Beta 353 Running Great

Sequence Generator
1

Beta 353 is running perfectly for me. I have run it for the past 2 nights for 11 hours each night on 2 scopes. 3 targets, 3 flips, about 90 total focus runs. Everything running without a hitch. Last night was particularly challenging. I had two spells of overcast weather. Recovery mode kicked in and when the clouds cleared, images were in focus and continued as if nothing had happened. I have temperature compensation enabled, and Smart Focus enabled. Smart Focus brought both scopes, which were way out of focus, into perfect focus at the start of the session.

This product is just about ready for prime time. All issues I have encountered during this Beta phase have been fixed. This is far and away the best version of SGP I have ever used.

A nice enhancement that has not been mentioned is in Beta 353. The auto-focus pack directories have a new naming scheme which makes it clear if some focus reruns were required. Here is an example from last night:
Capture1
Focus runs 1 through 7 were fine. Run 8 has 3 separate runs because of focus failures, so they were numbered 8_1, 8_2, and 8_3.

Also the focus graphs have a new naming scheme which makes their names unique. This makes it easy to copy all the FINAL focus graphs into one directory for easy perusal. Here they are for the runs shown above. The name even tells us that the Quality for the runs 8_1, 8_2, and 8_3 was very poor, a cloud came through, so only 3 tries were attempted.
Capture2

The new focus routine is working dramatically better than the old one did. The Quality factor returned by the new Quadratic fit routine allows the program to be able to reject those bad focus runs and not screw up the current good focus.

I have a couple of suggestions:

  1. It now only outputs a JPG for starting with the 5th image. That is good. However, if there are fewer than 5 images because the routine needed to start over at a different starting position, the final graph image should be output, so we have at least one JPG of the graph for each run segment. Currently no graph image is output, so we don’t know what happened with that run segment.
  2. I would personally like those images to contain the word FINAL so that they will all be fetched into my collection directory.
2

Thanks for the update. 358 should see all runs emit a “FINAL” or “NoFit” image (almost all runs… there are still some failure types that won’t, but the most common failure (bad fit, recenter, restart) will make one. Disclaimer: I read this right before I was about to release so it wasn’t exactly what the professionals call “tested”. If it doesn’t work right, I’ll clean it up.