far with many add-on tests and all OS versions, I have NEVER seen galaxies as pure white overbright blobs.

Bye Fridger

[jestr]

Dont know if this helps here,but I have the same problem with overbright galaxies,but only with some of my galaxy addons,where because they were made with older versions of Celestia in mind I combined a 3ds mesh (with hubble photo texture),defined as a nebula,and also included a galaxy definition using the old galaxy templates.I have found if I delete the galaxy part then the problem disappears.Cheers,Jestr

[Rich]

Sorry about that unspecific _stuff_ I posted earlier.

The last celestia full package I installed was the "Celestia FT1.1 " version. So, the "Celestia 1.4preFT1.2ish (11-16-05)" I just installed is using resources from the FT1.1 version.

When I use the "Celestia 1.4preFT1.2ish (11-16-05)" app only; the galaxies A2657, A44, A76, A102.... are over-bright; it almost seems that all the A galaxies are like this. When I use the above pre FT1.2ish celestia version as a "full package", (isolated from the full FT1.1 package) then I have no overbright galaxies - I cannot find any "A" galaxies.


Jestr - I'll look into what you mention in your post, Re: overbrightness- it seemed to me it may be something like that for me as well.

Thanks,
-Rich

[jestr]

I suspect it is where you have some addons installed which use the old galaxy templates,the A named ones sound like a Selden galaxy catalogue
(galaxy clusters?).Try moving them out of the FT1.2 folder and see what happens,Jestr

[t00fri]

...
When I use the "Celestia 1.4preFT1.2ish (11-16-05)" app only; the galaxies A2657, A44, A76, A102.... are over-bright;
Thanks,
-Rich

Rich,

I am even more confused: What is the FT1.2ish ending supposed to mean in practice? Really we can only follow up bug reports arising from proper installations of the full FT1.x distributions. Please don't forget I did many many tests with many many galaxies and add-ons and could NOT see anything you seem to describe. So in case there is something buggy nevertheless, we can only spot it if we all cooperate and try to describe precisely what we are seeing.

+++++++++++++
From your Axyz galaxy names (the origin of the respective data file you again did NOT cite!) I can see that these must be Selden's old galaxy data, and the A-galaxies are entries referring to the Abell galaxy catalog ( galaxy clusters) . It would be just great if I was told this next time in bug reports . Fortunately I know my way around pretty well...

Now, by using blindly such outdated file formats in our FT1.2 release means calling for trouble.

It's quite obvious why you are getting problems since Selden's old list doesn't include the required entries of the galaxies /absolute magnitude/! Besides, various other crucial parameters for proper 3d display are missing, too. So the orientation in space of almost all the galaxies from Selden's files will be just incorrect, since all that information is missing...
++++++++++++

In summary: Selden's file format is incompatible with the new requirements for precision galaxy data. We coded a workaround for old /add-ons/ (with missing magnitude entry) but Selden's is a galaxy data file, NOT a proper galaxy add-on (placed in 'extras')...So just eliminate that file and you should be fine.

Bye Fridger

[jestr]

Hi Fridger,I'm pretty sure it is this file Rich is talking about
http://www.lns.cornell.edu/~seb/celestia/galaxy_clusters.dsc
but it isnt in the FT1.2 install,Rich must have moved it to his FT1.2 extras folder.It isnt an FT1.2 bug.Jestr

[t00fri]

Hi Fridger,I'm pretty sure it is this file Rich is talking about
http://www.lns.cornell.edu/~seb/celestia/galaxy_clusters.dsc
but it isnt in the FT1.2 install,Rich must have moved it to his FT1.2 extras folder.It isnt an FT1.2 bug.Jestr

Thanks, Jestr, for your help. You are certainly right (see my above mail). The main problem is that Selden's old files neither have absolute magnitude entries (hence the excessive galaxy brightness) nor orientation information. But the fact that people ignorantly mix together such outdated formats makes the hunting for REAL bugs a "chore"...


Bye Fridger

[Kolano]

Same @Kolano!

Galaxies/galactic clusters add-ons appear over bright.

WHICH one?? I have to reproduce the effect before I can fix it. So far with many add-on tests and all OS versions, I have NEVER seen galaxies as pure white overbright blobs.

Bye Fridger

Sorry I took so long to determine which add ons were the specific cause of the over bright galaxies. I didn't get around to parsing through all my add ons till last night. The specific add ons that cause issue are Selden's galaxy add-ons found at...
http://www.lns.cornell.edu/~seb/celesti ... html#3.5.2

All but the globular_clusters.dsc, which are not really galaxies, cause the issue.

What seems to be happening is that those add ons duplicate galaxies already present in 1.4.0pre-FT1.2. This causes them to be drawn twice, doubling their brightness. A few other add ons that contain galaxy defintions also produce these issues, but they usually only effect a single galaxy, making the impact less noticable.

[t00fri]

Same @Kolano!

Galaxies/galactic clusters add-ons appear over bright.

WHICH one?? I have to reproduce the effect before I can fix it. So far with many add-on tests and all OS versions, I have NEVER seen galaxies as pure white overbright blobs.

Bye Fridger

Sorry I took so long to determine which add ons were the specific cause of the over bright galaxies. I didn't get around to parsing through all my add ons till last night. The specific add ons that cause issue are Selden's galaxy add-ons found at...
http://www.lns.cornell.edu/~seb/celesti ... html#3.5.2

All but the globular_clusters.dsc, which are not really galaxies, cause the issue.

What seems to be happening is that those add ons duplicate galaxies already present in 1.4.0pre-FT1.2. This causes them to be drawn twice, doubling their brightness. A few other add ons that contain galaxy defintions also produce these issues, but they usually only effect a single galaxy, making the impact less noticable.

