Pointing Hubble

[selden]

One of the things one will be able to do with Celestia v1.5.0 will be to point Celestia's model of the Hubble Space Telescope toward any object in Celestia.

This Addon adds an image of Hubble's field of view so you can see just how small an area of the sky the HST actually looks at.

http://www.lepp.cornell.edu/~seb/Celestia/files/pointing_hubble.zip
(470 KB)

You can edit the included pointing_hubble.ssc to direct the telescope and its fov to look at the eye candy that you've installed.

This picture shows the HST investigating "The Pillars of Creation" in the Eagle Nebula.



[edit]6-jul-07
pointing_hubble_v0.7 is now available at
http://www.lepp.cornell.edu/~seb/celest ... le_v07.zip (86KB)

It now asks how many views to show.
It also allows you to rotate Hubble's field-of-view on the sky.




[/edit]

[Malenfant]

That's a pretty neat little addon - puts things in perspective somewhat (no pun intended )

[rra]

Selden,

One of the things one will be able to do with Celestia v1.5.0 will be to point Celestia's model of the Hubble Space Telescope toward any object in Celestia

how is this new feature (for Celestia 1.5.0) to be used
can one just select the object and use some
keyboard controls to adjust the orientation ?
Does this allready work on the latest CVS version ?


Ren?©

[ElChristou]

Hey, seems cool! I go for a test. Tx Selden!

[produit]

Hi,

I will love to have that for INTEGRAL.
For INTEGRAL, the problem is simpler as
if you download the files in my account
http://isdc.unige.ch/~produit/celesia
you will have always INTEGRAL pointing at the proper
direction. So the FOV has a well defined direction (along the long
axis of the satellite).
So it is just a matter of defining the good
size and form of the FOV.
This will in fact be very useful for INTERAL users.
If you can give me instruction how to achieve this
I will do it.

In INTEGRAL there are two concept of FOV:
The fully coded field of view (FCFOV) in which we have
maximal sensitivity. Then the sensitivity is decreasing
and reach 0 at the 0 coded field of view (0CFOV)

The FCFOV of the different instrument are
(all FOV are aligned with the fixed solar panels)
IBIS: square of 9x9 degree
OMC: square of 5x5 degree (this is in fact a real FOV)
JEMX: circle of radius 2.4 degree
SPI: Hexagon 16 degree on a side

The 0CFOV:
IBIS: square of 27x27 degree
JEMX: circle of 5 degree
SPI: Hexagon of 30 degree on a side

Degree are not typos, FOV of INTEGRAL are realy huge compred
to Hubble.

Thanks

[selden]

Ren?©,

This Addon was tested using Celestia built from CVS on 29 November. I suspect there will be changes to the Frame definitions which will require changes to the Addon for the final release of v1.5.0.

Unfortunately, Celestia does not yet support interactive realtime control of SSC objects, including Hubble. One can hope that realtime control will be available in a future version of Celestia.

The observation currently has to be defined in advance by editing pointing_hubble.ssc to specify the target and Hubble's roll angle. Beginning and Ending directives or a .q quaternion file could be used to observe a sequence of targets, but I didn't make it that elaborate.

In principle, one could simulate realtime control by using a screen overlay with the Lua Edu package to display the field of view, adjusting the size of Hubble's FOV overlay to match the viewpoint, but I haven't investigated that, either.

[selden]

Nicolas,

Is a graphical image available of INTEGRAL's fields of view? Or a parametric description of their relative and opening angles? They're large enough that the method I used for Hubble's fov might need some modification.

The method used to display Hubble's fov is relatively simple: a square model is placed 1km in front of the telescope with an appropriate radius. A diagram of the fov is used as its surface texture. Another object is defined immediately in front of Hubble to be used as the viewpoint destination.

To make an observation, after defining the object and roll angle in the SSC, you GoTo the viewpoint object and Track the FOV object.

There are several things that can be done to improve the addon, but I didn't want to spend a lot of time "polishing the apple"

For example, a model consisting of OpenGL lines could be used instead of a surface texture image. This would improve the representation of the fov outlines.

