3.01
...
·
 |
|---|
Attached is a screenshot showing three stacks of M106 data from three nights. my setup is in an observatory so the imaging train wasn't changed between sessions and I was therefore able to use one set of flats for all three nights. However, the flats have been applied incorrectly for the first night's data. As you can see, the dust spot in at the top of the image looks like it's convex and the one at the bottom is concave. It's as though the master flat has been applied upside down and is under/over correcting. Has anyone ever experienced this or can anyone suggest a fix? TIA!  |
9.89
...
·
|
|---|
There is shift between flat and lights for the 1st image. As it is probably empty sky where the embossed residual shadows sit the best course of action is likely to be to clone-stamp them out of existence. In case you are a purist you need to shift and/or rotate the flat until the embossed shadow disappears. And, yes, it happened to me in the past. |
|
3.01
...
·
|
|---|
Looking at the position of the two error marks, it looks like the master flat needs rotating by 180 degrees. The problem is, it’s correct for the other two stacks. The final image will be a stack of all data together (which I did first but then stacked the three nights separately to troubleshoot). If I rotate the flats, which I’m not sure how to do anyway, They will be incorrect for the other two night’s data. I’m not sure how this could have happened as the rig didn’t change between sessions.
|
|
9.89
...
·
|
|---|
It's not a 180 degree rotation, just a small rotation which can be linearized as a shift. Doing it is easy enough and yes, you'd need to treat the 3 nights separately and then average them together. Which I would do anyway.
|
|
3.01
...
·
|
|---|
How would I go about rotating the flats please? Does this need to be done before WBPP?
|
|
9.89
...
·
|
|---|
No WBPP please. If you have a mono camera then the following is what you need to use:  This will rotate the flat until you nail the right rotation. If you have an OSC camera the thing is a bit more complicated . First you need to calibrate your lights without a flat (so just darks and cosmetic correction, if you do it that way), then you debayer them. You also debayer your master flat. Now back to ImageCalibration, you just calibrate using the debayered rotated flat with the debayered lights until you see no longer the embossed shadow. Then proceed with the remainder of the process such as registering, normalization and subframe selection till you integrate that night's lights. |
|
3.01
...
·
|
|---|
Thanks for the info. I’ll have a look into this but it sounds like it might be easier to clone the artefact out. As it’s only the background, it shouldn’t affect the final image. I’ve got about 12 hours of data so far and am aiming for 20 so if I add more subs, the effect should also be reduced
|
3.61
...
·
|
|---|
Mind if I ask a side topic question?…. Still learning here. I’ve taken a set of flats with a flat panel. If I rotate my camera to better compose a shot, the previous flats are no longer useful right? So…I have to save flats from that imaging session if I want to use the data later??? Or do i just process that night and save the master lights for later? |
|
3.01
...
·
|
|---|
Just a quick update on this. I’ve managed to solve the problem. I think I used the wrong flats (rookie error) as i was imaging two targets that night
|
|
3.01
...
·
|
|---|
Sean Mc: That’s right. If you rotate the camera, you need new flats |
|
3.61
...
·
|
|---|
Well… that adds a new layer of complexity. I’m far too lazy for that.
|
|
3.01
...
·
|
|---|
The best thing to do is just take new flats for every session. It only takes a couple of mins if you are using a flat panel. I usually do mine the morning after.
|
