Focusmax workflow

I use focusmax for initial focusing. It uses Maxim for controlling the camera. Then I disconnect maxim and launch SGP. Subsequent focusing is done within SGP.

1-does anybody use a similar flow?
2-could SGP do the following:

  • disconnect camera
  • invoke focusmax which will connect to the camera(not sure if it can disconnect via a script, will have to check that)

Or is this going to be obsolete with the improved focusing in future versions of SGP.

The truth is, I have been putting off learning more about focusing within SGP and instead relying on focusmax.

Thank you

SGP doesn’t have anything like this built in. You could leverage the SGP API and write a script to do that. Then you’d call the script from within SGP and have it disconnect the camera, run focus max and reconnect the camera. But the API doesn’t have any provisions for starting a sequence. The API is mainly made as an interface to the devices, not an interface to the automation.

You might just give this a shot in SGP. Here’s my workflow:

  1. Set focus points per filter in filter control panel (Define Filter List)
  2. Select the filter I want to focus Click the “Focus for Lum” or whatever the filter is on the Focus control panel
  3. Capture frame, see how focus is. If it looks relatively decent (Stars are relatively in focus) then run Auto Focus
  4. Continue running Auto Focus until a nice V is created.

In 2.4 things will be similar. Except that step 4 won’t require as much interaction and step 3 will allow you to have worse focus.

I would take a look at auto focus in SGP as it’s quite good (if I do say so myself). The only caveat is that you need to start in decent focus as the current focus mechanism relies on a fairly narrow window for focus. Really that’s the main change in 2.4 is that the focus window is larger and more deterministic.


Jared, Any idea when 2.4 will be released? Looking forward to the auto focus changes!

Hopefully a beta will be out in a week or so. Some fairly important features were changed (Phd2 integration) so once those are verified there will be a beta.


Got a chance to test the focusing last night. got a nice Vcurve, but
increased the number of steps to 10. One question though, I was having an
issue with FocusMax when using my DSLR where it was having problems dealing
with the Bayer matrix. I was getting very high HFD values, apparently, they
recommend enabling 2x2 binning in the configuration to deal with DSLR bayer

Yesterday, in SGP I measured the HFR on a 5 seconds exposure and got a
result of about 2.19. I then took a 10 minutes exposure and got a 7.89.
Focus did not change at all. Does SGP take into account the bayer matrix
while computing HFR?


SGP uses the raw image data and does not account for the Bayer matrix.

Even if your HFR is artificially high, as long as they’re relative to one another your focus should still be fine.


Thank You. But this does however mean that I cannot rely on HFR measurements to determine “absolute” image quality. The frame grading tool will also provide “incorrect” values as well, but correct relative to each other. I will have to test this more to make sure that the behavior is indeed repeatable and consistent. The fact that I saw 2 different values based on exposure duration kind of concerns me. Let me collect more data.

Thank you.

Well on an object by object basis the HFRs should be relatively close. I’m not sure why you would have that large of a swing unless your AF images were taken at a higher ISO and the focus metrics where attaching to some hot pixels? Obviously a 10 minute image is going to have a higher SNR than a 5 second one. That’s really the only thing I could think of that could affect it. Also since DSLRs don’t necessarily have a linear response range that could be causing issues too.


I used FocusMax for years (in fact, a talk I gave years ago about manual focus got the authors to thinking about developing it). Recently I have been using SGP focus and have to say that overall I like it better. SGP can focus in place instead of going looking for a star, which makes much more sense. It also looks at more than one star which eliminates a lot of issues as well (like unintentional use of close doubles). I tend to just run a visual focus first within the SGP control panel to get close and then do a manual run or two of the SGP autofocus to nail it.

FocusMax does work better (or at least more quickly) on very narrowband filters due to it’s ability to look for a bright focus star. No way around this for SGP that I can see…

Take Care

I use my lum filter for all focusing and use the “Auto adjust focus per filter” option which uses the filter focus positions to account for any focus shift due to filter change. Focusing through 3nm narrowband filters is not fun…I try to avoid it :smile:


I guess I have just never trusted offsets. Probably just superstition.

As long as they’re setup correctly they work extremely well and are highly repeatable. Ken and I both use them. I think a handful of our users that are shooting narrowband use them as well.