Call for Aberrant Data

Hey folks,

Got messed up data? Sorry. Can we have it?

Any “aberration” is ok, but we are currently focused on optimizing the eccentricity measurement so if you have an image with a bunch of pill stars, we’d like to “harden” our methodology with as much data as possible.

If you have a place you use to share data, that’s great. If you need one, respond here and we’ll send you a dropbox link you can use.

Thanks in advance!

Here is my contribution.
These frames were discarded because having an eccentricity value above 14, measured with CCDI. Have also included a screenshot of CCDI so you have reference values for each frame.
Hope this helps. Looking forward to SGP’s eccentricity calculation.

1 Like

Sorry, can you explain this? Eccentricity is a value between 0 and 1 where 0 is a circle. Do you know what eccentricity of 14 means? Is it maybe an altered HFR / HFD… meaning that that measure of eccentricity is, with respect to the ellipse, the HFR / HFD as measured from the centroid along the semi-major axis?

This number is called “aspect ratio” in CCDInspector and here is the definition from the user manual.

Aspect Ratio:
Aspect ratio represents how much out-of-round the star image is. It is the ratio of the longest axis to the shortest axis of the star profile, expressed in percent. A number below 20-30% represents a pretty round star. A number of 0% represents perfectly round star, but you will most likely not see this in real images due to noise and measurement uncertainty.

I discard everything above 14-15%, that’s when eccentricity becomes noticeable.

1 Like

So, the aspect ratio is a mean value then? For the entire image or a region? Seems like it would be useful to have eccentricity measurements on the entire image certainly, but maybe also just on the edges to gain insight around misalignment aberrations like mirror tilt, etc. In other words, it seems like seeing a real time comparison of the edges with respect to the central area might give fairly rapid insight into the cause of the eccentricity.

  • Edge eccentricity > central eccentricity means some kind of misalignment or image train issue, etc
  • Both edge and central eccentricities high (and roughly the same) means that there was a capture issue (guiding issue, etc)

Maybe others too. I’m just trying to figure out which presentation of this data is most useful for capture (since that what SGPro attempts to do). Thoughts?

No mention in the manual if it is a mean value over the whole frame or only a region. I think it is the whole frame because CCDI has the ability to generate eccentricity maps for more details.

Your idea of a mean value in the center and on a user predefined edge area (like the “crop autofocus frames” option), sounds good. It gives addittional information that may indicate the eccentricity cause, call it, guiding, tilt, spacing, mirror flop, or just OTA performance. Though, I wouldn’t use this function to correct the problem. Theres is other software like CCDI, ASTAP or PI with complex funtions, like eccentricity maps, etc, for that.

I am with you on that idea.

PS: the eccentricity value should be visible in the history diagram

Thanks. One question I forgot to ask concerns the wording of the latest request for this feature:

Can you explain the specificity of the request? Meaning, why 3x3? Is the intent to have a tool that is smart enough to detect issues on connecting edges of the mosaic? Or something else that is not occurring to me?

No intention for smart tools or calculations here. This is just a quick visual inspection of the “complete” zoomed frame corners, edges and center for picky people to check elongated star or other optical aberrations (coma, etc). It is typically 3x3, because you want to check the 4 corners and the center.
Now that you mention… adding HFD and eccentricity value on each mosaic would be a cool feature, although you may not always have stars in each mosaic…

PS: Do you never zoom in and scroll to every corner to check your frames?

I don’t. I can’t say that I have ever been interested in doing this because I don’t know what it would inform for my particular set of gear (refractor + well-matched flattener). There is nothing I could do about any eccentricity here and general HFR deviations do an adequate job (for me) of identifying bad images from environmental issues. I do, however, see that, for folks who can do something about this (usually any kind of folded light path, but also maybe somebody trying non-matched flatteners), that it may be useful. Also, I think I’d like to use it as an additional filter toward HFR calculations where stars with elongation above some threshold would just be discarded.