Camera Labs

.
The finished (I hope!!!) image can be seen in its own thread here.

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

Hi folks,

I thought it might be fun to present a 50% crop of the North America nebula taken from the image in this post.

North America Nebula (NGC7000)

The image processing was covered in some detail in that thread whilst the shooting information for the source image was covered a few posts ago.

When I look back on my efforts towards the beginning of this thread I think it's fair to say I've come a long way. Maybe when I finally get my TEC 140 telescope (it's been on order since April!) I might have even farther to travel. I hope so.

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

Hi folks,

This is an almost worthless comparison but here goes:

The sharper image was taken with my dedicated astronomical CCD camera (FLI ML16803) by combining three 200 second exposures through red, green and blue filters (equals 600 seconds of red, 600 seconds of green and 600 seconds of blue). The lens used had a 165mm focal length, was fully open at f/2.8 and the pixel width is 9 μm. The less sharp one was taken with my 5D Mark II by combining fourteen 120 second exposures. The lens used had an 85mm focal length, was full open at f/1.2 and the pixel width is 6.4 μm.

I used PixInsight to register some appropriately sized crops and then opened both in PhotoShop. I left the ML16803 image alone (it is just as it originally left PixInsight after the RGB images were combined except for a white balance adjustment) and then tweaked the 5D2 image to match the colour of the brighter nebulosity as well as I could (at the expense of adversely affecting the star colours and less bright nebulosity ) and then applied additional processing to the 5D2 image to reduce the star sizes. Yes, those stars were even larger, though not hugely so.

The comparison is unfair because:

1. The lenses used were different and the 85mm may well be less sharp at f/1.2 than the 165mm at f/2.8.
2. The exposures were different and taken on different nights.
3. The ML16803 had a 40% resolution advantage after factoring focal length and pixel size.
4. The 5D2, like any unmodified DSLR, has filters which cut out the deepest reds (Hα) picked up by the ML16803 with the red filter in place.
5. The exposures weren't made with a view to comparing performance. I'm just offering what I've got to hand.

I am confident that I achieved the best focus that I could with the 5D2 (live view) so is the 40% resolution advantage of the ML16803 enough to explain the difference in star sharpness? Probably not, IMHO, so that leaves the anti-aliasing filter of the 5D2 (the ML16803 doesn't have one as it is a monochrome sensor) as a prime suspect. As most know its purpose is to defocus the light before it reaches the sensor so that each of the red, green and blue pixels in the sensor's Bayer matrix get a chance to respond. The 5D2 is going noticeably less deep than the ML16803 but it's hard to quantify why. Different nights, different exposure times, AA filter plus Bayer matrix, higher QE from the ML16803...

Am I suggesting that DSLRs just can't compete with dedicated CCD cameras. Absolutely not! A dedicated and cooled CCD can certainly do stuff that it would be very hard to replicate with a DSLR but for a good deal less than the cost of a dedicated astro CCD camera plus associated glass one can buy a good DSLR, have it modified to let through more of the red stuff if that's what is wanted and buy a pretty fancy lens with a moderately fast f-number to make up for some of the DSLR sensor limitations. The large sensor in the DSLR will still likely offer wider fields of view even with a slightly longer focal length lens to make up for the blurring of the AA filter.

No, I don't regret for a second going for my ML16803 as, with a cooled high efficiency monochrome sensor, it will enables me to do stuff I couldn't otherwise easily manage and the lack of sensor noise at -25°C is wonderful. But a DSLR is still a darn fine tool for astrophotography.

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

Whoa, I almost don't understand anything, but the difference between the two images is soooo huge, it's almost a no-brainer.
So why the long explanation

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

Hi Thomas,

It would be good to see how well a modified DSLR can do. My 5D2 blocks most of the Hα which will account for a lot of the difference. Star sharpness can probably only be gained by using a longer focal length lens. A DSLR couldn't, in all fairness, compete if one was after the really faint stuff as thermal noise would get in the way but the North America nebula is relatively bright.

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

Focal length and pixel pitch seem to support your story but I couldn't believe that this is the major reason for the huge difference.
So in layman's terms: what is the most most important difference in the setup of both versions that causes the IQ-difference?

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

.
I'm not sure there is a single main factor. Some astronomical CCDs are colour so it can't just be the Bayer matrix. For nebulosity my guess in answer to your question is the IR filter but maybe popo can answer that one as he has a modified DSLR?

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

Ok, I'm only some 3 weeks late to this thread but I must have missed it at the time and just found it looking for something else.

On the resolution differences, I'm going to take lens and sensor in turn.

