Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

In this Discussion

BPP Questions

Adam,

I have just completed the BPP videos in Fundamentals and I have a few questions.

In the Execution video, under Star Alignment, you select a reference image of 13 Oct whereas the target files are for 10 Oct.  Should the reference image be the same date as the set of target files?

Under what conditions can a Master Flat be reused over several nights?  At the very least the image train has not been taken apart allowing dust to enter, and the CCD temperature is the same for all flats and lights.  Any other conditions?  What about rotating the camera after meridian flip?  Can BPP distinguish different PAs like it does binning and filters?

Thank you.

Alan

Comments

  • Hi Alan,

    Regarding the reference frame- given that the differences in shifts between frames are dominated by dithering- you can choose any good image as a reference. If you happen to have a fraction of images that are shifted by a large amount- then probably not using them as a reference is wise. The wouldn't want the minority to control the usable area. 

    You are right. As long as the image train has not been affected- flats will continue to characterize the system. Removing the camera, losing the focus, changing the optics... etc etc will all end badly for having the flats match the data. 

    For many systems (it depends), matching flats with rotation angle is important. With regards to a GEM- if you are not rotating the camera to keep the same field orientation - then the one set of flats will do the job. If you rotate your camera 180 to keep the same orientation (likely to keep using the same guidestar) then flats for this configuration are usually necessary. 

    BPP does not on it own distinguish. There are a couple of "tricks" to make life easier. In fact I think there was something about this in the PI forum recently But to be honest in the case of a flip...you have two sets of data. So just running BPP two times for the two orientations isn't that bad. You just change the flats and load new data (keep the biases and darks of course).

    As you know, this is one area that BPP could be improved in. Having one addition bit of logic to assign a value to a key (such as something in the Header or filename) would make it possible to do what you want. 

    -the Blockhead


Sign In or Register to comment.