M57 the ring nebula

[NickK]

Ok, I still need to sort out some noise in the corner (I think this is down to not setting the cooling low enough):

M57, the ring nebula
This is scaled from 3376×2543 16bit (17MB) to 800xY 16bit:

Full resolution crop:

 

[juno_first]

How come it has to be cooled?

 

[NickK]

Noise. The colder the CCD, the less noise occurs. Normally it operates as 40degC below ambient, so about -25degC at the moment. At night you're picking up few photons from faint objects. The less noise means you'll see more signal (more picture).

To make the image above:
* 10 x 60 second 'light' exposures - actual images.
* 10 x 60 second 'dark' exposures - images with the lens cap on to find the noise introduced by the camera over a 60 second exposure.
* 10 x 1 second 'flat' exposures - images with a electro-luminecent panel on the end of the scope to discover the flatness of the field and to detect any dust motes.

I'm not an expert at the moment but by processing for that image was:
1. stack darks to give a strong image of the bad (hot/stuck) pixels.
2. Make a bad pixel map from the stacked darks
3. apply the bad pixel map to each of the flat frames to correct bad pixels
4. stack the corrected flat frames
5. apply the bad pixel map to each of the light frames to correct the bad pixels
6. apply the flats to the light frames
7. stack the corrected light frames together using stars to align
8. Adjust the image using Digital Development (DDP) algorithm.

I think I need to apply the darks at some point..

 

[juno_first]

Ouch, thats a whole lotta work to remove image noise.
Very interesting stuff.

 

[PhillyDee]

Quite a nice focus on that! What do you use for guiding? Just a mount, or have you gone with a guidescope?

Whats the glow in the bottom right off?

Nice image BTW!

 

[NickK]

The camera pixels are 5.4um across and with the scope, this works out that each pixel is 1.66 arc seconds across (an arc second is the width of a ten pence piece at 3 miles away). The scope has a 675mm focal length at f6.3. The optics are good for at least 1 arc second but the mount pointing accuracy without periodic error correction is a good deal lower (in the tens of arc seconds).

I use an application that measures the sharpness (flux density) to focus the scope. I still do it manually rather than use a focusing motor.

At full resolution the image showed a small elongation of the stars unfortunately. The lack of a good alignment results in blurring of the target, so the ring could be a lot sharper (5minute exposures had star trails). I used an off-axis guider (my old 16ic camera and a program called PHD) but as it was the first time I've used it - I was unsure of the results and I don't think PHD was working correctly.

I think the glow is amp noise or the natural noise gradient of the CCD. Having the cooling on should remove that (as should apply the darks).