Clear all

Please note our new Downloads page here

2023-01-19: APP 2.0.0-beta13 has been released !

!!! Big performance increase due to optimizations in integration !!!

and upgraded development platform to GraalVM 22.3 based on openJDK19


We are very close now to  releasing APP 2.0.0 stable with a complete printable manual...


Astro Pixel Processor Windows 64-bit

Astro Pixel Processor macOS Intel 64-bit

Astro Pixel Processor macOS Apple M Silicon 64-bit

Astro Pixel Processor Linux DEB 64-bit

Astro Pixel Processor Linux RPM 64-bit

Data calibration and post processing of Canon EOS Rebel T4I (650D) data

Quasar Admin
Joined: 6 years ago
Posts: 3426
Topic starter  

Hi all,

This is a workflow on a small subset of data of Kevin Malone


OTA: C8 and a focal reducer (which Kevin is experimenting with to find to right reducer distance).

Camera a Canon EOS Rebel T4I (650D)

Light frames : 6 x 300 seconds iso 400

flats frames: 5 x 0,2 seconds iso 400

flat dark frames 5x 0,2 seconds iso 400

dark frames: 6x 300 seconds iso 400

APP Kevin1

 First step:

 I have created a Bad Pixel Map of the darks and flats following:


 The Bad Pixel Map only needs to be created once per year (or several years, depending on you camera's sensor).

app Kevin2

 Second step:

creation of the master calirbation frames, Kevin supplied bias frames as well, but since we have darks en flat darks, I don't use the bias frames because the bias signal and BIAS pedestal is already contained in the darks and flat darks.

So I loaded all darks, flat darks en flats into APP and clicked on "calibrate" in 2) CALIBRATE.

This created 2 master darks (1 for the light frames and 1 for the flat frames) and 1 master flat.

Let check if the calirbation works using the l-calibrated image viewer mode :

app Kevin3

Obviously something is wrong in the flat calibration. I have experimented with using the bias frames instead of the flat darks for the flat frame calibration but it gives the same result. These flats will not work I am afraid. We can test this in APP with the "correct vignetting" tool in 9) TOOLS.

Using a Kang Weiss model with geometric factor (third model is the model list) we can correct the vignetting, this indicated that the supplied flat frames are bad. Unfortunately the "correct vignetting" tool wont'correct the dustspots that are on the sensor, so this is not a solution now.

app Kevin4

I suspect the flat frames have a problem with corrector of the C8, this doesn't seem to be corrected by the flats. Maybe it's a good idea to create the flats next time with a longer exposure time and ISO100 possibly to get better quality flat frames.

 Third step:

I will now integrate the calibrated light frames, I loaded the lights, the Bad Pixel Map, the Master Dark for the lights and the MasterFlat (created from dark subtracted flats).

app Kevin5

notice the marks in the Frame column:  D F BPM

This indicates that the lights will be calibrated with the supllied master dark, master flat and the Bad Pixel Map. You don't need to save the calibrated frames. The frames will be automatically calibrated in the steps up until integration. Debayering of the data is also done automatically at the right time.

3) ANALYSE STARS used defaults

4) REGISTER  turned on dynamic distortion correction

5) NORMALIZE defaults


  • composition mode full,
  • MBB 5%,
  • LNC 1st degree 1 iteration,
  • no outlier rejection since I have only 6 frames
  • pixel interpolation, Lanczos 3 with no under/over shoot turned on


And I pushed "integrate" in 6) INTEGRATE

St avg 1802.0s NR x 1.0 LZ3 NS full eq add sc BWMV nor AA RL MBB5 1stLNC it1 St

Notice the bend borders. This is dynamic distortion correction at work. This corrects the optical distortion of the C8 which is quite considerably. Furthermore, the bad flat-field calirbation is shown and we also notice a clear gradient.

Using the "remove light pollution" I have tried to correct the image. Correcting the bad flat-field calirbation is hard, but by setting the area slect boxes near the borders of the clear illumination difference circle, we are able to correct this to a high degree:

Kevin LPcor

Notice how I have placed the area select boxed in circles aroung the flat field calibration artefact surrounding M101.

From looking at the stars in the corners, I would think that the corrector is not well placed yet. M101 in the center of the Field Of VIew is sharp, but the stars in the corner are not. We see very strong aberration in the star shapes. So Kevin has some more work in getting the reducer to work properly.

app Kevin6

 Fourth step:

After the remove light pollution action, I have used the calibrate background tool, the star color calibration tool and the selective color tool.

After star color calibration, you will usually see a green cast over the image. You can correct this by injecting the opposite of magenta in the green channel in the selective color tool.

Before selective color:

St avg 1802.0s NR x 1.0 LZ3 NS full eq add sc BWMV nor AA RL MBB5 1stLNC it1 lpc lpc lpc cbg csc St

After selective color, green channel + 0,85 magenta (twice, click on apply to apply the settings and do another iteration).

M101 KevinMalone St

Finally, to get completely rid of the Illumination circle due to the bad flat-field correction, the light pollution removal tool could be used several times 😉






White Dwarf Customer
Joined: 6 years ago
Posts: 16

The failed flats calibration/illumination circle ended up being via light being phased back on through to sensor after first reflection off clipin filter then focal reducer. It seems it's one or the other with this system, either the filter or the reducer but not both. What was learned off this set of data being trialed through APP was that APP's stacking is far superior to both DSS Adaptive or Sigma. Although I didn't do final processing and finish work in APP the stacking really allowed for some good software bin/deconvolution work in another processor I'm more familiar with..Thanks Mabula for your hard work!! I submit for your review a screen grab highlighting APP's stacking over DSS. APP version is the middle version with DSS Adaptive on left and DSS Sigma on right. The resolution gains really impress!

~Kevin (KTMan) Malone

Quasar Admin
Joined: 6 years ago
Posts: 3426
Topic starter  

That's good to hear 😉

I am happy you're getting to the bottom of your illumination circle problem. Also happy to hear that you like APP's integration qualities!

Can you post the screengabs again? Possibly something went wrong? I am eager to see the differences 😉



White Dwarf Customer
Joined: 6 years ago
Posts: 16

DSS Adaptive on left, DSS Sigma on right, APP in middle. It was real friendly to getting the most resolution..

DSS Adaptive APP DSS Sigma stacking ST processed

Quasar Admin
Joined: 6 years ago
Posts: 3426
Topic starter  

Great Kevin, the difference is stunning 😉 !