Creating 3-Color Images
Overview:
If you intend to visualize both the
energy and intensity in
astronomical images, a 3-color image might be a good solution,
especially if intended for the general audience or public/press
releases. This page describes how to create 3-color images from both
UVOT and XRT data, or from a combination of UVOT and XRT data.
Read this thread if you want to: Create "true" 3-color images in different energy and intensity scales.
Last update: Nov 27, 2005
3-color images can be created in various ways. Some of the more common
methods are described below.
This thread assumes that you have created three FITS files in three
different filers (UVOT) or energy bands (XRT).
1. DS9:
SAOImage DS9 is an
astronomical data visualization application for files in FITS format.
It consists of a set of GUI application available for most computer
platforms and provides visualization support for a number of
popular astronomical analysis software packages, such as the
Swift
FTOOLS.
You can use DS9 either in command-line mode or in GUI mode to create
3-color images.
Command-line mode:
In
your terminal execute the command
ds9
-rgb -red <red
image.fits> -green <green image.fits> -blue <blue
image.fits> |
The 3-color image will be visible in the main DS9 display, while a
pop-up "RGB" window gives you control over the currently selected image
to adjust levels, scales, etc., in the main window.
After the right scale and levels have been selected, click on "File" in
the main menu and next on "Save Image as". Your can chose between
different file formats such as JPEG and TIFF. If you want to create a
Postscript file, click on "File" and then on "Print". Chose "Print to
File" and enter a file name (the default will be "ds9.ps"). If you
intend to create 3-color images for print media, select a resolution of
300 dpi and check the "Interpolate" button. Many journals and
newspapers request 4-color files (CMYK) for
reproduction, which requires that you click on the "CMYK" button. Since
your computer screen is only able to display 3 colors, make sure to
make test printouts on a 4-color printer to check the results.
GUI mode: You can also simply
start the DS9 application, then click on "Frame" and "New Frame RGB".
The "RGB" window will pop up. Select the "Red" channel in the "RGB"
window, click on "File" and "Open" in the main DS9 window to load your
red image. Do the same for the green and blue channels.
In both cases you can overlay source markers and text using DS9 region
files.
The example below shows UVOT 3-color images of SN2005ke/NGC3173 from
summed V, B, U, and
UVW1, UVM2, and UVW2 images, created with the commands
$
ds9
-rgb -red uvv_sum.fits -green ubb_sum.fits -blue uuu_sum.fits -region
circle.reg |
$
ds9
-rgb -red uw1_sum.fits -green um2_sum.fits -blue uw2_sum.fits -region
circle.reg |
2.
Photoshop and FITS Liberator:
The
FITS
Liberator is an increasingly popular plugin for
Adobe
Photoshop to create "press-ready" color images from FITS files.
The
Adobe
Photoshop is a commercial product which offers a wide range of
image editing capabilities. The wider range of image manipulation
tools,
such as
stretchable image scales, makes the FITS Liberator a superior tool
compared to the DS9. Please note that Adobe Photoshop is only available
for Windows and Mac platforms. A FITS Liberator plugin for The Gimp,
popular on Linux platforms, will not be available in the near future.
3.
CIAO Tool "dmimg2jpg":
Experienced Chandra users might find it easy to create 3-color images
using existing CIAO tools.
The Chandra Interactive Analysis Of Observation,
CIAO, is a popular high-level
data analysis package for the reduction of Chandra and other X-ray
data. If you
are familiar with CIAO, a fast way to create 3-color images is
to follow this
CIAO thread
which creates a "true" 3-color image from any three files in FITS
format (not only obtained from Chandra observations).
4.
IDL Tools:
You can experiment with countless other tools, such as IDL image
libraries.
For example, you can pass a 2-D image through the three color vectors
that make up a color
table.
IDL> thisDevice = !D.NAME SET_PLOT, 'Z' LOADCT, 5 TVLCT, red, green, blue, /GET SET_PLOT, thisDevice
|
Make sure the 2-D image data is scaled into the 256 colors.
Next, create and fill up the 3D image array:
IDL> thisImage = BYTSCL(image) s = SIZE(thisImage) image3d = BYTARR(3, s(1), s(2)) image3d(0, *, *) = red(thisImage) image3d(1, *, *) = green(thisImage) image3d(2, *, *) = blue(thisImage)
|
Finally, write the JPEG file using the WRITE_JPE
IDL> WRITE_JPEG, 'myimage.jpg', image3d, TRUE=1, QUALITY=
|
JPEG uses a "lossy" compression scheme in which some of the original data can be lost
when the file is de-compressed. The "Quality" parameter sets the amount of information
you wish to keep. The default value is 75.