DSLR and planetary imaging

[jambouk]

A previous thread (http://stargazerslounge.com/topic/201073-televue-powermate-on-a-11-edgehd/) has made me think about the use of DSLRs in planetary imaging.

In essence, why not use a modern DSLR for planetary imaging?

For example, the Canon 6D has pixels 6.55 microns, which corresponds to a f/ratio of about f/30 for optimal planetary imaging, which would be about right on a 2x powermate on the back of a f/12-f/15 scope. In video mode, the camera can record at 25 fps in low compression (ALL-I) mode at 1920x1080.

So, my questions are:

1. How could you alter the "gain" to make sure the image wasn't over exposed in the video?

2. What happens with regards "binning" when selecting the 1280x720 50 fps low compression mode?

3. Would there be any scope in just doing rapid fire non-video exposures at say 1/60th second at 5 fps in large jpeg mode for say 3 minutes; only 900 frames, but might the resolution in the good frames outweigh the poorer resolution of the 3000 frames shot on a webcam?

It would be nice to have a load of clear, good seeing nights to play with this.

James

[jambouk]

I think i've just worked out, even in Video mode, I can alter the ISO and the shutter speed, so that would effectively allow me to alter the 'gain', so question 1. is already answered.

James

[cgarry]

Why do you think a 'webcam' has poorer resolution?

At the end of the day you want NO compression, so for me low compression video and jpeg are both out.

Cheers,

Chris

[jambouk]

I'm not sure. This is what I really want to get to the bottom of, and essentially have a go.

When the manual talks of compression, what is this relating to? Binning in some form? If so, how much? What are the problems with compression once any stacking software has broken the video up into single frames?

James

[Space Cowboy]

There are some half decent planetary shots taken with modern DSLR cams but they are not in the same league as webcam/planetary cams i.e DBK, ASI or QHY.

[cgarry]

Hi James,

Compression for video and images is either lossless or lossy.  Lossless compression means that the original data is available bit for bit, but has the downside that it does not compress the file size as much as lossy compression.

Raw images from a DSLR use lossless compression (good!), jpeg images and video from a DSLR use lossy compression (bad!).

Lossy compression produces very small files, but has the disadvantage that the data is 'approximated' in order to compress it.  This means that when you zoom into the images there are obvious artefacts and the original fine detail has been lost or distorted.  With lossy compression you trade off between quality and file size.  By choosing a higher quality value you will get less image artefacts at the expense of a larger file size, but you will always get some artefacts.

Lossy compression is great for looking at videos and images with the human eyes as that is what it was designed for, but it is a different matter for stacking software where that lost detail can never be recovered.

Having said that, as Stuart says there are some nice planetary shots posted on this forum from guys breaking the 'rules' and imaging using DSLRs in video mode.  Though in my opinion they cannot touch the quality of images produces by dedicated fast frame rate cameras (avoiding the term webcams because most of them are not webcams).

Cheers,

Chris

[jambouk]

That's useful, I think I'm getting there....

So this compression, what is actually being compressed?

If on the DSLR sensor, during video mode, one pixel detects 10 photons, and its neighbour detects 4 photons, and the next one detects none in a single frame, what does compression do to this information?

James

[michael.h.f.wilkinson]

One simple problem is that at the 5000mm focal length I use for imaging planets, I only need 640x480 pixels even on Jupiter in my scope (0.3Mpixel). Uncompressed, or using lossless compression like LZW, I can transmit this data at huge rate through USB 2.0. Your 6D sensor is a 20.2 Mpixel sensor, but only an area of 640x480 is needed. You are collecting 67x more data than needed. This means that you need a far faster link and disk to keep up with the data rate or you need far more compression. Besides, the AD-converter has to be 67x faster, which adds noise too.

For the moon the situation is different, because a significant area of the sensor will be filled. You can afford to take far fewer shots than with a web-cam, and you need less stitching to make a large mosaic. Simply taking a sequence of shots works well.

[jambouk]

Thank you all.

James

[Russe]

Great info! Thanks!!