Discrepancy between Focus Max report and ACP reporting

[Jeff Lewis]

I had FocusMax fail more than once while testing new settings. I expected the failure, FocusMax reported it correctly but ACP reported success. See the attached log and screen shot.

Note that FocusMax Log reports the AcquireStar Failure while ACP reported success.

-Jeff

[Bob Denny]

I'll have to at the focus max API to see if there is a way to pick up those errors from Acquire star. Or maybe I'm not checking the status in the acquire star code. I am right in the middle of it now so I will look tomorow.

[Dick Berg]

Hi Jeff, I would wonder how it could possibly be possible that, in searching within a 40 degree square of the catalog, FocusMax could not find a single star. This is what your testing was about? I understand that ACP and FMx are out of sync here, but ACP by itself would surely find focus stars in such a large area, regardless of the catalog used. Have you been able to tune up your settings so that FMx doesn't fail?

Is the data reported by ACP correct? or is that data from some previous focus run? I was just curious.

It does look like there could be an API communication issue.

[Jeff Lewis]

Dick,
I was testing using different catalogs. I had been using GSC and have never had any issues. The only parameter I changed was the catalog PinPoint uses, tested USNOA2 and UCAC4.
Oddly I did a test solve with each of the catalogs on the field I was using for focusing, and they all solved. Then running the focus routine via ACP resulted in the issue I reported.
I did not test directly with FocusMax, removing ACP from the equation would reduce the variables, and I will do that next.

During the test I changed the catalog settings in ACP, PinPoint and FocusMax, which might not have been necessary but done, again to limit the variables.

The data in the ACP log file is for the last Luminance focus run that succeeded (ie HFD value and focus position), the acquire star fail was for an attempted V filter focus run.
Note the ACP log indicates Autofocus through 4 but the data displayed is filter 0

Here is a log from a good run, before changing catalogs

[Bob Denny]

During the test I changed the catalog settings in ACP, PinPoint and FocusMax, which might not have been necessary but done, again to limit the variables.

I assume you meant Visual PinPoint? Settings in different host programs do not interact, by design. The PinPoint engine has absolutely no internal memory of parameters/settings. They are 100% provided by the host program. So only FocusMax settings have any influence over its operation including its usage of the PinPoint engine.

More later...

[Bob Denny]

One thing I noticed: The 18-Jun log shows FocusMax's log info embedded into ACP's log. On the 19th it is gone. Did you turn this off? If so, why (just curious)?

The simple code in ACP for detecting FocusMax AcquireStar failures follows the FocusMax docs. ACP calls AcquireStarAsync() to kick off the process, then ACP monitors AcquireStarAsyncStatus until it changes from -1 (in progress) to either 1 (success) or 0 (failure). After that it reports. To see this code in AcquireSupport.wsc search for for

' (3) Start AcquireStar Focus

Actually, ACP goes further, requiring both AcquireStarAsyncStatus to be 1 (success) and also the HalfFluxDiameter to be non-zero. Both must be true for ACP to report success.

Conclusion: FocusMax is reporting success when it actually fails.

[Jeff Lewis]

Oddly none of the FocusMax log entries for the failed runs showed up in the ACP log. FocusMax is configured to send the log to ACP and that has not been turned off.
Also, good to know about the independence of catalog settings across applications.

-Jeff

[Bob Denny]

I wonder why the log sending to ACP stopped between the 18th and the 19th. What version of FocusMax are you running?

[Jeff Lewis]

Running FocusMax version 3.0.8.14

[Bob Denny]

OK, I was wondering if V4 beta... thanks.