Results of nights imaging

[rcheshire]

Well. Not so good - lots of colour, but that's as far as it goes. Posting to get some feedback - wouldn't bother otherwise. Pretty crappy result.

Problems with dark calibration, but that is fixable. Bias and flats applied only. Would anyone care to comment on the streaking? Thanks for looking.

[DavidU]

Rowland, show us the image first:D

[adman]

what camera are you using?

[rcheshire]

There now - it's getting late.:P

Canon 1000D - not cooled.

[Stu Ward]

Hey
I'm no expert, looks a little noisy, but apart from that I think it's beautiful !!
Really nice pic.

Stu

[irwjager]

Hey Rowland,

Been a while - hope you're well :)
The streaking is somewhat hard to see with all the JPEG artifacts, could you maybe post a 1:1 crop or host the full image somewhere?

Since you have ST, download the latest 1.1 Beta and try the Chroma Noise Filter in the Develop module. For people with Photoshop, the technique discussed here (http://www.digitalphotopro.com/technique/revolution/blur-color-noise-in-photoshop.html) will do the same.

It's hard to judge the streaks from the attached JPEG, but if they're horizontal or vertical (most likely due to read noise), try using the (experimental) Band module in ST 1.1 beta. For people with Photoshop, Noel Carboni's Astronomy Tools Action Set does something similar (but it's not adaptive). For best results try using it while the data is still linear (unstretched).

Cheers,

[renormalised]

How many subs taken and with what camera???. How many subs bias, flat and dark frames?? ISO rating, F ratio etc etc. Was it shot piggyback or prime focus???. We need to know those to be able to help. You have a lot of JPEG compression noise there, however that streaking could also be just noise from the pic...not having enough frames to pickup your S/N ratio. It may also be read noise, like Ivo mentioned. You can get plugins which will help remove this.

[rcheshire]

Thanks guys.

Thanks Carl. As there is no obvious answer here are the details.

New image - over saturated and cropped the centre section to illustrate the problem, which is image wide, depending on star density near the corners.

Camera 1000D - noise reduction both off - not cooled.
Lens 200mm prime f/3.2 - CLS-CCD filter.
ISO 800
Ambient temperature ~10 C.
Polar scope alignment - 15 minutes no appreciable drift.

38 frames: 1x 15, 2x 10, 2x 7, 31x 5, 2x 3 minutes - 3.5 hours

I use PixInsight for the heavy lifting. So there are technical differences in relation to calibration of lights darks and flats - please take it up with the PI developers. I wont take up space explaining it here. I suggest reading their explanation rather than debating it here.

Problem is the same with and without calibration.

40 bias frames - 10 darks and 10 flats.

So what is causing the streaking?



Ivo. I've been good just busy how are things with you? I'll be sure to download the latest ST.

[irwjager]

Did you dither while taking your image?
The streaks seem reminiscent of non-dithered data combined with very good polar alignment & a wide field. Does your very last frame align perfectly with your very first frame? If there's almost no difference you may want to try dithering next time you image.

Still, I like the result (very nice colors!) and it should be fixable with some processing.



Good to hear. Busy myself preparing for AAIC 2011 and the usual daily freelancer craziness :)

[rcheshire]

Thanks Ivo. Hmmm... I think I can identify a problem. It seems that unless high speed is selected on the hand paddle there is no real movement in RA or DEC. It would have been easier to interrupt the drive between shots for 5 - 10 seconds or so and reframe occasionally.

Great to have the time and opportunity - work....

[rcheshire]

Folks thanks for your input. The problem is twofold. Poor calibration - failure to remove hot pixels causes the effect to carry over from one image to another during registration. If the images are aligned by hand resolution is at pixel level, and as I understand it during registration the streaking effect is occurring at sub pixel levels.

If you are a PI user, which uses temperature independent scaling during calibration, there is a temptation to take only a few dark frames. Temperature is not such an issue but a good master frame is required all the same.

Hope this is useful to someone.