SGP Log Viewer with new HFD log

Sequence Generator Auto Focus
1

Hi.

Since I upgraded to SGP V4. The new HFD method will help a lot than the PinPoint function, and the curves runs good.

But when I uses latest Log Viewer, it gets empty result :frowning:

2

Please attach a log file and I’ll have a look.

Mikael

3

Thanks Mikael.

I’ve shared a recent log on Google Drive:

Regards

4

@Jared, @Ken

It looks like the message with the HFD value for each focus step is missing in the log file when running AF in ASTAP mode.

In normal SGP mode it looks like:

[03/18/20 01:08:14.206][DEBUG][Main Thread][AF;] Calculating HFR with sample size: 5
[03/18/20 01:08:14.314][DEBUG][Camera Thread][AF;] Auto focus HFR calculated at: 5.32
[03/18/20 01:08:15.335][INFO][Camera Thread][AF;] Incremental Fit Result (1): quality=%; a=0; b=0; c=0;

while in @esazx log file it reads:

[11/18/20 21:46:29.629][DEBUG][Main Thread][SQ;AF;] Star list contains 123 stars…
[11/18/20 21:46:29.629][DEBUG][Main Thread][SQ;AF;] Calculating HFR with sample size: 3
[11/18/20 21:46:30.497][INFO][Camera Thread][SQ;AF;] Incremental Fit Result (2): quality=%; a=0; b=0; c=0;

I couldn’t find the HFD value logged elsewhere.

Thanks, Mikael

5

So it’s not recorded :frowning:

who do I need to report this issue?

6

The mesage above was addressed to Jared and Ken, so we just wait for them to respond.

7

II’ll address this. It will come out like:

SGP:
Auto focus HFR calculated at: 5.32

ASTAP:
Auto focus HFD calculated at: 5.32

PinPoint:
Auto focus FWHM calculated at: 5.32

It will be out in the next SGP4.

Jared

1 Like
8

Also adding one for the validation frame:

Auto focus validation frame calculated at: 1.23

Jared

1 Like
9

And I’m also adding in the star count found for these if that is helpful:

Auto focus HFD calculated at: 7.21 with 176 stars

Jared

1 Like
10

Hi Jared,

this is great. These enhancements will facilitate to compare the AF results using SGP HFR and ASTAP HFD for AF metric.

Bernd

11

Looks very good to me. I’ll make the changes as soon as possible and issue a new version of the AF Logviewer.

Mikael

12

Mikael,
I’ve only just found your Logviewer, but what a useful piece of software. Thank you.
Terry

1 Like
13

Hi Mikael,

Just wondering what is the outlook for an updated version of AF Log Viewer with support for ASTAP HFD metrics?

The most recent version on SourceForge was uploaded c3 months ago so I guess you are still working on the changes. Or is there a more recent version elsewhere?

Also I am wondering if it might be possible to calculate a ‘correlation coefficient’** for the FocPos vs Temperature charts? I am thinking the higher the correlation the more we can rely on the number of measurements we have taken to give an accurate measurement of the temperature coefficient and filter offsets. If correlation is low even after collecting a reasonably large number of data points we might start to suspect the accuracy/repeatabilty of the focuser to return to exact position!

Mike

**

  1. Correlation measures the strength of relationship between two or more variables.

  2. When the nature of relationship between variables is known, it is easy to predict the value of one variable when the other variable is known.

1 Like
14

Hi Mike,

A new version 1.24 was sitting on my computer for some time, I was planning to do some more changes, but now uploaded it on Sourceforge. Please test it, as I haven’t been able to test it against my logs (still running Win 7 on my machine, need to upgrade to use SGP 4):

About the correlation measure, that is already included. For a linear regression, the R^2 is exactly the square of the correlation function. For a perfect correlation between temperature and best focus position it yields R^2=1, no correlation yields R^2=0. The R^2 value is computed and displayed in the title of the Position vs. Temp graph.

Best, Mikael

15

@mikaelA

Hi,

I’ve installed v1.24 and it seems to work OK. The AF curves look good but I’m not seeing a very good FocPos/Temp correlation - having said this I must admit the temp range is quite narrow which I think magnifies the impact of small FocPos changes. I think for good calculation of filter off-sets. coefficients, etc, a good number of samples is required over a decent temperature range.

If it helps you at all, here is a link to some SGP V4 logs with ASTAP AF metrics. I see the graph is showing HFR rather than HFD but the values displayed are HFDs.

Thanks for the update.