Also, there is a slight parallax offset of the fov since the viewpoint won't be exactly at the correct pov: Celestia stops before you get there. This can be reduced by placing the fov object much further away from the telescope. I'll do that for the final release for v1.5.0. One also can eliminate this parallax by specifying a precise pov location relative to the telescope either in the GoTo menu or by using gotolonglat in a script. I'll add such a script in the final version.

[rra]

Selden , regarding your add-on :

I tried it at work, but al I get is a white flat square, non-transparant ,
with black lines and markers as shown in your previous post.
At my work I don't have openGL 2.0,
could that be the problem ??


Ren?©

[selden]

Ren?©,

I think this has to be considered a bug in Celestia. I'll make sure Chris hears about it, just in case he doesn't read this thread.

The CMOD model specifies an emissivemap texture so that the lines glow. Emissivemaps are new in v1.5.0.

I don't see anything in what Chris wrote about the new cmod texture maps which says that they require OpenGL 2.0. I don't think they should and that the failure is just due to an oversight.

[buggs_moran]

I wonder if there is a file out there containing the dates, RA and Dec pointing info for things like Hubble, Chandra, etc. that could be used to create an ssc with all pertinent directions over the years...

For that matter, I wonder if there is a similar spreadsheet that contains mission dates, and orbit data for all of the shuttle flights so when the shuttle was in orbit we could go "see" it... like we do with a spacecraft like Voyager...

[selden]

I've generated an updated Addon which works around the transparency bug.

http://www.lepp.cornell.edu/~seb/Celest ... le_v02.zip
(1.1MB)

The original images were not a power-of-two on a side. I've resized the fov image from 901x901 to 1024x1024. New graphics cards, like Nvidia 6nnn or greater, can use non-power-of-two sizes. Celestia v1.5.0 will automatically scale non-power-of-two images so they'll work on older cards, but there seems to be a bug.

The Hubble fov image looks much better in the OpenGL 2.0 render path than it does in the other paths. I suspect not much can be done to improve the older rendering paths. I guess I'll have to investigate using a model instead of an image sooner rather than later

[produit]

Hi,

there is a sketch of the FOV of INTEGRAL in:
http://integral.esac.esa.int/AO4/AO4_Policies.pdf
page 21
This sketch with numbers and explanations in my email should do it.
The best will be to have lines drawn the same way the coordinate Ra Dec
grid line are done.

Tell me if this is sufficient or if you need more explanations.

Thanks

[ElChristou]

...I guess I'll have to investigate using a model instead of an image sooner rather than later ...

I confirm the white square if not under OGL2.0...
A model would be really great for a better resolution in all case...

[selden]

Nicolas,

That should be enough.

Thanks!

[selden]

I've updated the Addon to be
http://www.lepp.cornell.edu/~seb/celest ... le_v03.zip
(500KB)

I replaced the image of the fov by a 3d model. It doesn't use lines. Getting those right would have taken too much time and effort. Instead, it's a translucent model of the various detectors and labels.



It seems to work reasonably well.

[Cham]

Selden,

your fov model could be improved easily :

you don't need to add any depth to the pieces (green arcs and letters). Made them flat (no depth), so they'll look better (no transparent parts with confused rendering) and they'll have less polygons. And the green arcs really need a smoother edge (it looks crude).

[selden]

Cham,

Yes, all of those could be improved.

This is only v0.4. v1.0 will be better

[ElChristou]

Cham,

Yes, all of those could be improved.

This is only v0.4. v1.0 will be better

Selden, have a look at your mailbox...

[selden]

Nicolas (and anyone else who is interested),

An Addon is now available to display the fov of Integral:

http://www.lepp.cornell.edu/~seb/celest ... tegral.zip (31KB)

(Yes, it's misnamed: it assumes that the ssc which defines Sol/Earth/INTEGRAL is doing the pointing.)

It uses glowing lines in space (a CMOD line model) to outline the fields of view. Please let me know if it's OK.

[Cham]

Selden

your model is crashing Celestia. Total lock-up.

Apparently, it is the cmod which is doing this, but I didn't wanted to reproduce the painfull crash, so it need a confirmation.