Functional QUICK LOOK Requirements
General requirements
The quick-look operations will all be accessible from the
image display tool.
Obviously, there must be communication with the image display for the purposes
of specifying positions with the cursor, and for overlaying graphics.
The general rules are the same for the image display, and are repeated here:
- control will be by GUI inputs (ie, graphical buttons, text boxes)
and/or mouse;
- mouse button functions will be continuously updated/displayed;
- default parameters should be accessible for user modification and
resetting to nominal values.
Quick-look operations must be easy to use. As they may be less
self-explanatory than the image display operations, there should be
an "on-line help" button to provide detailed descriptions of the procedures
and operations.
There must be hardcopy capability (with proper stamp) for any plots or
text output generated by the quick-look operations.
List of Quick-Look Operations
General Image Analysis:
- Row/column plots; average over many rows/columns;
- Averaged row/column plots for a user-specified region (mouse drags
out box to specify the region).
- Plot along an arbitrary vector. These coordinates should appear
on the plot abscissa: absolute x, absolute y; relative distance from start.
The plot must be labeled with vector (x1,y1), (x2,y2) length and position angle.
The width of the data strip along the vector (eg, 1,3,5 pixel) should be
user-selectable;
- Print out array of pixel values in small (eg, 9 x 9 pixel) box;
- Statistics within a box; user drags out box of desired size.
The list of possible values should include:
number of pix mean rms about mean sum
nx,ny of box median min max
location mode coords of extrema
percent of values above/below 3,4,5 sigma
It should also be possible to see a histogram, and to select the same stats
above with/without (a) sigma clipping and (b) "masked" regions (eg, bad pixels)
excluded.
Direct Image Analysis:
- Centering on cursor-selected feature (positive or negative).
Different algorithms for centering are: centroiding; edge-weighted centering
of features of known width; location of peak (positive or negative) pixel.
All three (and other?) center values could be displayed. For interacting with
other operations (eg, PSF analysis below) the centering algorithm
could be user-specified.
- Radial profiles;
- PSF/object analysis: sky, peak, FHWM, rms to gaussian fit (or other PSF
fit), S/N estimate, approximate magnitude (based on standard zeropoints or local
calibration -- see Note 1).
There should be a "clip CRs" option in this analysis;
- Cursor readout of celestial (and other) coordinates, including updating
the coordinate zeropoint by marking reference objects (this means
there must be a "private" WCS for this function);
- Ability to measure distances (a "ruler"): drag out vector, display length
in pixels and arcsec, and angles (wrt image, north, horizon).
- Various graphic overlays:
- circles (of specified radius);
- coordinate grids (eg, RA-Dec; Alt-Az);
- catalogued object positions;
- slitmask outline.
Spectral Image Analysis:
- Cursor readout of approximate wavelength (by individual slit)
plus celestial coords along slit. Slit/object ID must also be displayed;
- Calculation of approximate redshift upon identification of a spectral
feature, and overlay of rest-frame wavelength scale;
- Overlay of object positions;
- Statistics of spectral line profiles: FWHM, rms to gaussian fit, skewness;
- Simple spectral extraction (with options of displaying object+sky,
object, and/or sky). Must include a S/N calculator. Extraction should
also list FWHM of object along slit. Traced location of object peak
(or center of object window) as a function of wavelength should be
shown on image display (see Note 2).
Calibration Analysis
These operations may be included as separate tools, rather than part of
quick-look. They are included here for completeness
- Analysis of stellar images on a focus frame;
- Analysis of images to compute gain and read-out noise.
---------------------------
Note 1: "Standard" magnitude zeropoints for typical conditions must be
tabulated for each filter. Coupled with gain and exposure information (and
filter ID), this enables quick estimate of throughput and/or transparency.
However, we should also have the ability to specify "private"
zeropoints appropriate for, eg, a particular image taken under poor
conditions. Under these conditions, a star of known magnitude is selected
and a zeropoint is calculated, and added into a "user" section of the
zeropoints table.
Note 2: Quick spectral extraction may be best provided by easy
access/interface with more sophisticated spectral extraction routines
in the automated spectral reduction package.
Design Notes
The number of different quick-look operations suggests that we should use
keyboard inputs for selecting various operations. Examples: "r" keystroke for
radial plot, "c" for column plot, for display of coordinates, etc.
Andrew C. Phillips / Lick Observatory
Last modified: 07 Mar 96
phillips@ucolick.org