Page translated by Claude — switch to Italian to read the original article.
This tutorial, created by Edoardo Luca Radice and Vittorio Liberti of the Astrofili Veneti (http://www.astrofiliveneti.it/), aims to analyze and solve one of the possible problems that could be encountered when using PixInsight’s BatchPreprocessing with the Default values.
Using BatchPreprocessing with the default values, the final result was the following (as an example, only the image obtained with the SII filter is used).
It was thought that the brighter part depended either on gradients (unlikely) or, more probably, on an overcorrection of the vignetting due to an incorrect flat field, but this hypothesis too was soon set aside since the setup used should not give rise to vignetting.
The only viable route, which later proved to be the correct one, was an incorrect calibration of the images. After various tests we came to the conclusion that the problem lay in an incorrect calibration of the flats with the darks (below you can see the master dark used and, as can be noted, the bright areas seen in the light above closely trace the areas where the amp glow is present in the master dark). For the record, the sensor used is a CMOS.
In practice, having deselected the DARK optimization in BatchPreprocessing, the FLATs were being calibrated with an excessive DARK, which created dark areas corresponding to the amp glow, simulating the presence of vignetting.
After calibration, therefore, these areas turned out to be much brighter.
The last test performed gave the hoped-for result and consisted in obtaining a master flat for each filter, calibrated and integrated separately and not through batchpreprocessing.
In practice, the checklist to get to the Batchpreprocessing is this:
1. Have the Master Bias with the same sensor temperature and gain as the Lights;
2. Have the Master Dark with the same temperature, gain and duration as the Lights (it is better not to use dark optimization when amp glow is present; this requires that the lights have their respective darks of equal duration!);
3. Have the master Flat obtained by calibrating the flats of each filter with the Flat Darks (so for each filter the optimal thing would be to take the flat darks of equal duration, temperature and gain). If the FLAT exposure times are short and the sensor is cooled, for simplicity it is possible to calibrate the flats also ONLY with the Bias (as done in this tutorial!);
4. Once you have all the masters, then you can put everything into BatchPreprocessing (thus checking the options for using the Masters and unchecking the Dark optimization, see fig. 5).
So, to be precise, the ideal would be first of all to take the Flats for each filter, take the respective flat darks, obtain the Master dark and calibrate the Flats with the respective master dark and the master Bias. Integrate the calibrated Flats to thus obtain the Master Flat (for each filter).
Let us now move on to how to obtain the calibrated Flats (I remind you that in this tutorial I calibrated the flats with the master Bias since the flat darks were not taken).
1. Calibration of the Flat frames.
If we have the Master Dark of the Flats then we also check the Master Dark item and select the respective Master Dark. In this example only the Master Bias was used.
- Integration of the calibrated Flat frames.
And at the end we obtain the Master Flat (an operation to be repeated for each filter used).
We can then move on to BatchPreprocessing.
- BatchPreprocessing.
Final result
The images above are obtained by feeding everything into the “BatchPreprocessing” script and letting it generate the master flats itself.
The images below are obtained instead by first generating the master flats and then running BatchPreprocessing.
WARNING: You must be careful, however, that the master flats obtained contain in the Fit Header the indication of the filter used, otherwise an error could be generated during the process because the system is unable to match the master flats to their corresponding lights.
Since we are at it… let us see in detail how to change the Fit Header with PixInsight.
1. let us open our image and in the FILE menu select FITS Header
- Select the image to be checked using the drop-down menu at the top of the window: as can be seen in the image, the filter used is already indicated in the Light frame… but let us pretend it is not there and that we want to insert it.
As can be noted, there are three columns: Name, Value, Comment.
It will be enough to add, in the respective rows found in the window, the values we want to be displayed (in my case I added: FILTER, OIII, Filter used).
And click on Add
As can be noted, the values entered in those fields are now displayed together with those already present.
However, you must apply the work just done to the image. So let us click on the now-famous small blue triangle present in the bottom-left window…
…and let us drag it onto the image. Only at this point will the image’s Fit Header contain the desired annotation. To test it you can close and reopen the Fit Header and check. To delete the annotation, just select it and click on Remove… and then apply the change to the image.
If this concerns many images, the advice is to create an Image Container and then apply the Fit Header to it, so that it is applied to all the images at once.