The ASIAIR leverages an autofocus algorithm which is very simple, but enables the use of low cost focusers that may have backlash.
Essentially, the focuser will compute a V-curve as normal, but instead of returning to the bottom of the curve, an action subject to backlash, the focuser takes you out past the bottom of the curve, and iteratively works down to the bottom of the curve in smaller steps, until the HFR matches the min HFR recorded during the V-curve.
This would be a no-brainer to have as a feature and would enable much more consistent performance out of setups that do not have perfect backlash-free focusers. Yes I know there is backlash comp in focuser drivers, but this is hardly perfect, since this backlash comp number can change depending on the altitude angle of the scope.
@Bray_Falls1
Based on your description of the ASIAIR focusing routine, it seems to be taking a lot more focus frames than necessary. All precision mechanics (not just focusers) have to deal with backlash and those methods are well understood.
Autofocusing needs to be as efficient as possible to avoid wasting time. The following is an effective autofocusing routine. Note – the camera is always moving against gravity.
- Move the focuser outward to a point well outside of best focus.
- Move the focuser a fix number of steps (eg. 5, 7 or 9) inward taking frames.
- Use those frames to compute the best focus point. Camera is at its most in position.
- Move the focuser outward to the best focus position PLUS a backlash amount.
- Move the focuser inward to the best focus position.
The camera arrived at the best focus position moving in the same direction (against gravity) as it was moving when the focus frames were taken. This produces a very accurate focus position even with a somewhat poor quality focuser.
The number of backlash steps is not critical – it just needs to be greater than the actual backlash of the focuser.
Charlie
Yes I understand, but in practice this simplified version is not how it actually plays out. The load on the focuser changes depending on the sky angle, and when combining things like temperature change, mirror flop, stiction, etc, a “one size fits all” backlash comp step doesn’t actually work.
By computing the best HFR and working into the focus point with a few extra exposures, you can avoid all of these problems. To me, this is worth paying an extra 5min of AF exposures per evening to be absolutely sure focus is perfect. Im just saying having this as an optional routine would be a good idea.
For example, I can almost never hit perfect focus consistently with my rokinon 135 and ZWO eaf system in SGP, due to variant backlash. But inside the ASIAIR, it is perfect every time. There is an extra time cost yes, but it makes it work. And it makes me not want to use SGP.