Ice in Space Joint Imaging Project?

[cventer]

Ice in Space Joint Imaging Project ?

Eddie T’s idea to create a giant mosaic of LMC got me thinking.
<FONT size=3><FONT face="Times New Roman"><?xml:namespace prefix = o ns = "urn:schemas-microsoft-com[img] /><o:p></o:p></FONT></FONT></P><P style=[/img]
What would be very cool is if we planned an Ice in Space GIANT mosaic.
<o:p></o:p>

There are enough good imagers on this group to make this a reality. It would need a project manager, which I can't volunteer for as I just don’t have the time right now, but we could jountly pick an interesting area of the Sky.


The Project manager could divide the area up into squares and list them with coordinates. People could then volunteer to image certain squares that are then marked as taken. Over a period of 6 – 12 months we could build up a fantastic image that could make a giant poster. Possibly one of a kind in the world. We would then need a team of image processors to put the final thing together.


We could display it online as kind of a giant Jigswaw puzzle. As people finish an image, the thumbnail completes that sqaure. eeveryone can watch the squares build up over time to form the image.
<o:p></o:p>

Obviously a few logistics to work out as different people have different focal lengths etc… but if we break it up small enough, then people with wider imaging rigs can just take 2 or 4 areas at once.
<o:p></o:p>

We would need a good acronym for the project that reflects its coolness.


Ice in Space Joint Imaging Project – IISJIP does not quite do it for me, so get your thinking caps on.


What does everyone think. ?
<o:p></o:p>

Best Regards

Chris Venter

[avandonk]

Good Idea!Can't see any problems with different focal lengths as Registar can easily handle this. To make the mosaic use Registar to produce individual registered images and then stitch these together using something like Imerge.
Here is a pic done this way.I left it bright to show that the join is almost invisible.I have been thinking of doing a mosaic with my 300mm lens.This sounds even more ambitious.

avandonk

[[1ponders] (Paul)]

Luv It!! How would you imagine going about processing the shots so they all "look" the same. With 300D, 350D, 20D, 20Da, SBIG and goodness knows what else lurking round the site that in itself will be a challenge.

I also imagine matching resolutions would be a challenge as well. May be make it a requirement that everyone buy Orion 80ED

How about IISMIP, Ice In Space Mosaic Imaging Project

[seeker372011 (Narayan)]

love the idea, doubt if anything like that has ever been done before!

[Striker (Tony)]

There are enough good imagers on this group to make this a reality

Dam well l'm out....how about noob with 2 weeks of experience...does that qualify...lol

Sound like a fine project...I vote Paul as project manager...congrats Paul.....hehe

[cventer]

We would need to set some gudelines as far is the image scale in arc/sec per picel that each frame of the mosaic would need to be.

This may be a range eg 1.5 - 2.5 arc sec per pixel. We can scale images with tools like Registar , but obviously we cant scale a 3 arc sec pic to 1 arc sec resolution without it lookign blocky and pixelated.....

ED80 with say 300D gives 2.54 arc sec per pixel, with a 20D its around 2.2 arc sec per pixel. The range we would have to consider to allow the most number of people to take part is probably in the < 2.5 arcsec range. Higher res can allways be binned or scaled to match without affecting quality.

An ED80 with 300D at native f7.5 gives a FOV of 85 x 123 arcmin. A Takahashi FS102 with 300D gives 63 x 95 arcmin fov. An 8inch SCT with 6.3 reducer gives 40 x 61 arcmin (at 1.21 arc sec per pixel). So we would have to work out what size FOV a standard UNIT of the mosaic is.

I have not done Mosiaics before so some gudinace from the more experienced. (Eddie T are you listening ?) would be good.

Best Regards
Chris Venter

[[1ponders] (Paul)]

You want to make me project manage Tony Have you seen my shoelaces? I buy shoes that don't have them

Chris, fur us newbies like myself and Tony, a quick lesson in working out those arc second sort of details would be appreciated. I only started imaging in Feb this year and while I sorta understand the figures your talking about I have no idea how to work them out.

Yeah....yooo hooo, I know your busy Eddie...but some advice......

[cventer]

Ok, no problemo

The hard way is:

Sky Coverage in arc-seconds/pixel = 206.265 / (focal length in mm) *(pixel size in microns)

Then to work out your area covered by your chip is simply a matter of multiplying your chips resolution in pixels by this number. A 300D has 3072 x 2024 pixel array. So multiply 3072 by your arc sec per pixel and you get length of sky coverage. Multiply 2024 by your arc sec per pixel and get the height.
The easy way is to download a free tool to do it for you in one step.

http://www.newastro.com/newastro/boo...camera_app.asp

Best Regards
Chris Venter

[[1ponders] (Paul)]

thanks

[Robby]

Sounds like a great idea. All that math throws me, but give me a piece of the sky & I'll image it
Kinda like a Google-Space.... or perhaps the SpaceOnIce project.
Cheers

[rowena]

This is something I was actually interested in doing myself... but for me personally i was thinking of mapping the sky at a certain focul length with my camera.

The other thing I may be able to help with is processing power. Depending on how much we need and how long the processing will take. Currently I have two machines sitting idle 80% of the time with 3Ghz+ cpu's and 512mb RAM+.

as cventer mentioned we need soome guidelines for arc seconds per pixel aswell..

with a 300mm lens on my 300D is capturing around 0.68755 arc seconds per pixel? from the below calculations... which is 206.265/focul length. is this correct? or have i got this wrong... with this calculation then the smaller the focul length the more pixels per arc second... and that would be incorrect.. because being zoomed u need to do multiple shots to get the same areaa as a smaller F.L and therefore you are capturing more data with the larger F.L in the same area covered.