Kolano,

these are the same /outdated/ data files as used by Rich above. Same comments apply. Selden's data files are close to /3 years/ old and have NOT been adapted to the recent new galaxy file format. The lack of both /absolute magnitude/ and orientation entries prevent an acceptable display in Celestia-FT and the forthcoming Celestia-1.4.0 with FT1.2 being completely integrated. Also in earlier versions the galaxies from those data files will be incorrectly orientated in space and will have incorrect brightness. In addition there is considerable data overlap with the now official deepsky.dsc from FT1.2, with /all/ 10600 galaxies from the revised precision NGC/IC catalog (Steinecke 2005).

This is a vivid example that it is always MOST crucial to supply precise information about the data sources used in connection with bug reports.

Bye Fridger

[Rich]

Hi Fridger,I'm pretty sure it is this file Rich is talking about
http://www.lns.cornell.edu/~seb/celestia/galaxy_clusters.dsc
but it isnt in the FT1.2 install,Rich must have moved it to his FT1.2 extras folder.It isnt an FT1.2 bug.Jestr

Thanks, Jestr, for your help. You are certainly right (see my above mail). The main problem is that Selden's old files neither have absolute magnitude entries (hence the excessive galaxy brightness) nor orientation information. But the fact that people ignorantly mix together such outdated formats makes the hunting for REAL bugs a "chore"...


Bye Fridger

Jestr, Fridger, thanks - that fixed it.
I find the Celestia Resources directory one of my configuration management challenges due to the amount of 3rd party add-ons: Am I correct in saying that when a new celestia version comes out, in general the 3rd party models and textures are still OK to use, but 'data'-type files like the "galaxy_clusters.dsc" become suspect? When I download a celestia package, I had been updating my existing CelestiaResources files with any newly downloaded corresponding ones, but leaving alone any other existing CelestiaResources extras/data files (versus removing them)

-Rich

[ajtribick]

Given the change of file formats between 1.3.2 and the 1.4.0/FT versions, maybe a major version number change is in order?

[t00fri]

Given the change of file formats between 1.3.2 and the 1.4.0/FT versions, maybe a major version number change is in order?

I think it would be much better if add-on authors would try to update their add-ons on ML from time to time, such that synchronization with Celestia is effectively maintained. Another option would be to add information on ML concerning the required Celestia versions.

Of course in Celestia development, we try to retain as much file compatibility as possible in the course of time. However, e.g. Selden's galaxy file did /neither/ have magnitude information NOR information on galaxy orientation. And it's almost 3 years old and seriously overlaps with the content of the much more precise official deepsky.dsc file (that is now also in CVS). Personally, I am not much motivated to tediously code a workaround for the case of galaxy data where practically every single galaxy has /necessarily/ incorrect orientation in space and wrong brightness ...

If the add-ons on ML are never updated, an increasing number of fake bug reports will come from such outdated sources. I think this is of some concern, since obviously a number of ML downloaders are pretty unaware of what kind of objects they actually have installed...and frequently seem to be unable (or unwilling?) to trace them back to the sources on ML.

Bye Fridger

[selden]

I've added comments on my "catalogs" web page saying not to use the galaxy catalog file with Celestia v1.4.0. Of course, people who already have the catalog probably will never read that warning.

*sigh*

[t00fri]

I've added comments on my "catalogs" web page saying not to use the galaxy catalog file with Celestia v1.4.0. Of course, people who already have the catalog probably will never read that warning.

*sigh*

Many thanks Selden. I really think a more general version info implementation at the ML would be a most useful addition.

Bye Fridger

[ElChristou]

Guys, well... ... I will need more than a few days to end the templates because:

- I'm really busy with my normal work...
- The task become much more complex because of the use of minimum 4 layers of pts and several version of Toti's bmp2pts...

I think I will release two sets, one in lowres and the other one, full res with templates around 500/600 ko each...

Later,

Bye.

[t00fri]

Hi,

people trying ElChristou's technique of overlaying several different kinds of bmp2pts templates will end up with brightness values in the fourth column of the .pts template that are significantly >1.0. Which is BAD.

Here is a 4 line Maple (Lua ) code that properly normalizes such resulting multy layer templates. You may easily carry the steps over into the computer language you know best:

Code: Select all

> S:=readdata("SBb.pts",[float,float,float,float]):
> nops(S);
                             10650
> bmax:=max(seq(op([i,4],S),i=1..nops(S)));
                            1.70079
> writedata("X.pts",[seq([S[i,1],S[i,2],S[i,3],S[i,4]/bmax],i=1..nops(S))]):


It's very easy to understand:

The first line reads in here an unnormalized SBb.pts template into an array called S. The second line counts the number of "records" (with [x,y,z,brightness] values, each). It's 10650 in this example. The next line calculates the maximum brightness value (4 th data column) which is 1.70079 here. Finally, the data are written back onto some file "X.pts" where the brightness column has been divided by the maximum value in S, such that all values remain <= 1.0, as it has to be!

That's it.

Bye Fridger

[Boux]

You can also load the file into a 4-column Excel sheet (or your favorite OSX/Linux app) and do the normalizing stuff right away on the 4th column which stores brightness.
The big advantage of Excel and siblings is that one can create several versions of the same *.pts with various brightness level normalization, just playing with the formula and saving at will
BTW is it just me or El Christou's template is broken? Looks like data is missing on some lines (only 3 instead of 4 entries).

EDIT!
I have checked some other templates and there is also some data missing here and there in the 4th column.
Question: are those missing data for black (value 0.0000... invisible) points?
If yes, are all the black points in the template loaded in and processed by Celestia anyway or are they discarded for faster rendering?
I am away from my Celestia box now and cannot look into the code.
Cheers!