Better? Why? How does an arbitrary change to the data make it better?

In actual fact the current code has an undocumented feature where the full set of HFRs is averaged, the standard deviation is determined, values that are more than two standard deviations from the original mean are removed and a new 'average' is determined.
This is statistically dubious, not only does this reduce the mean but the amount by which it is reduced will depend on the data. Maybe extended objects will be included sometimes but not in others.

The median is at least a valid way to determine a central value.

What benefit does this have, other than be more complex?

From a look at the code the HFR array is already sorted so getting the median is easy.

If that is what is in the code then I would leave it as it is, in a binomial distribution 2 standard deviations within the mean would equate to 95% of samples, with those outside being discarded, so pretty much it is already discarding outliers then.

I'm not sure whether using an actual median value would be better in all cases, perhaps it would be nice to have the option to configure it either way.

I wonder if the HFR of stars on a field is always a Poisson distribution. There are probably not enough of them in a focus frame to determine that and use that statistical property.

I am trying to introduce PSF in the CCD simulator to better test the different focus algorithms. SExtractor support will be there too relatively soon. We also need to display an interface for the SEP settings.

We also need to better document the features in the tooltips that appear in the UI.

Anyone has unused work days to spare?

-Eric

Not really. This is standard, called sigma clipping. And the winsorizing I mentioned is an improvement on this for a better determination of the standard deviation. Both methods are quite well analyzed and accepted in statistics for removing outliers.

Something I heard of at work was that people were analysing the waste from gold mines to see how much residual gold was left and if it was worth reprocessing it using more modern methods.

Lots of analyses were done, and some came up with negative amounts of gold. You can't have less than nothing so they were set to zero and the results averaged - and there was enough gold to be worth processing. I'm not sure how much waste they got through before they discovered the error of their ways.

It looks as if this is just the same, one side of the distribution is being clipped and that will inevitably affect the average value.

Hi!

I am nowadays using full field focusing, and rather long exposure times to get good SNR (4-10 secs, depending on filter). But I do have one issue that I am unsure of how to handle:

There is about 250 tics offset between my V filter and the LRGB filters. Often, I start the night with a run of variables, using the V filter. Then I want to switch to LRGB imaging. THis is where my issue is. I have problems getting Ekos to handle the offset here, and find a good focus 250 tics less than it was on, on the last variable target. I have an initial step size of 300, but that is OUT, and I need to move inwards. If it started in, instead of out, this would not be a problem. As it is, it goes 300 tics OUT, and since it is off focus to begin with (not in the CFZ), I often end in a very bad focus situation.

I've tried using a short focus-job inbetween here, but it often does not help. It sure gets the right filter in place, but not the right focus position. And offset in the filter dialoge is not what solves this.

I use Polynomical, SEP for star identification. Any ideas on how to handle this better?

Magnus

Magnus

I'm not at a computer, so my memory may be wrong.... But I think you can specify the approximate offsets between filters in the filler focus widow, which I guess should be able to start you v filler in a better position.

Hi!

No, to the best of my knowledge (it's day now so I can't test until later), it does not do this. The offset there makes the focuser move a fixed amount of tics after focusing - this to make it possible for using another filter for focusing, which is not parfocal with the current filter. For instance, I use Lum to focus instead of Ha, but there is an offset of 100 tics between them. So after focusing with Lum, focuser moves 100 tics and shifts to Ha.

If I did that, I'd need to use the V filter as my focus filter, and then have an offset. Which is not what I want, for several reasons (among them, wanting to use Lum for focusing with my parfocal RGB filters).

If I set the offset but did not use V as my focus filter, I'd still have the problem of finding focus the first time when shifting from V to Lum.

What I need is a way to handle the fact that there is an offset between V and Lum that makes a shift from one to the other problematic.

Magnus

Not true. Yes, the scenario you describe is one of the applications, but basically the entries set up a mesh of relative focus differences. And they are automatically applied whenever a filter change happens, not only after focusing. If done properly, you can focus on any filter, then change to any other filter, and it should be (about) in focus. Definitely good enough to start an autofocus run...

So if you have 100 difference between Hα and V, focus Hα and then switch to V, it will shift the focus 100 units. If you focus V, then switch to Hα, it will change the focus by -100 units....

Ah, OK. I'll try it out tonight!

Thanks.

Magnus

Hello!

I've now tested Linear substantially for two nights, for my variable star shots. And it is quite obvious that I now use a slightly smaller aperture in measuring those frames. That is, focus is perceptibly better, not just very reliable. So - Thanks, Hy! This is a clear improvement!

Trad off is of course speed. It takes more steps than Polynomial. In my case, I compensate by focusing slightly less often, seems to work nicely.

Magnus