Paul Deeter:
I use blink to weed out the obviously bad (really bad) frames prior to WBPP. I don't feel this step takes that much time. However, I'm not capturing 500+ light frames either. Just my two cents. Clear skies, Paul.
John Hayes:
Sometimes I might have as many as 800 x 130MB files and I consider blinking to be an essential tool to weed out frames that weighting and filters might not deal with properly. Passing high thin clouds or other effects can cause issues that shouldn't be included in the stack. The blink tool is my number one most used tool in PI.
- John
Reg Pratt:
I'll use blank to weed out periods of clouds and then let APP do the rest. Running 2 scopes at short subs means thousands of subs at the end of each session. Far too many to blink it all.
I find that even with a big number of frames, it is a fairly fast method.
I find that even with a big number of frames, it is a fairly fast method.
*
Depends: If you have 500+ subs and each with 120MB (full frame) then even the loading into Blink is not negligible assumung an 'average' PC is used.
CS, John
Indeed, you could play around with SS formulae and conditions to auto exclude low weight assigned frames and see what works. Although, if you have invested time to capture that many frames it is almost inevitable to invest a LOT of time for pre and post also, just to be sure you are taking full advantage of your dataset
andrea tasselli:
Kidding aside I don't blink because it is far easier to sort subs according to number of stars detected and reject those with an obvious drop in detected ones (clouds coming through...), as shown here:
You can do other sorting for fwhm, weight, eccentricity and so on and drop them from the stack as needed.
).
Ben Koltenbah:
An interesting thread. I'll chime in with my two cents (now four cents with inflation) worth:
I Blink. I blink often and early.
With so many subs to sift through, I need to quickly flip through them to first remove the egregious (see P.S. below) ones with airplanes, errant clouds, owls, wife turning on the backyard light, etc.
At times past, I have attempted to skip the Blink visual check and solely use SubframeSelector (SS), and I was surprised to find that some of the frames SS passed were those that I would definitely have rejected. With the many parameters available on which to rack, stack and sort the frames, this is not surprising. I am finding, however, that the newer parameters like PSFSignalWeight and such are doing a better job than some of the older ones I used before.
Given my past issues with SS, I usually open a second instance of PI on which I load the same list of files into Blink as I have into SS on the main window. This allows me to do a visual spot check on the frames while I adjust the selection criteria.
All the Best,
Ben
P.S. Whoops! The last guy used "egregious", so I'll change mine to "painfully obvious to the most casual, distracted of observers".
Ben Koltenbah:
An interesting thread. I'll chime in with my two cents (now four cents with inflation) worth:
I Blink. I blink often and early.
With so many subs to sift through, I need to quickly flip through them to first remove the egregious (see P.S. below) ones with airplanes, errant clouds, owls, wife turning on the backyard light, etc.
At times past, I have attempted to skip the Blink visual check and solely use SubframeSelector (SS), and I was surprised to find that some of the frames SS passed were those that I would definitely have rejected. With the many parameters available on which to rack, stack and sort the frames, this is not surprising. I am finding, however, that the newer parameters like PSFSignalWeight and such are doing a better job than some of the older ones I used before.
Given my past issues with SS, I usually open a second instance of PI on which I load the same list of files into Blink as I have into SS on the main window. This allows me to do a visual spot check on the frames while I adjust the selection criteria.
All the Best,
Ben
P.S. Whoops! The last guy used "egregious", so I'll change mine to "painfully obvious to the most casual, distracted of observers".
Chris White:Ben Koltenbah:
An interesting thread. I'll chime in with my two cents (now four cents with inflation) worth:
I Blink. I blink often and early.
With so many subs to sift through, I need to quickly flip through them to first remove the egregious (see P.S. below) ones with airplanes, errant clouds, owls, wife turning on the backyard light, etc.
At times past, I have attempted to skip the Blink visual check and solely use SubframeSelector (SS), and I was surprised to find that some of the frames SS passed were those that I would definitely have rejected. With the many parameters available on which to rack, stack and sort the frames, this is not surprising. I am finding, however, that the newer parameters like PSFSignalWeight and such are doing a better job than some of the older ones I used before.
Given my past issues with SS, I usually open a second instance of PI on which I load the same list of files into Blink as I have into SS on the main window. This allows me to do a visual spot check on the frames while I adjust the selection criteria.
All the Best,
Ben
P.S. Whoops! The last guy used "egregious", so I'll change mine to "painfully obvious to the most casual, distracted of observers".
I blink. Ben hit on the reason why.... SS often passes images that I would reject and rejects images I would pass. Analysis software for FWHM in particular is confused by astigmatism and eccentricity. So unless you have a perfectly corrected field for the entire chip (which most of us do not have) a visual inspection seems to be the most robust. Even with my poor eyes, I can blink through images very fast to determine roundness and shape. I use a very fast scope, so I might easily have 500 or more subs in a given project. I always run through blink twice. First pass is zoomed out to see if clouds rolled through. Second pass is at 2:1 to look at star shape. I can blink through a couple hundred images in just a few minutes. Anything I miss will be an outlier for rejection algorithms...
Subframe Selector takes a long time, and I dont trust the results... I never use SS.
Hi Chris, thanks for chiming in and confirming that SS is not doing what would be obvious when doing manually.
Interesting you state that even with 500 subs you blink through in 2 rounds in just a few minutes. I have serious problems with 500+ subs at full frame (120MB). I use an iMac with 64GB RAM.
Thx, John