Help ! Getting crazy with Master Offset...

[gloicq]

Hi there !

I'm using PI for some months and everything worked fine till now. I bought a Moravian CCD G2-8300 and I wanted to work on its first lights. Unfortunately, nothing works...even if I use the AIP process icones.
Let me explain you :

- I shoot my calibration files : Darks, Offsets and Flats with the same conditions than the lights.

- I made my masters with the AIP process icones (so parameters are correct cos lots of people use them without any problem)

- I calibrated the lights with the masters. Surprise : my lights were a lot darker than the originals files. Well, as I never did it with a CCD, I thought "Oh...nevermind, maybe that's normal !" (https://www.dropbox.com/s/xk744l3et431zom/Image%201.jpg?dl=0)

- I proceed with the Stellar Alignment but...

Registering target image 1 of 84
Loading target file:
C:/Users/.../Pictures/M42 07_01_16/C_HA_0096_1.xisf
Loading image: w=3358 h=2536 n=1 Gray Float32
25 FITS keyword(s) extracted.
Structure map: done
Detecting stars: done
11399 stars found.
* Reference image: Limiting to 2000 brightest stars.
* Target image: Limiting to 2000 brightest stars.
Matching stars: done
*** 0 star pair matches found - need at least six matched stars.
* Previous attempt failed - this is try #2
useScaleDifferences=false
* Reference image: Limiting to 2000 brightest stars.
* Target image: Limiting to 2000 brightest stars.
Matching stars: done
*** 0 star pair matches found - need at least six matched stars.
* Previous attempt failed - this is try #3
useScaleDifferences=true
* Reference image: Limiting to 250 brightest stars.
* Target image: Limiting to 250 brightest stars.
Matching stars: done


After a few tries, that made me upset : stars are almost at the same place on the lights files !
So, I thought the problem would come from the dark images after image calibration.
I asked why to different Pixinsight users I know and they told me that wasn't normal at all to have darker calibrated images even after STF...(Background is BLACK)

CALIBRATION WITHOUT MASTER OFFSET

So we tried to find why the calibrated images were so dark : finally, I tried a calibration without the Master Offset and the result was good. So, for me, the problem comes from the master offset

NEW MASTER OFFSET

So convinced that the problem came from the master offset, I've taken 80 new offset and made a new master offset with them. Tried a calibration but always the same results.
I also tried with an offset in place of the Master Offset and the resulting image was clear enough to perform a star alignment...but still not with the Master Offset.

(https://www.dropbox.com/s/mhlf87x9mdmutx0/Capture%20d%27%C3%A9cran%202016-01-11%2012.04.29.png?dl=0)


A NEW DAY, A NEW HOPE...well...kind of...

After working on this problem during hours yesterday, I went to bed thinking that tomorrow will be a better day...fool was I ! Today, I tried a new calibration with the same offset and same parameters than yesterday and...I obtain the SAME dark result than with the Master Offset...what the hell happens to Pixinsight ? I'm getting crazy today !

There is the process icones that I use : https://www.dropbox.com/s/8s9oc8ngecim096/1_JMpretraitement-vp2.xpsm?dl=0

The Master offset : https://www.dropbox.com/s/1eu23ngoc583949/MasterOffset.xisf?dl=0

Offset : https://www.dropbox.com/s/zkh1ajknluea1cp/O__0002.fits?dl=0

A light image : https://www.dropbox.com/s/hskaxxsx9zrfgzj/HA_0018.fits?dl=0



Help me please...

[gloicq]

Need help please...UP

[gloicq]

Up ! Nobody can help me ??

[pfile]

Ha image - not very many stars are there generally speaking.

but from your log: 11399 stars found.

this means that your images are probably not calibrated properly and staralignment has simply found hot pixels.

rob

[gvanhau]

Hello.
If I Understand it correctly:  If you calibrate your images using a master offset (I asume master bias?) you are getting a almost black image, but if you dont use the master bias result is ok?

If so, I think that some how you are substracting master bias twice.
There are several options to calibrate your images, and you have to be consistent with one of them.

For example:

you can do:   
                    integrate bias-> master bias
                    calibrate flats: flat - master bias -> calibrated flat (Im assuming here that flats are very short exposures)
                    integrate calibrated flats -> master flat
                    integrate darks (wo calibration) -> master dark
                    calibrate lights : (light - master dark) / master flat == calibrated ligt.

but you can also:
                    integrate bias-> master bias
                    calibrate flats: flat - master bias -> calibrated flat (Im assuming here that flats are very short exposures)
                    integrate calibrated flats -> master flat
                    calibrate darks : dark - master bias -> clibrated dark
                    integrate  calibrated darks -> master dark
                    calibrate lights : (light - master dark  - master bias) / master flat == calibrated ligt.

If you see, in the first case, I dont subtract master bias from light, because it is already included in the master dark.

So, in order to help you,  it would be necesary that you show the whole calibration process, icluding how you created each of the masters.

Geert

[bmhjr]

I am having the same issue since upgrading Pixinsight.  When I include the Master Bias, which is the same one I have been using in past, then the Light Calibration does not work.  The calibrated lights look very dark with a few star/pixels If I dont include the bias, it looks as expected.

I follow this manual work flow
                    integrate bias-> master bias
                    calibrate flats: flat - master bias -> calibrated flat (Im assuming here that flats are very short exposures)
                    integrate calibrated flats -> master flat
                    calibrate darks : dark - master bias -> clibrated dark
                    integrate  calibrated darks -> master dark
                    calibrate lights : (light - master dark  - master bias) / master flat == calibrated light

The image with bias is on left, the image on right is before calibration.  STF applied to both

Something is crazy!

Thanks
Bill

[bmhjr]

I am having the same issue since upgrading Pixinsight.  When I include the Master Bias, which is the same one I have been using in past, then the Light Calibration does not work.  The calibrated lights look very dark with a few star/pixels If I dont include the bias, it looks as expected.

I follow this manual work flow
                    integrate bias-> master bias
                    calibrate flats: flat - master bias -> calibrated flat (Im assuming here that flats are very short exposures)
                    integrate calibrated flats -> master flat
                    calibrate darks : dark - master bias -> clibrated dark
                    integrate  calibrated darks -> master dark
                    calibrate lights : (light - master dark  - master bias) / master flat == calibrated light

The image with bias is on left, the image on right is before calibration.  STF applied to both

Something is crazy!

Thanks
Bill

Ok, I found my problem.  My approved lights were RGB instead of gray scale.  After the updates, the DSLR format settings were defaulted.  After I changed it back to Pure_RAW and reran the subframe selector output frames, the calibrated light looks fine with Master Bias subtracted.  DOH!

Thanks
Bill