Crux Enhanced with Focus Magic

[avandonk]

This is an image of Crux from data taken with a Canon 300mm F2.8L and a Canon 5DH. It is actually at nearly twice (X1.9) the native sensor pixel size. That is 47 Mega Pixels versus 12.5 Mega Pixels. It has to be this large to show the detail.

Large image 8371 x 5553 pixels and 18.5MB.

http://d1355990.i49.quadrahosting.com.au/2010_08/crux_FM.jpg

If you click anywhere on this image in your browser to show the image at 100%. The full image size is then 2.5 metres wide!

This is an animated Gif showing the Focus Magic enhancement. 1.7MB

http://d1355990.i49.quadrahosting.com.au/2010_08/fmc1.gif

Is it real? Here is an animated Gif of the enhanced image with the same area of sky from the Deep Sky Survey. (DSS_1) 1.7 MB

http://d1355990.i49.quadrahosting.com.au/2010_08/dssfm1.gif


I have managed to get far better resolution from my sensor by dithering while collecting data and stacking upsized.

Focus Magic seems to improve the image further without the artefacts that star reduction or Richardson Lucy enhancement produces.

Here is the above large image downsized to the native pixel resolution of the sensor at the same jpg quality 4532 x 2887 pixels 6MB.

http://d1355990.i49.quadrahosting.com.au/2010_08/crux_FM_NAT.jpg


This is an animated gif showing what the same area looks like at the native pixel resolution of the Canon 5DH sensor. 1MB.

http://d1355990.i49.quadrahosting.com.au/2010_08/crNat.gif


With these methods I have managed to make the 12.5 Mega Pixel sensor of the Canon 5DH perform like a 47 Mega Pixel sensor. This can only happen if the lens or telescope has far better resolution than your sensor or in other words the image is under sampled.

Bert

[renormalised]

Amazing detail in there, Bert. Great shot:)

[Shiraz]

hi Bert
Thanks for posting such a clear demonstration of the benefits of super-resolution reconstruction through dithering. Its a great way to increase the flexibility of a system isn't it? - if the seeing is poor, process normally, if the seeing is good, use super-res to increase the image scale (by 2 say) and increase the number of subs (eg by 4x). All without changing the system at all - its sort of like having a "resolution" switch on the side of the camera. I guess you could get the same flexibility by binning images from a larger camera in poor seeing, but then you would have to buy the larger camera in the first place. The only problem I have found is that the tracking mount need to be pretty good - super-res very effectively enhances trailed images and I have ended up with long, thin stars instead of eggs at times. BTW, I use Nebulosity, which automates the process, linking to PHD to do random dither between subs. Sorry about the long-winded post, but this is a fascinating subject.

[multiweb]

Very interesting read Bert. Thanks for taking the time to post all your results. :thumbsup:

[tornado33]

Well done, I must try dithering on my next imaging run.
Scott

[luigi]

Very very nice read and images, I'm intrigued by this technique, but I didn't fully understand how it works, how do you dither your image?

The crux area is really superb one of the most detailed photos of Crux I've seen well done!

[Paul Haese]

So many stars and the resolution is very nice.