Lens focal length difference: 1.94x more for the CCD system.
Pixel pitch: While the 5D2 has a finer pitch, the bayer pattern only have 1 in 4 red area coverage where most of the H-alpha would sit. I would use a rough rule that a bayer pattern camera is only half its nominal linear resolution when it comes to fine colour. That's 1.42x more for the CCD system. I'm assuming the AA filter is just the right strength here.

Combine the above I make it the CCD image has around 2.8x the linear resolution of the DSLR output.

On the DSLR IR filter removal, I don't believe that has any quality benefit other than increasing the sensitivity and related factors from that.

I can offer a H-alpha only output from the 450D with a 135mm lens from some data I gathered before I got PixInsight, so I'm reprocessing that now. By my calculation, the CCD system should only be 1.4x higher linear resolution than this combination. Will post that later when it's done...

Of course, all the above totally ignores field of view coverage, and is only looking at pixel level resolution.

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.

Here it is, hope Bob doesn't mind I linked his earlier image for ease of comparison. While I present mine in greyscale, it is only the red channel output from the DSLR. I have tried to rotate and scale it by eye similarly to Bob's earlier sample. It is roughly 50% linear scale, so as mentioned earlier due to the bayer pattern I don't think I'm losing much at this shrinkage level.

Oh, looking at the source, the focus was off on this part of the shot. I have a suspected deviation between lens axis and sensor plane so the two are not adequately perpendicular. Which means one part of a shot can be in focus where another isn't. So possibly if things were perfect I'd have better resolution.

From my earlier calculations, all else being equal, Bob's CCD system should be 1.4x the linear resolution, and Bob's 5D2 output about half.

I used 12x2+8x4 minute exposures to get the above. Never mind the focus, my tracking isn't perfect either...

_________________
Canon DSLRs: 7D, 5D2, 1D, 600D, 450D full spectrum, 300D IR mod
Lenses: EF 35/2, 50/1.8, 85/1.8, 135/2+SF, 28-80 V, 70-300L, 100-400L, TS-E 24/3.5L, MP-E 65, EF-S 15-85 IS
3rd party: Zeiss 2/50 makro, Samyang 8mm fisheye, Sigma 150 macro, 120-300 f/2.8 OS, Celestron 1325/13
Tinies: Sony HX9V.

.
Interesting result and you are most welcome to use my example for the comparison. I think it tends to bear out my assertion that DSLRs can be a great tool. I still wouldn't trade in my CCD though...

Bob.

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.

Hi folks,

Still a preliminary result so far as processing is concerned but I thought I'd share my first Hα mosaic of the main stars of the constellation of Orion. The original is 4000 x 6500 pixels but I've linked to a 1024 pixel wide version here and here is a forum friendly 1024 pixel high version to look at:

Once I'm fully satisfied with the image I'll update it and post full details but essentially what one is looking at is the large area of faint Hα emission which, as you can see, is much more extensive than the Sword (M42) visible with the naked eye. The Flame nebula is visible just to the left of the left-most star in the belt of Orion and the Horsehead nebula lies below. The stars were added into the image from separate RGB exposures and I made an attempt to include the RGB data of the brightest nebulosity as well. That proved extremely tricky as, although PixInsight allowed me to easily composite 200, 40 and 8 second RGB exposures (HDR needed due to the extreme, er, dynamic range!!!), it took me quite a while to figure out how to coax PhotoShop into including that nebulosity together with the Hα in a way that wasn't totally gross! That compositing may still need more work and might actually benefit hugely from more narrowband imaging at emission lines other than that of Hα (that will be for next year) but the benefit even with the data I have right now is some natural colour from the Flame nebula and even a slightly fuzzy rendition, due to the small scale, of the Running Man nebula. I'm not totally sure about the colour in M42 compared to the artificial colour I use for the Hα nebulosity - I'd hate to use that colour for the extensive Hα but am I justified in altering it inside M42 to match the Hα? So many decisions!

I'm also not too happy with the way the brightest stars bloated out so maybe I need to revisit the area and capture a full set of 40 second exposures of the whole constellation rather than the more limited set of such exposures I used to do the HDR of M42. A work in progress...

Update: The final result was posted here.

Bob.

P.S. My previous effort with a DSLR (here) shows that my investment in time, money and, dare I say it, learning is starting to pay off. I still feel pretty close to the bottom of that learning curve, though.

P.P.S. Extra processing might have to wait as, after nine months, the new telescope (a TEC 140) has finally arrived. I'm in love...

_________________
OM-D E-M5 + ED 12-50mm, H-PS14042E, H-F007014E, L-RS014150E, M.ZUIKO DIGITAL 45mm 1:1.8.
Gitzo GT1541T, Arca-Swiss Z1 DP ball-head.
Astrophotography: TEC 140 'scope, FLI ML16803 camera, ASA DDM60 Pro mount.