Xguiphot can be used to do various kinds of quick-look photometry: including checking photometric quality, determining good values for the aperture sizes, and making quick measurements.
To begin the tutorial type the following:
cl> xgex3
After the gui is actived the first image in the image list m92b1.fits is displayed in the image display window, and the gui is ready for user input.
Xguiphot can be used to check the photometric quality of a series of images of the same field taken at various times throughout the night.
In order to perform these checks properly we need to set traps for saturated stars, normalize for exposure time, and set the gain and readout noise parameters as follows:
1. move the mouse to the Phot menu
2. select the "Impars pset ..." item
Now move the mouse to the impars parameter editing panel and:
1. move to the imaxdata entry box
2. backspace over the existing entry
3. type in 32000 and press CR
4. set ikread and ikgain to "" in a similar manner
5. set ireadnoise and igain to 75 and 11 respectively
6. press the Apply button
Note that the user does not usually need to set ireadnoise, and igain because the correct information is usually in the image headers. However this is older data so these keywords are missing and values must be set explicitly. Users with non-NOAO data may need to alter the ikread and ikgain parameters to match their header keyword formats.
Now step through the 4 images using the n keystroke command in the image display window, while checking that the values displayed in the the impars parameter editing panel igain, ireadnoise, ietime, iairmass, ifilter, entry boxes are present and correct for each image.
Everything appears to be ok, so return to the first image in the list by typing the command :imnumber 1, and dismiss the impars parameter editing panel. Note that it also possible to move backwards and forwards in the image list by selecting the "Files panel ..." item from the Files menu and clicking on the desired image name, or by selecting the "Next image" and "Previous image" items from the Images menu.
Now we are almost ready to make a few trial measurements. Before doing so however bring up the photometry plots panel (so we can examine the results for individual stars) as follows:
1. move the mouse to the Results menu
2. select the "Photometry plots ..." item
or
1. move to the image display window
2. type the g keystroke command
Now move the mouse to a likely looking star in the image display window and either
1. type Spacebar
or
1. press the left mouse button
2. select the "Point and measure" item
Examine the object region display and the semi-major axis plot (the same as a radial profile plot for circular apertures). Does the default aperture radius look like a good value or is it too big? Press the Activate button under the major axis plot and expand the sky region by placing the cursor on the sky region and typing Y a few times to expand the plot. The default aperture looks too big to me. To change it press the Deactivate button and:
1. move the mouse to the Phot menu
2. select the "Photpars pset ..." item
Change papertures parameter from 15 to 10 and press CR, press the Apply button, and remeasure the last star with the Spacebar key. Repeat the previous instructions until you are satisfied that the aperture size is reasonable and dismiss the photpars panel.
Now look at the sky region and sky histogram. The sky annulus looks like it should be a bit closer to the star. However before changing it, check to see how flat the sky is as follows:
1. move to the skyplots button in the photometry plots panel
2. press the left mouse button
3. select the "Theta profile" item
The sky looks reasonably flat around the star so go head and change the default sky aperture as follows:
1. move the mouse to the Phot menu
2. select the "Skypars pset ..." item
Change the srannulus and swannulus values to 15.0 and 5.0 respectivley, hit the Apply button, and then move to the image display window and remeasure the star. These new values look good so dismiss the skypars and photometry plots panels and activate the photometry table panel as follows.
1. move the mouse to the Results menu
2. select the "Photometry table ..." item
3. type the i keystroke command to redisplay the image
or
1. move to the image display window
2. type the t keystroke command
3. type the i keystroke command to redisplay the image
Select one star to measure in the first image and measure it with the Spacebar keystroke command. Move to the next image with the n keystroke and measure the same star. Repeat for the remaining two images. Examine the results in the table. Note that the image names are too long to be fully written out, but that the first two images are B filter images and the second two are V filter images. The exposure times are short and the images were taken over the course of 20 minutes or so, so there should not be too much difference in airmass. The measurement repeatability appears consistent with the formal errors. Do you agree ?
In quick-look mode xguiphot can be used to determine the optimal photometry aperture sizes for doing stellar photometry.
To a first approximation the optimal aperture for maximizing the signal-to-noise ratio is approximately equal to the fwhmpsf.
To estimate the fwhmpsf:
1. move the mouse into the image display window
2. type :imnumber 1 to display the first image
3. move to a bright isolated star
4. press the left mouse button
5. select the "Show model plot ..." item
Examine the bottom of the plot and note the value of Hw:. I get a value ~1.88 for the star near X,Y = 145,288. This means that the fwhm of the psf is around 3.8 pixel and would be a good initial choice for a measuring aperture.
To measure the width of the same star on the next 3 frames:
1. type the n keystroke command
2. move the mouse to the same star and type x
3. note the value of Hw:
4. repeat steps 1-3 for the next two images
By comparing x, y, amplitude, size, and shape measurements and looking at the image, you should be able to see that the second V image appears to be somewhat trailed relative to the B images and the first V image, and that the seeing looks worse as well. The "optimal" aperture radius for all 4 images is somewhere between 3.5-4.5.