This will be a good and interesting project to work on!

Originally Posted by cventer

Ok, no problemo

The hard way is:

Sky Coverage in arc-seconds/pixel = 206.265 / (focal length in mm) *(pixel size in microns)

Then to work out your area covered by your chip is simply a matter of multiplying your chips resolution in pixels by this number. A 300D has 3072 x 2024 pixel array. So multiply 3072 by your arc sec per pixel and you get length of sky coverage. Multiply 2024 by your arc sec per pixel and get the height.
The easy way is to download a free tool to do it for you in one step.

http://www.newastro.com/newastro/boo...camera_app.asp

Best Regards
Chris Venter

[cventer]

The calculation depends on what 300m lens you have. If its the canon f4 then your focal lengh in is 308mm (apperature x focal ration. ie 77 x 4) This gives 4.95 arc sec per pixel on the 300D.

The Pixel size in microns for the 300D is 7.4

So calculation is : 206.265 / (focal length in mm of 308) *(pixel size in microns of 7.4)
26.25 / 308 * 7.4 = 4.95

Best Regards
Chris

[[1ponders] (Paul)]

Actually I'm a bit confused about the pixel size Chris. In some reviews I've read it at 7.4 and in others it was stated as 7.1. Have there been two versions of the chip released at some point?

[avandonk]

Here is a mini attempt at a mosaic.One image taken with an 85mm lens and two with a 300mm lens.One picture shows all three stacked with Registar.The other only the 300mm images but they are still in register.To maintain the resolution of the 300mm images the combined file of all three ends up to be 500M in size.

To set up a framework for images produced by longer focal lengths we could use
300mm images as a way of registering them without distortion as the field of
a good 300mm lens almost has neglible distortion.

I have found if you try to make mosaics out of small images without a framework the mosaic starts to distort rapidly as you add more images.

What are your thoughts on this as a strategy?

I have done another with tornado33's fine image of the Trifid Nebula superimposed on a 300mm image just as a test.

[cventer]

Thats a smart idea using the wide field shots to register the more detailed shots. I like it !!!!!

[cventer]

Looks like interest is here, so:

Its now time to decide on an area of the sky we wish to Image.

My vote is we focus on Southern Area to make it unique to Southern Hemisphere.

So if we image and area that encompasses 47 Tuc, LMC, SMC, Crux, Eta Carinae, Taruntula, Omega Centauri etc..... I dont have my star charting software here to print an image of a proposed area....

The good thing is we probably have a lot of images of objects in this area allready.

Best regards
Chris Venter

[EddieT (Eddie)]

Hi all, I wasn't listening, but just discovered the thread. How do you guys keep up with each other ?

I like this idea a lot, but it is frought with difficulties and would be a triumph of astronomical proportions if it succeeded (poor pun intended). The biggest mosaic I have done was 28-tiles, The Milky Way mosaic and that was only in B&W. It took me over a month to match the component images. Colour adds a whole new dimension of difficulty to the mix.

Even with B&W images only, all taken with the same equipment, from the same location on successive nights, the background variation creates a pretty great problem to overcome. When you include colour, the background variation goes from shades of grey to different hues of colour, which is much more difficult to compensate for.

The alignment issue is very real when mixing images from different plate scales and resolutions, but can be overcome and is the least of the problems.

More of an issue, is the Signal-to-noise ratio. This is where it gets very iffy. THE SNR can be calculated and if everyone sticks to acheiving roughly the same SNR, then it *may* be possible to have all of the images looking similar enough to join together and appear to be one, but in practice this is very hard to do. The SNR is determined by many things. Sky brightness, camera noise characteristics, ambient temperature, focal ratio, focus, tracking, etc, etc. If the component images are too different (and it doesn't take much) you will end up with a patchwork quilt rather than a final single image that looks like one image.

I do this full-time these days and my estimate of 12 colour frames for the LMC mosaic taking 6-months, was made assuming 12 clear, moonless nights....Sounds easy but in reality, well as we all know, the clouds only clear when the moon is between waxing and waning gibbous! For mosaics a moon that changes in brightness every night will make the final task very difficult to say the least, introducing background gradients that are a royal pain and changing the SNR from image-to-image.

I don't want to rain on the parade, I just want you to know what your getting yourselves in for Perhaps when you decide who will be involved, you should start with a small project and see how it goes. I would be more than happy to be proven wrong and have you all create a great image and say "Sheesh what a drama queen that Trimarchi is!".

[avandonk]

EddieT is correct it is difficult for all the reasons he gave.So initially why dont we choose just a small area and try to image that.Here is one I cooked before.

I think we should at least try.I am willing to take the 300mm Image of the area of choice and then people can take images even at random within the area and as they come in superimpose to show what has been captured.Once we have the area covered we can then start to adjust individual images to obtain a match in signal to noise,resolution, colour balance etc.

I for one have seen many superb images but it is very hard to imagine how they are all related and where they are on a naked eye view.Large fields overcome this.Just my two bobs worth.

[EddieT (Eddie)]

I look forward to seeing the outcome. I can't contribute, not enough time and too many projects, but I think it's a great project and look forward to seeing how it progresses.