pdosmall.gif (6581 bytes)

The Minor Planet Observer
Palmer Divide Observatory

2007 Shoemaker Grant Recipient

Planetary.gif (18781 bytes)


External Links
(opens new window)

Buy MPO Software

MPO Support

CALL web site

Society for Astronomical Sciences




MPO Connections - Focusing

MPO Connections offers several different ways to focus both before and during a night’s run. For those using RoboFocus or Optec TCF focusers, temperature compensation focusing is allowed. This automatically adjusts the focus as the temperature changes. This feature is not available for focusers that do not have built-in temperature sensing.

Different Focusing Methods

There are many different methods for focusing. A very popular method is using the maximum pixel value in a star. If several images are taken, then the average or median value is found and used to compare against the value obtained at a different focus position. Another method is the so called "half-flux". This is a bit different from the Full Width Half Maximum (FWHM) in that it finds the radius of a circle in which half the flux from the star is inside the circle and the other half is outside. The average value from several images from each focus position is used to determine which way to move the focuser. This is the value used by all the focusing methods for Connections. Where the difference lies in focusing is how those values are obtained and applied.

Half-flux V

This method requires you to calibrate focus by running a special routine. The process starts significantly inside focus and moves out and then past focus. A plot of the HF value versus position shows two straight lines, a "V", converging to a point of best focus. The focusing routine uses the slopes of the derived lines in conjunction with readings taken at two locations to determine where focus is in relation to the V and attempts to move the focuser directly to the focus position. This method is good when using a focuser with absolute positioning, such as a RoboFocus or OPTEC, which use stepper motors.


The program takes an image at the current position and then moves IN and takes another reading of focus. This is used to determine which way to move towards focus. A series of images is taken until the focus starts to get worse. The focuser then moves in the opposite direction in the smallest step size until it goes just past focus. The focuser is moved one step (the smallest) back out. This is the best method when using focusers that do not have absolute focusing, such as a DC motor focuser like the LX-200 electric focusers.


This method works somewhat like the V method but without benefit of a calibration run. Using two focus positions and the HF readings, the program interpolates (and/or extrapolates) the focus position. This method is good when using a focuser with absolute positioning, such as a RoboFocus or OPTEC, which use stepper motors.

Auto-Focusing in Connections

Auto-focusing in Connections uses the method of the three above that you've selected in the configuration. Whether you invoke auto-focusing manually from the Connections focuser page or from within a script, the auto-focus method proceeds through these steps:

  1. Takes a 2-second exposure at 2x2 binning to find a suitable focus star. The 2x2 binning shortens the download time for this preliminary step. A "suitable" focus star is one that is not too close to the edge of the frame and has a minimum SNR.

  2. The exposure time is adjusted to produce a good SNR, with a maximum exposure of 3 seconds.

  3. A set of exposures is taken at the current position at 1x1 binning to find an average value.

  4. The focuser is moved IN and another set of exposures is taken. The change in the half-flux values is used to determine how to proceed, based on the selected focusing method.

On average, the fastest method is the Half-flux V approach since the correct position of the focuser can be approximated after the preliminary steps above. The other methods require more exposures. However, even they are relatively efficient if the focus is not too far off the ideal position. This would likely be the case after doing an initial focus at the start of a run and then using the AutoFocus command in a script to "tweak" the focus during the night. In tests using the Iteration method using a RoboFocus, the process generally takes 90-120 seconds, while the half-flux V method takes between 60-90seconds. These compare very favorably with other focusing software results.

For more about focusing in MPO Connections, download the Connections manual (PDF).


This site is best viewed at 1024x768 and higher

This page was last updated on 01/19/11 05:14 -0700.
All contents copyright (c) 2005-2011, Brian D. Warner
Trademarks are the property of their respective owners