In some cases it is desirable to pick an aperture which includes "all" the light of the stars on a series of frames taken at different times through different filters, or at least enough of it to get past seeing and filter effects to the point where the correction is constant. One way to determines this radius is to take measurements in several apertures and see where the magnitude versus radius curve levels off.
Before beginning make sure that the sky annulus setting is reasonable. From the model fit plots above a good first value for the position and width of the sky annulus looks like 15 and 5 pixels. To set the sky annulus parameters:
1. move to the Phot menu
2. select the "Skypars pset ..." item
3. change srannulus to 15.0
4. change swannulus to 5.0
5. press the Apply button
6. press the Dismiss button
or alternatively
1. move to the image display window
2. type :srannulus 15.0
3. type :swannulus 5.0
Next change the the single aperture radius value to the range of values specified by 1:10:1, which means measure the flux through aperture radii from 1 to 10 in steps of 1 pixel. To make the change
1. move to the Phot menu
2. select the "Photpars pset ..." item
3. change papertures to 1:10:1
4. press the Apply button
5. press the Dismiss button
or alternatively type :papertures 1:10:1 in the image display window.
To make the measurements:
1. move to the image display window
2. type :imnumber 1
3. move to the Results menu
4. select the "Photometry plots ..." item
5. move back to the image display window
6. point to a star in the image display, e.g. X,Y = 145,288
7. press the left mouse button
8. select the "Point and measure" item
Note that only the largest aperture is overlaid on the object region in the plots panel. Now move to the plots panel and:
1. press the objplots button
2. select the "Curve of growth" item
3. press the Activate button
4. move the cursor to the "flat" part of the curve
5. hit the Y keystroke and examine the plot
6. press the Deactivate button
Note that although there are 10 aperture measurements there are only 9 points plotted in the curve of growth plot. This is because the curve of growth plots shows magnitude differences not the total magnitude.
Does 10 look like a good choice for the largest aperture? It seems a bit hard to tell from the plot. Scroll the text in the upper part of the plots panel to see magnitude printed as a function of aperture radius. What do the numbers look like ? Change the papertures string to 1:15:1 and srannulus from 15 to 20 and remeasure. Measure several more stars. For this image I think 10 actually does look like a good aperture radius. Do you agree.
Now go to the image with the worst seeing and guiding, i.e. xapphot$data/m92v2.fits, and repeat the above measurements. Note that the radial profile is not really smooth indicating that the star is trailed. This is confirmed by looking at the object axis ratio plot. However the star's curve of growth does appear to flatten out around 10-11 pixels like the first image. So what do you think is the best aperture to get good photometry of all the stars in the m92 images ?
Xguiphot can be used to make quick measurements of extended objects. Before beginning this example:
1. move to the image display window
2. type :images xapphot$data/gcluster.fits
3. type :objects ""
Select the object at 233,380 and zoom in on it by double clicking the middle mouse button.
To define an elliptical object aperture and a concentric elliptical sky annulus of the same shape for this object
1. move to the image display window
1. type the z key to delete any existing object list
2. type a to enter the object drawing menu
3. type a to select an object and annulur sky
4. move to the object of interest and type e
5. move to the semi-major axis and type e
6. move to the semi-minor axis and type e
7. type e to select an elliptical sky annulus
8. move to the inner radius of the sky annulus and type e
9. move to the outer radius of the sky annulus and type e
10. type q to quit
To measure:
1. move the mouse close to the center of the object
2. type the . keystroke.
The measurement apertures will be drawn on the image display. They should be very close to the marked apertures except for the possible effects of recentering. If desired turn off recentering by activating the cenpars parameter editing panel from the Phot menu, and changing calgorithm from "centroid" to "none" or, alternatively type :calgorithm none in the image display window.
The results are printed in the status line. They show that for the apertures I marked, the computed object axis ratio and position angle derived from a moments analysis are ~0.82 and 42.0 What are your values? What is the actual object and sky geometry ? To see this, activate the List panel from the Lists menu and examine the last object in the objects list. I see numbers like 0.80 and 46.0, so my aperture is not optimal. How do your measurement apertures compare with the computed geometry? Using a combination of the values in the list panel and the results in the status line, enter a new object into the list via the List panel "Enter object geometry" box and hit the Enter button. Remeasure it, by clicking on the new object in the object list, and clicking on the Current button.
Now bring up the plots panel from the Results menu. Select the "Display last result" item from the Results menu. Examine the major axis profile. Does it look smooth and flat? It should be if the elliptical aperture is a good match to the elliptical galaxy. How about the sky plots? Is the histogram reasonable? Are the major axis and position angle plot flat as they should be if the sky annulus matches the elliptical galaxy.
To quit tutorial III press the Quit button.