Dithering with a Constant Temperature Optic

[avandonk]

I now keep the temperature of the 300mm lens constant (20C) with two Kendrick dew heaters and a thermostat. Focus does not change with any change in ambient temperature. In fact it does not even change from one night to the next as long as the fridge and lens are at their set temperatures.

Guidemaster has this nifty ability to move your mount in between exposures while guiding. It is set by specifying a radius in pixels of your guide camera. I now control the cameras shutter with Guidemaster by means of a relay box which also does the guide control. Shoestring has USB adapters that do this.

Here is an animated GIF of nine cropped jpg (8min at f5 and 400ISO)images of the same sensor area straight out of the Canon 5DH. You can see the hot pixels are not moving. This of course means when you median stack a set of images signal to noise is even further enhanced. Resolution is also increased depending on how you process.

Animated GIF 4MB
http://d1355990.i49.quadrahosting.com.au/2009_05/dither01.gif


Some resulting images

Carina 9MB
http://d1355990.i49.quadrahosting.com.au/2009_05/cardith1.jpg

Rho region 5MB
http://d1355990.i49.quadrahosting.com.au/2009_05/rhodith.jpg

Swan and Eagle nebs 6MB
http://d1355990.i49.quadrahosting.com.au/2009_05/swaneagle.jpg


The other nice thing now, is that I can set four different exposure lengths with different numbers of exposures. Very handy for HDR data aquisition. If there is no chance of cloud I can then have a long nap while the data collects.

Bert

[TrevorW]

Nice images bert

[gbeal]

Dithering I understand, but haven't used yet Bert, Maxim allows it as well.
I couldn't work out why you wanted the lens temp constant, until the concept of not needing to refocus appeared., clever.
While I didn't delve into the huge files you linked to, the three that accompany the posting very very nice. One thing I did wonder was the "hatching" criss-cross pattern on the Rho one, what is it?
Regards,
Gary

[avandonk]

Gary the faint reflection nebula intensity is barely above noise hence the fridge. In fact the vertical and horizontal banding of the sensor is what you are seeing. Without dithering it would be far worse. That image needs far more data and a larger dithering radius.

As you can well imagine a 300mm lens made of Magnesium Alloy will contract with temperature. The depth of field of the Canon 300mm lens is tiny about 2mm at 6M at f/2.8. I used to have to adjust the stepping motor that focuses the lens all night. Melbourne can start of at 20C in the evening and by morning it is down to 2C. This of course happens on lovely clear nights. Keeping the lens temperature constant is by far the best solution. I should have thought of it years ago.

The title was an attempt at wordplay.

Bert

[Insane Climber]

Those images are incredible, Thanks for sharing your technique it was quite enlightening especially for a beginner like myself.

Cheers
Jas

[jase]

A lovely panel of images Bert. I particularly enjoy M16,M17 in the same FOV. Its a great pairing, similar to that of NGC6334,NGC6357. The images don't look as though they've been through the HDR process. Maybe they have, and you've mastered the technique as I noted on earlier images that colour balance post HDR didn't seem quite right. These images look the part. Really nice work.

No argument from me on dithering. If an imaging rule book existed, I'd make it compulsory and list it not too far behind a big and capable mount. Dithering is often overlooked. The indirect benefit it brings in managing noise through data rejection is advantageous.

[iceman]

Fantastic work, Bert. I'm sure that all of this work you're doing with your temperature control, HDR, etc is simply ahead of its time.

In the months or years to come, everyone will be following your lead.

Lovely images.

[avandonk]

Thanks for the kind words of encouragement all. As I am retired this is now my full time job. It has many benefits Mike, the main one being it saves me from losing my marbles. A lot of what I am doing is driven by knowledge gained on this and other forums or is that fora.

Yes Jase all those images are HDR's. I have worked out that colour balance etc should be adjusted consistently and carefully for all images used to generate the final tone mapped image. That way the final image only needs a tiny tweak. Like everything it is a matter of practice.

The Carina image is from 10X(2m, 4m and 8m) at 400ISO and f/5 with the Canon F2.8L 300mm lens. Fridge at -5C. So the sensor was at 12C.

I reprocessed the data at 6k pixels wide. The Canon 5DH sensor is 4.3k wide. The sampling vertically and horizontally is about 0.7 that of the diagonal. This is due to effectively square pixel sensors. It has more to do with the intrinsic spacing of the sensor pixels.

This stacking with dithered images should make the sampling more uniform. So here is a large image 7433x5175 pixels in size. If you view it on a typical monitor at 100% the full image is 2.2m wide. I used a bit of RL enhancement.

Large image 15.7MB
http://d1355990.i49.quadrahosting.com.au/2009_05/ecarlarge.jpg

If you look carefully the the hand of the fickle finger is just visible. not bad for a 300mm F.L. optic!


Bert

[jase]

Top stuff Bert. The 300mm lens delivers an amazing FOV.

[Prickly]

Hi Bert,

Amazing detail for sure. Always wondered how you got such incredible resolution with these shots. Thanks for the tips.

I was reading about dithering in nebulosity last night. Can you combine dithering and drizzle or is it one or the other? These are amazing results but wondering how drizzle would compare also?

Cheers
David

[prokyon]

Hi Bert,

I am fascinated by the resolution of these pics!
Especially the Rho region. I made a pic of this region some weeks ago. To compare with yours.... :whistle:

[lacad01]

Really like these wide-field shots, amazing detail and resolution :thumbsup:

[tornado33]

Lovely images. I must try dithering. I can do it manually as I hand guide anyway.
Scott