Conlin Hill Observatory -- 42 7' N, 71 54' W

Techniques - Imaging problem: streaks in images


September 21-24, 2003 Lately I've noticed streaking of my fainter images particularly those made with the HA filter or when objects present weak signal using one of the RGB filters. That said, the streaks are apparent even on fairly strong LUM data. Below, LUM and RED image on the left are dark subtracted, aligned and combined in CCDSOFT. The images on right are aligned and combined with no dark subtraction.

Clearly the streaks are image noise spread along the diagonal by the alignment of the images. As a side issue I need to reduce the amount of shift seen here. I may have been out of balance for these shots. The images were guided with my 201xt autoguider.

It is also clear that the subtraction of dark frames before combining helped to reduce, but did not eliminate, the effect. The streaking in the weaker Red data makes the final reduced, aligned and combined image unusable.

An experienced imager I spoke with guessed that the problem was with the temperature of the dark frames. He suggested that the temps of the darks and the images didn't match or that the camera was having trouble holding a steady temp due to an overly agressive temperature set point. To address this issue, I set the temperature point to that recommended by CCDSOFT (-5c in this case), and took a new set of darks immediately following the imaging session. I checked the FITS headers to determine the temperature at which each image was taken. In all but 4 images, the LUM, Red, and dark frames were taken at -4.8c. The 4 discrepancies were at 5.2c, with 2 dark and 2 red frames exhibiting the difference. This strikes me as being pretty close to identical temperatures, and yet the streaks remain.


I received more on-line advice telling me to try using Sigma combine software. The author of one of these utilties, Ray Gralak, contacted me and got me going. His software can be found at Sigma Home Page. It worked wonders! Just using it to recreate my dark frames produced much smoother images. I'm grateful to Ray for his help and his Sigma software! I found I still likes CCDSOFT to combine my bright LUM images, but SIGMA for the fainter and noisier color shots. I followed his on-line instructions for adjusting settings for both the new combined dark frame and the color light frames, they worked well. Here are the re-done LUM and Red images. Both reduced with the new dark, LUM combined with CCDSoft and the red combined with SIGMA:

Major improvement! It was great to complete my first color galaxy image, and in fact, my first galaxy since May 4, 2003, a full 4 months! Here's the final image: gallery-galaxies-ngc6946-001.htm

<< back

Home   Biography   Equipment   Techniques   Observatory   Gallery   Wish List   Favorites