Guide Scope versus Off-Axis Guider analysis

– OR –

How I Learned to Love Hate Differential
Flexure in my Newtonian Reflector

I've long suspected my imaging rig of having serious Differential Flexure problems, particularly when imaging through my Newtonian. Here the term Differential Flexure refers to flexure of the the Imaging Camera (Canon Rebel XT [350D] DSLR) with respect to the Guiding Camera (Philips ToUcam webcam). This Differential Flexure could have any of a number of mechanical causes:

  1. Sagging of the Imaging Camera within the Newtonian's Focuser
  2. Sagging of the Newtonian's Focuser's draw tube
  3. Sagging of the Focuser's point of attachment to the Newtonian's tube
  4. Flexing of the Newtonian's tube itself
  5. Shifting of the Newtonian's primary and/or secondary mirror(s)
  6. Flexing of the mechanical coupling of the Guide Scope's rings to the Newtonian's rings
  7. Sagging/shifting of the Guide Scope in its rings
  8. Sagging of the Guide Scope's Focuser's draw tube
  9. Sagging of the Guiding Camera within the Guide Scope's Focuser

and probably several other possible sources. My gut tells me that of all of these items, #3 and #4 are the dominant culprits. That is, the weight of the DSLR causes the Focuser to sag/flex/deflect with respect to the main tube, and may well cause the main tube itself to flex. Of course, this "movement" is not seen at all by the Guide Scope nor, therefore, by the Guiding Camera, which merrily and happily continues to keep the Guide Star centered in its field. While so doing, however, the actual field "moves" within the Imaging Camera, due to this Differential Flexure, thereby causing elongated stars and therefore poor images.

I devised a simple experiment to determine whether the poor guiding/tracking I was getting was due to Differential Flexure, or whether it was due to just having a cruddy mount. To do this, I tried using an Off-Axis Guider, a gizmo which eliminated the need for a Guide Scope and instead diverts a tiny bit of light to the Guiding Camera that would otherwise be going to the Imaging Camera. The important point is that the Guiding Camera and the Imaging Camera are both "looking through" the main scope; i.e., the Newtonian, so any mechanical flexure will be "seen" by both cameras equally. That is to say, if the Newtonian flexes during an exposure, the Guiding Camera will see this flexure instantly, and will therefore begin sending guiding commands to the mount to try to correct the problem. That's the theory at least.

Well, based on my tests, this theory holds up pretty well. Compare these two animations, one created from an image set using a Guide Scope, and the other using the Off-Axis Guider. (Thanks Doug for letting me borrow the Off-Axis Guider.) In both cases, the images were calibrated (dark- and flat-frame corrected), registered (aligned), white balanced, and stretched using a combination of IRIS and JimP, and then assembled into animations using ImageReady. I recommend watching one of the full 8-frame animations several times through before doing likewise with the other one (rather than quickly glancing back-n-forth between the two).


Off-Axis Guider
γ Cygni region
8-frame animation
10 minute exposures @ ISO 400

Guide Scope
NGC 7380 region
8-frame animation
8 minute exposures @ ISO 800



Notice how consistently tight the stars are in the first animation, the one using an Off-Axis Guider. Conversely, notice how (often) poor the guiding is with the second animation, the one using a Guide Scope. Notice also that the exposure length using the Off-Axis Guider was actually longer than the exposure length using the Guide Scope. (Generally, the longer the exposure, the greater the opportunity for something to go wrong!)

There's really no comparison here. This proves to me that my mount is perfectly capable of yielding very nice images. What remains is for me to either: (a) switch to using an Off-Axis Guider; or (b) seek out and eliminate all sources of Differential Flexure. Using an Off-Axis Guider can be difficult and cumbersome, so eliminating Differential Flexure is high on my list of priorities. Help anyone!? <grin>