After some discussion on the QHYCCD yahoo forum, turns out that this is because of the steep light paths the the f/2 hyperstar system entails, combined with the large CCD chip used to image. The data is there as can be seen from the red channel:
The other channels show severe banding. Green:
And Blue:
As can be seen, the flat looks ok but is just not able to correct for these diffraction bands:
And the bands make it very hard to correct for it and bring out the details. People indicated that the Baader 35nm Halpha filter migt do a better job. The reason being that the steep light path basically shifts the wavelength by about 6nm and this leads to diffraction patters resulting in the color bands. The Baader 35nm, being pretty wide for a Halpha filter still captures the Halpha even though the shift happens and hence should preserve the data while cutting out all the light pollution. I tried stacking them without flats as well, to rule out the flats and the same problem exists. But oh well!! After 200$ for the Astronomik UHC, I think I am back to using the IDAS-LPS-P2
One thing that can be noticed here is that the chip seems to be offcenter. This was indicated by Sander after seeing my flats. I think the collimation effort pushed it a bit off-center. I can try and recollimate to bring it back to the center, but for now I think I am going to continue with this setup and the LPS to see what kind of results I get.
Posts
2 comments:
This bands may be corrected. Like many other risidual optical gradients (caused by spectral peculiarity of coating/interference filters of optic elements located near to focal plane). The root of this problem - is difference of night sky background spectrum and flat-box(or daytime sky) spectrum.
You should spend a (moonless!) night taking night sky-flats. Many subframes of starfields with shifting between frames. Substract dark, normalize sequence and add frames with strong Sigma-Clipping (3 iteration, sigma 1.5-1.7).
After that you will get sky-FLAT for night-sky spectrum.
Then take usual FLAT (with flatbox or by daytime sky).
Divide that SKY_FLAT by FLAT, and you'll get "correction-file". Debayer it. Blur it with median filter(~3-5pixels) to elimitace residual hot pixels, then blur it about 20-50 to reduce overal noise level. Than save it and use any "lifetime" (for moonless nights only, and until big re-collimation of the optics.).
Using is simple - divide images by it like usual FLAT, after full (D-F-B) Calibration and CFA-conversion.
Igor Chekalin.
Thanks Igor! I will try this. And sorry for the delayed response. Was out of action for a bit
Post a Comment