Milky Way, LMC and SMC dims in 1.5.0 cvs Sep 28

ANDREA · Post #1by **ANDREA** » 28.10.2006, 00:56

I've installed the 1.5.0 cvs release dated Sep 28th, and testing it I found a strange effect.
After checking the deepsky.dsc dated Sep 28th, I found that dims are:
Milky Way= 5e+04
LMC= 1.596e+04
Well. as shown in the image, tha LMC looks more than half the MW, instead of less than 1/3 (and the SMC too looks much bigger than the due):

BTW, Just to be sure, in the image the galaxy luminosity is the maximum.
Someone can help me to understand if I'm wrong or it's a bug? :oops:

Thank you.
Bye

Andrea

ANDREA · Post #2by **ANDREA** » 26.11.2006, 23:02

Sorry for repeating this post, but there are two possibilities to explain why none answered me:
1- it's such a stupid question that it's not worthwhile replying;
2- the problem exists, but none admits it. :wink:

Waiting to know if I'm stupid or not, I inform you that the image remains the same even with the latest cvs installed (Nov 24th, I think), and the MilkyWay.png file is dated Sep 11th, 2006.
Thank you for your time.
Bye

Andrea

Post #3by **selden** » 26.11.2006, 23:41

Andrea,

The radius of the LMC specified in deepsky.dsc
is 1.596e+04 and the radius of the Milky Way is specified as 5e+04 so their relative sizes ought to be reasonable.

The LMC is specified as Irr, but I can find no Irr template png image.

The Milky Way uses its own CustomTemplate png image.

My guess is that the algorithms for deciding how large to draw the two different types of models (irregular vs template) are not using compatible calculations.

When I replace the "Irr" Type by "E0", the relative sizes look more reasonable.

I know that Fridger has been planning to upload some corrections to the Galaxy code and templates. Hopefully this will be one of the problems that is fixed.

ANDREA · Post #4by **ANDREA** » 26.11.2006, 23:45

selden wrote:Andrea, ... My guess is that the algorithms for deciding how large to draw the two different types of models (internal vs template) are not using compatible calculations. I know that Fridger has been planning to upload some corrections to the Galaxy code and templates. Hopefully this will be one of the problems that is fixed.

Thanks a lot Selden, a fast and exhaustive reply, as always. :wink:

Bye

Andrea

Post #5by **selden** » 26.11.2006, 23:47

Not quite so exaustive at first!
I added a comment about testing E0 sizes while you were responding, so you may not have seen it.

ANDREA · Post #6by **ANDREA** » 26.11.2006, 23:54

selden wrote:Not quite so exaustive at first!
I added a comment about testing E0 sizes while you were responding, so you may not have seen it.

Yes Selden, I missed it, now I see the problem.

Thank you.
Bye

Andrea

Post #7by **t00fri** » 27.11.2006, 00:04

selden wrote:The LMC is specified as Irr, but I can find no Irr template png image.
...

There has never been an Irr template since Irr's are generated in galaxy.cpp...

I'll look into that issue. I have never really checked that Irr code, since it's not by me. We'll see.

Bye Fridger

ANDREA · Post #8by **ANDREA** » 27.11.2006, 00:23

t00fri wrote:
selden wrote:The LMC is specified as Irr, but I can find no Irr template png image.
...
There has never been an Irr template since Irr's are generated in galaxy.cpp...
I'll look into that issue. I have never really checked that Irr code, since it's not by me. We'll see. Bye Fridger

Thank you Fridger.
Bye

Andrea

Post #9by **t00fri** » 28.11.2006, 20:08

ANDREA wrote:
t00fri wrote:
selden wrote:The LMC is specified as Irr, but I can find no Irr template png image.
...
There has never been an Irr template since Irr's are generated in galaxy.cpp...
I'll look into that issue. I have never really checked that Irr code, since it's not by me. We'll see. Bye Fridger
Thank you Fridger.
Bye

Andrea

Andrea,

indeed, whoever coded the Irr galaxy stuff in galaxy.cpp
produced a size error by a factor of 2.

Here is the proof.

I entered the SMC galaxy twice in deepsky.dsc. Once as an Irr (default) and once with identical coordinates and parameters as a spherical E0 Hubble type. Since the two images now appear exactly overlayed, one can clearly see that the pancace-shaped Irr version is too large by a factor of two relative to the spherical E0 shape. Both types have been entered with exactly the same radius.

I have fixed this along with fixing the orientation of the arms (to counter-clockwise) of our MilkyWay.

See
http://www.celestiaproject.net/forum/viewtopic ... ight=milky

I'll commit the changes soon to CVS.

Bye Fridger

Post #10by **t00fri** » 28.11.2006, 20:09

ANDREA wrote:
t00fri wrote:
selden wrote:The LMC is specified as Irr, but I can find no Irr template png image.
...
There has never been an Irr template since Irr's are generated in galaxy.cpp...
I'll look into that issue. I have never really checked that Irr code, since it's not by me. We'll see. Bye Fridger
Thank you Fridger.
Bye

Andrea

Andrea,

indeed, whoever coded the Irr galaxy stuff in galaxy.cpp
produced a size error by a factor of 2.

Here is the proof.

I entered the SMC galaxy twice in deepsky.dsc. Once as an Irr (default) and once with identical coordinates and parameters as a spherical E0 Hubble type. Since the two images now appear exactly overlayed, one can clearly see that the pancace-shaped Irr version is too large by a factor of two relative to the spherical E0 shape. Both types have been entered with exactly the same radius.

I have fixed this along with fixing the orientation of the arms of our MilkyWay (to counter-clockwise) .

See
http://www.celestiaproject.net/forum/viewtopic ... ight=milky

I'll commit the changes soon to CVS.

Bye Fridger

ANDREA · Post #11by **ANDREA** » 28.11.2006, 22:45

t00fri wrote:
ANDREA wrote:
t00fri wrote:
selden wrote:The LMC is specified as Irr, but I can find no Irr template png image.
...
There has never been an Irr template since Irr's are generated in galaxy.cpp... I'll look into that issue. I have never really checked that Irr code, since it's not by me. We'll see. Bye Fridger
Thank you Fridger. Bye Andrea

Andrea, indeed, whoever coded the Irr galaxy stuff in galaxy.cpp
produced a size error by a factor of 2. Here is the proof.
I entered the SMC galaxy twice in deepsky.dsc. Once as an Irr (default) and once with identical coordinates and parameters as a spherical E0 Hubble type. Since the two images now appear exactly overlayed, one can clearly see that the pancace-shaped Irr version is too large by a factor of two relative to the spherical E0 shape. Both types have been entered with exactly the same radius. I have fixed this along with fixing the orientation of the arms of our MilkyWay (to counter-clockwise) .
I'll commit the changes soon to CVS. Bye Fridger

Now it's very clear, Fridger.
Thank you for solving the problem.
Bye

Andrea