[fitsbits] Question(s) regarding development of proprietary FITS manipulation software. . .

Mark Calabretta mcalabre at atnf.CSIRO.AU
Mon Aug 27 21:20:31 EDT 2007


On Mon 2007/08/27 11:32:19 -0400, Michael Williams wrote
in a message to: Thierry Forveille <Thierry.Forveille at obs.ujf-grenoble.fr>
and copied to: fitsbits at donar.cv.nrao.edu

Michael,

The "I" in FITS originally stood for "Image", now it really stands for
"Information".  To get a better understanding of what FITS is about I
suggest installing 'fv' and using it to explore its test files and a
selection of FITS files that you gather from the internet.  fv can
display astronomical data as images, plots, histograms, etc as
appropriate.  In fact, I doubt that there is anything that you would
want to do with a general FITS file that fv doesn't already do,
http://heasarc.gsfc.nasa.gov/docs/software/ftools/fv/

To get a better understanding of the visualization aspects, try
installing Karma and use 'kvis' and 'xray' to play with 2D or higher
dimensional images (not binary tables, etc.).
http://www.atnf.csiro.au/computing/software/karma/

To view general images with proper world coordinate handling, try
ds9: http://hea-www.harvard.edu/RD/ds9/

>1)  Is a FITS file really similar to, say, a Microsoft Word =20
>document. . .in that it can store *any* kind of data (images, mp3s, =20
>pdfs, minivans) with the only thing that can be counted on being the =20
>"standard" header?  It seems that my problem is the fact that I'm =20
>expecting FITS to actually *fit* (pun intended) into some specific =20
>category with absolutely specific syntactical and semantic requirements.

In some respects FITS is a bit like the Roman alphabet, it only provides
a framework for communication.  Knowledge of the Roman alphabet allows
you to read something written Italian, German, French, etc. without
understanding what the words mean.  It doesn't work so well for English
or Polish which should give you an idea of how FITS, likewise, has been
corrupted by usage!  Various garnishes, accents and the like, that have
been added to the Roman alphabet in some languages also have an obvious
parallel in everyday FITS.

FITS does also have some semantic content that is interpretable by
software, mainly associated with the interpretation of physical
measurements.  However, most of FITS' semantic content is really only
understood by software written by the originators of the FITS file,
and also by humans reading the comments in the header.

>2) What is the significance of *any* imagery in FITS?  Are the images =20=
>
>literally live pictures that were taken, or some arbitrary graphical =20
>representation of data?

Two examples:

1) Rotational synthesis radio telescopes don't "take photos" in
the same sense that some optical telescopes do.  They measure the
components of the Fourier transform of the sky brightness distribution
("visibilities") and these are later synthesised into an image.
FITS can store both the visibilities and the image.  The resulting
image can certainly be "viewed" but it is a visual representation of
something that cannot be seen.  Such images typically have artifacts
that require special algorithms to remove.  The purpose of generating
such images is usually to measure the brightness and the precise
position on the sky of particular radio sources.  Images of extended
sources taken at different wavelengths are sometimes compared - these
may contain profound astrophysical content, e.g.
http://antwrp.gsfc.nasa.gov/apod/ap020920.html
http://antwrp.gsfc.nasa.gov/apod/ap020812.html

2) The detectors on board an orbiting gamma-ray telescope record the
direction and strength of each incoming photon.  These individual bits
of data can later be combined into a type of 2D histogram, a sort of
"image", showing where the photons came from.

Have a look through the backlog at "Astronomy Picture of the Day" for
further examples.  Today's just happens to be aposite.
http://antwrp.gsfc.nasa.gov/apod/astropix.html

>3) How do I determine what type of image manipulations are legitimate =20=
>for any type of image data?  Should I allow sepia toning?  Should I =20
>allow them to run photoshop filters on the pictures?  I mean, what, =20
>really is the usefulness of any manipulation on the image and what =20
>are the most common techniques?

Play with the image display options in kvis, I think you'll get the
idea pretty quickly.

>4) What is the goal of scientists when it comes to examining =20
>another's FITS files?  What kind of "information" is truly gathered =20
>from the "data"?

In a nutshell, the basic aim is to understand the origin and evolution
of the universe.

Mark Calabretta
ATNF




More information about the fitsbits mailing list