With version 1.0.5, StarSieve now includes a new and exciting tool that
provides a map of quality of focus across the entire field of view of the
camera. At a glance, you can see if the camera is tilted at the focal
plane, or if the optics, focal reducers, or field flatteners , or other
accessories are causing significant field curvature resulting in focus
shift across the image.
Here are a couple of images generated from actual camera frames. The first shows a normal field-curvature variation of focus. Darker areas are in focus, brighter are more defocused. The second image is an example of a map produced by a slight tilt of the camera in a 2" focuser holder:
The idea of a curvature map was first suggested to me by Mike Cook, who even prepared a sample screen of what it should look like. I had some similar ideas in mind for future releases, but Mike prompted me to work on this sooner, and I'm very excited by the results. Thanks Mike for the suggestion and the help in testing of these enhancements!
Curvature Map is a plot that assigns various colors to the quality of focus at different points in the image frame. StarSieve extracts thousands of stars from each image and computes their FWHM. Then, a polynomial function is fitted to the distribution of FWHM values. This is then plotted using varying colors, as follows:
Black - lowest FWHM
Blue - slightly defocused
Green - more defocused
Red - significantly defocused
For example, in the first image above, the center of the image is best focused (lowest FWHM). The left side of the frame, showing some pink/red coloration is the most defocused compared to center.
Curvature map is a good way to judge focus variations throughout an image. By moving the mouse around the map, you'll see a popup tooltip that will display the FWHM variation at that spot.
Curvature Map can also tell you whether the frame was focused at the center, or at the edge of the field. Here is an example of a frame where best focus is to be found at the edges of the image. Note that the same optics were used to produce this field as the image #1, but the focus was adjusted so that the stars at the center of the image were more out of focus than at the edge:
Curvature Map can be calculated from a single image, or from a collection of
images. Select the images you want to use for creating a Curvature Map (see
hints below for how to pick images for Curvature Map calculation).
Then, click on the Curvature Map... button at the bottom left of the StarSieve display:
When StarSieve completes the analysis of the selected images, an Image Viewer window will pop with the calculated Curvature Map in color. Try to move the mouse around the map to see how much the various areas of the image differ from perfect focus (note that the FWHM values displayed are in pixels):
Note that clicking the right-mouse button anywhere in the image brings up
the same pop-up menu with zoom-level choices, as well as, a new option to copy
the image to clipboard. This allows pasting the image into a paint program or
into a document. Additional option included in the menu is the ability to define
the display range. This option gives you the choice of defining what FWHM value
is mapped to black (Minimum setting) and what value to the brightest red (Minimum+Range).
These values are automatically set for best dynamic range display when the map
is first displayed. If you have multiple maps open at the same time, you can
apply the same display range settings to all by clicking on Apply to All
button. This will allow a direct comparison across multiple maps by using the
same stretch parameters for all images.
What is a good minimum/range setting? If the median FWHM value is, say 3 pixels,
Ideally, the following criteria should be applied to the multiple image selection to produce an accurate curvature map:
1. All images must be of the same size
2. Images should be acquired during the same session with the same optical configuration, preferably with little or no focus variations between the frames. The images do not have to be of the same area of the sky, but should be close-by to avoid the effects of mirror-flop and flexure on focus.
3. Use StarSieve to measure images, and pick only ones that are very close to each other in FWHM and Aspect Ratio.
4. Use images with a good sprinkling of unsaturated stars throughout the image frame. For example, a globular cluster would not be a good image, as all the bright and bloated stars will be concentrated at the center of the glob.
5. Use StarSieve dark frame reduction, or images that are already reduced by dark frame subtraction.
6. Use 3 or more consecutive images for building a more accurate curvature map. This increases signal-to-noise of the measurement and produces a more accurate representation of the field curvature.
7. Use good sampling: if FWHM values are well below 2 pixels, the FWHM measurement will not be performed accurately, and the resulting curvature map may appear distorted due to uncertainty in determining the correct stellar profile. One way to deal with undersampled, short focal length configurations is to defocus the telescope slightly before taking a test image. This will result in greater FWHM values that can then be more accurately measured by StarSieve.
If you'd like to measure a single image, it should contain an evenly spread-out star field of unsaturated stars, and should be dark-frame subtracted for best results.
The example maps above were generated with various configurations of a TEC140 refractor. Here is an example of a 10" SCT with a mirror flop problem, and suffering from tube cool down issues:
Back to Top