Re: Images

[Tass Mailing List <tass@mail.alembic.net>]

Date: Tue, 02 Jan 2007 19:31:58 -0500
From: Doug Welch <welch@physics.mcmaster.ca>
To: Tass Mailing List <tass@mail.alembic.net>
Subject: Re: Images


Tom et al,

I think that it would be far more valuable if the pass through the CDs
and DVDs resulted in them being publicly available in an image archive.
There are a number of places the images could be stored. I know the CADC
would be interested and I have an arrangement with them for uploading
images to be made available.

The image headers contain everything which is needed to make sense of
the images - even to the extent of dealing with duplicates. I would
recommend an effort of feeding CDs and DVDs which would most directly
result in images being made available on well-maintained, high-bandwidth
sites and not just producing an intermediate cross-reference list for
image header information.

Cheers,
Doug

Tass Mailing List wrote:
>
> Date: Tue, 02 Jan 2007 16:50:46 -0600
> From: Thomas F. Droege <droege@fastmail.fm>
> To: tass@tass-survey.org
> Subject: Images
>
> Looking around, I estimate that I have images on 2000 CDs and 200 DVDs.
> Best guess is about 500,000 images.  I would like to get them in a form
> where images containing a particular star could be found.  This is
> important if one tries to use the measurements, since funny things can
> happen to a particular image.  An example is a faint satellite trail
> that crosses a star image.  If faint enough, the data reduction does
> not reject the measurement.  Both V and I images then look somewhat
> brighter, simulating some sort of outburst.  To check this one really
> needs to look at images.  There is probably too much data to be put up
> on a data base.
>
> I propose to start cataloging the image disks.  This is a good mindless
> project for a sick person.  Just set up a system and start feeding all
> the disks through it.  I would propose a small program that would do the
> following:
>
> 1) Read all the fits headers on the disk.
> 2) For each write a summary line.
> 3) Output a file with the name tass_images_n where n is a sequential
> number for each disk. The file would contain a summary line for each
> image pair.  Note that we only have image pairs and we always have image
> pairs.
> 4) I would then go through all the CDs/DVDs and generate the summary
> files.
> 5) I would then label each CD/DVD with it's sequential disk number and
> store them away in 50 disk cake boxes.
>
> It would then be relatively easy to query the summary line file for say
> all the images containing a particular star.  This is not perfect since
> we only have the approximate center of each image stored in the fits
> header.  One could then extract disks 10,457, 2068 relatively easily
> from the stack of cake boxes.
>
> I am willing to read all the disks and set up the cake box stack.  I am
> also willing to answer a reasonable number of queries by manual
> extraction of the images.  If this gets to be a burden, then we have a
> good thing - the data is interesting - and one could go to the trouble
> of loading all the images in some data base.
>
> At the moment, we can't even find a single image set spanning the
> northern sky.  This might be useful and could be easily constructed once
> we had the summary line file.
>
> First we need to agree on what should go in the summary lines.  Some
> things:
> 1)Disk Number
> 2)Image Name
> 3)ra of image center
> 4)dec of image center
> 5)JD t1ime of exposure
> 6)exposure length (we have 100 and 200 second exposures)
> What else?
>
> Then I would need a program to do this.  Any volunteers?  Does not seem
> like a big task for some of you.
>
> Tom Droege
>