search

mmiesch / stereo-pB

leverage STEREO polarized brightness coronagraph images for CAT
0 stars 0 forks source link

STEREO pB -> Bt inconsistent #1

Open mikemarsh-mo opened 2 months ago

mikemarsh-mo commented 2 months ago

@mmiesch I think at least one of the issues that you might have been having with this experiment of comparing the raw STEREO img total brightness images with your attempt to recreate these is at the line below. When you are calling SECCHI_PREP to combine the images, it will be applying additional calibrations such as flat field, calibration factor from DN to solar brightness, etc. So you're not comparing like-for-like when you compare your calculated calibrated total brightness to the uncalibrated img files.

You could try to instruct SECCHI_PREP not to apply the calibrations (via the keyword /CALIBRATE_OFF). This may give you the images that are equivalent to the img files. Although, I'm not sure if you tried just combining the polarised sequence and multiplying by 2/3 without SECCHI_PREP?

Also, another small issue is that you are applying the smoothing mask via the /smask keyword. This smoothing won't have been applied to the raw img files, so that's another difference there.

https://github.com/mmiesch/stereo-pB/blob/aa2efb1baadba78f43f524c33f32d587ec64576b/RT/combine_stereo_pb.pro#L5

mmiesch commented 2 months ago

Thanks @mikemarsh-mo for your comments.

I did start out with noble intentions not to use IDL! I tried just summing up the pB images and multiplying by 2/3 all in python, as captured here. And then I compared them with the unprocessed tB images. But I found some discrepancies, as described in that first email to Bill Thompson that I forwarded. Then he suggested I use secchi_prep. I did find agreement there so I proceeded that way.

However, looking back now, it might be the case that I just did not take into account the exposure times properly in my original python implementation. So, it is worth revisiting.