So confused about cameras

[Datalord]

To my regret I have read too much about cameras for my own good. And not sure I actually know much more. 

First off, my 6D. I love it. Will I be able to use this clip filter to take proper narrow band images? 

http://www.astronomik.com/en/clip-filter/clip-filter-fur-canon-vollformat/clip-filter-eos-xl-mit-astronomik-h-alpha-ccd-6nm-a.html

 

Second, I've read about all the cameras and one thing I can't get through my head is how you use a 4 MP camera for anything of a decent size? Is it because you drizzle and thus expand the size 2x or 3x? Does anyone have experience with this? Coming from 20MP it feels super wrong to go with 4.

Third, I also read that a DSLR at iso 1600 has a much better sensitivity that any CCD, so a CCD would need a lot more exposure (2x-8x) than what my 6D needs. Also seems bonkers to me. Does it have any truth to it? 

[Davey-T]

The basic difference is your DSLR has a Bayer matrix covering the actual pixels, a CCD has separate pixels that can be R,G,B,L,Ha, O111, S11 so even if there are less off them they're a lot more versatile.

A narrow band filter in a DSLR will reduce its data capturing capabilities even more.

Lot's of great DSLR astro images around so no reason not to use one.

Dave

[steppenwolf]

A DSLR camera like yours has plenty of pixels so it produces high resolution images but that is only part of the equation. To produce a colour images, as Dave has said, a DSLR camera has a matrix of filters (red, green and blue) sublimated on the sensor's surface which means that 50% of the receptors collect green light, 25% collect red light and the last 25% collect blue light. This works fine with daytime images and with two caveats, fine with night sky imaging. The two caveats are the IR blocking filter that all DSLR cameras have in front of their sensors have the unwanted side effect of filtering out some of the Hydrogen Alpha (Ha) light so necessary for imaging nebulae and the fact that green is the least important colour in astronomical imaging yet 50% of the sensors collect it! Also, only the red receptors would be able to collect the all important Ha component.

You can in fact use an Ha filter with a DSLR camera but it will not work very efficiently for the reasons stated above.

With regard to resolution and, therefore, sensor size and receptor size, larger receptors are an advantage in low light conditions, i.e. those conditions experienced in deep sky imaging. Likewise, many deep sky objects are large so a large sensor producing a wide field of view is an advantage. However, it is more important to achieve a good sampling rate (measured in arcseconds per pixel) with your specific telescope than to just consider resolution for resolution's sake.

The world of astrophotography is very different from that of 'terrestrial' photography as the requirements for light collection are so very different. In an ideal world where money grew on trees, a large sensor mono CCD camera with big pixels and a set of large filters covering Luminance, Red, Green, Blue, Ha, OIII, SII and NII would be the weapon of choice!

[vlaiv]

I would like to add the following in response to ISO setting on DSLR and sensitivity.

No real sensitivity is gained by cranking up ISO - it is just a multiplier (there are some benefits of having variable gain when your ADC resolution is not enough to cover full well capacity, and effects on read noise with some cameras, but that aside) - so if your DSLR collects 10 photons and on high ISO setting you get 1000 units - there is no additional information there, because 5 photons will produce value of 500 so signal of 10 photons will still be twice as much as that of 5 photons (1000/500 = 2). Photons are discrete in nature, and anything that amplifies photon count by multiplication adds no value to recorded signal. At least this is so in astro photography where captured image undergoes a lot of post processing in order to overcome huge magnitude of differences in captured signal - in daytime photography, ISO makes sense since there is enough light and you most of the time, don't need to alter image in order for it to show captured scene (so being visually brighter or less bright does play part, and noise levels are really low compared to signal - not so in astro photography).

[pete_l]

3 hours ago, Datalord said:

Second, I've read about all the cameras and one thing I can't get through my head is how you use a 4 MP camera for anything of a decent size? Is it because you drizzle and thus expand the size 2x or 3x? Does anyone have experience with this? Coming from 20MP it feels super wrong to go with 4.

A 4MPix camera is more than adequate. A full-HD monitor is "only" 2MPix and nobody complains about the image quality on one of them. Even if it is a 27-inch monitor and you're sitting within arm's reach of it on your desk.

One could argue that since a 4K monitor is "only" 10MPix, that having anything of higher resolution is wasted, since you would never be able to view that natively unless you have cropped the image down.

 

Oh yes, one thing I nearly forgot. Most DSLRs have an anti-aliasing filter in front of the sensor. So while a DSLR has a nominal "resolution" of the sensor that always forms the central theme of the sales literature, you never get that resolution unless you remove the filter, or buy a camera with it specifically omitted - like the Nikon D800e.

The reason for the filter is to eliminate fringes that occur occasionally in domestic photos. But they also kill a noticeable amount of sharpness. Astro cams don't have that filter. So their "true" resolution is more comparable with a higher quoted DSLR resolution camera.

[Datalord]

12 minutes ago, pete_l said:

A 4MPix camera is more than adequate. A full-HD monitor is "only" 2MPix and nobody complains about the image quality on one of them. Even if it is a 27-inch monitor and you're sitting within arm's reach of it on your desk.

One could argue that since a 4K monitor is "only" 10MPix, that having anything of higher resolution is wasted, since you would never be able to view that natively unless you have cropped the image down.

With the slight exception that if you order prints, the extra pixels matter quite a lot. Also, you mention cropping. The extra pixels does give you a bit of leeway to crop an image. But I get you point. 

Thanks to you guys for giving me a bit more to digest. I looked at a QSI690 as a pretty good compromise. Would that in your opinion be an upgrade to the 6D?

[Davey-T]

The QSI683 will give you a bigger FOV compared to the 690 at slightly less resolution, it's what I use on my Star 71

Dave

 

[ollypenrice]

42 minutes ago, pete_l said:

A 4MPix camera is more than adequate. A full-HD monitor is "only" 2MPix and nobody complains about the image quality on one of them. Even if it is a 27-inch monitor and you're sitting within arm's reach of it on your desk.

One could argue that since a 4K monitor is "only" 10MPix, that having anything of higher resolution is wasted, since you would never be able to view that natively unless you have cropped the image down.

 

 

Well... I'm not so sure. 1) This assumes that you only want to make images which can be seen in their entirety at approachng full size on a large computer screen. This does not describe many astrophotographers and certainly doesn't describe me. I like to produce images which can be seen in their entirety on a large screen but which can also be zoomed in on so as to explore key regions in detail. I guess what I'm saying is that current screens aren't big enough to be satisfying, so we image beyond their borders and then scroll around the image to enjoy it in a different way.

2) Having more pixels than the PC screen makes your instrument more versatile. You can shoot small objects which would not give a good scale if fitted to screen, but which can be cropped and zoomed in on whie still being well resolved.

My key point is that the PC screen size and resolution is arbitrary and not, for me, a determining factor in choosing kit.

I do take note of useful resolution, though, and wouldn't image below about 0.8'' per pixel since I don't think it can be reliably resolved.

Olly

[Gina]

I'm with Olly on this - I've never even thought of matching my astro image to the screen resolution.  Cropping and the ability to zoom in to examine an image in more detail have always applied for me.