Which CCD (or CMOS) Camera for Deep Sky Imaging

[Decimus]

Hi Everyone,
I have a budget of between $4000 - $6000(max) with which I intend to purchase a deep sky imaging camera. I know this runs counter to what those who seek the sharpest images would say, but for me, a monochrome camera is not on my short-list, simply because, with clear skies being so rare nowadays, the time factor involved in imaging with various filters would only reduce imaging opportunities. I have a Nikon D850 which is a great camera, but not suited to the deep sky imaging I would like to do. I am not especially interested in planetary imaging, and as for the Moon, I think my DSLR (when attached to my WO refractor) can readily match images from other cameras, CCD or CMOS. I will be imaging with the WO GT102 (F6.9) and a WO reducer/flattener 8.

This limits my choices to colour CCD or CMOS cameras, though I prefer the former. Does anyone have any suggestions? Cameras from QSI, QHY and ATIK come to mind, as well as the Starlight Xpress Trius series (like the 26C). A broad selection, I know :D

FLI models seem to be very expensive.

Does anyone have any suggestions/recommendations? All comments and suggestions welcomed.

Cheers,
Richard

[Atmos]

Do you have a flattener? Do you want a full frame or smaller sensor?

I can think of a few good colour CMOS sensors :)

[Decimus]

Hi Colin,
I have the WO reducer/flattener 8, still sitting there unused in its box since I bought it last year - it's BIG, too big, in my opinion for the WO GT102 (with which it is entirely compatible, according to WO), and IMHO, almost guaranteed to cause tube flexure in what is only a 4" telescope. I suspect I will sell it and buy something smaller.

As I have just discovered, sensor size is not everything, (pixel pitch, well depth, etc seem to be more important), but I guess a sensor of APSC or four-thirds size and up to full frame, would be great (and with a weight under 600g?) CMOS sensors are gaining ground (like the one in the ZWO ASI1600MM), but there are those that say CCDs still have the edge for overall image quality.

Would love to get some suggestions from you - and thanks for your reply!

Cheers,
Richard

[Atmos]

It’s on the cheaper side but the ASI071 is good. APS-C, good dynamic range, 14-bit output. On the upper side is the QHY367 but that may be more than the $6,000 budget.

[traveller]

Hi Richard,
Matching of optics to your camera is critical.
This is a good online tool for you to play with https://astronomy.tools/calculators/ccd_suitability
I have a ASI1600, which matched well with my FLT 110, and the ASI1600 (including the new pro OSC version) has a long list of happy owners!
Cheers
Bo

[Decimus]

Hi Colin,
Thanks for the advice. I certainly wouldn't mind the QHY367 which was near the top of a list of about 10 models. I was also thinking (dreaming) about the very pricey Atik 11 000 (9 micron pixels and full frame, their flagship)....if I keep saving....and then there's the Atik 490EX (within budget) and the Trius SX 26C (as mentioned earlier) which appears to be roughly APSC -sized CCD....

I guess I'll just keep looking until I am ready to buy, which hopefully, won't be too far away. Thanks again.

Cheers,
Richard

[gregbradley]

The idea that colour sensors reduce imaging time is not correct. They generally speaking have lower QE so they take total exposure time to get to the same level as filtered imaging with mono sensors. CCD cameras are typically around 60% QE for mono and about 28% or less for one shot colour (at least half as sensitive). I think the reality is more like 1/3rd as sensitive when all factors are considered.

But you can get caught out with an incomplete set of filtered data when using a mono sensor, so that is true. But then the colour data from the same time period from a colour sensor will be noisy so whilst you may get an image it may not be that pretty either.

I have used both and colour has its advantages but where you see its weaknesses most are in dust areas, dim areas of deep sky objects. Those will be noisy and are hard to get clean compared to filtered mono imaging.

So a 2nd hand QSI 683 WSG8 with filters is about the best setup I have seen in recent years. Many award winning images, the autoguiding function seems to work well and the smaller filters it takes allows using the better narrowband filters as the cost is kept done.

You would easily get one in your budget range. I see them on Astromart or CNN for about US$3000 - $3500 with filters.

Are these ASI type colour cameras much better than a modded DSLR? Especially say a full frame modded 6D (US$1000 or less 2nd hand)?

Full frame trumps these little ASI sensors. They are tiny.

Greg.

[Decimus]

Thanks, Bo. The Astronomy Tools Calculator site is always on my list of favourites, but I usually use the other tools (like FOV calculator); so I'm glad you mentioned the CCD suitability tool which I had completely forgotten - it's very handy.

Re the ZWO ASI1600MM (monochrome), one of my friends down here in Tassie has one and the images he has produced with it are excellent. I don't know what the colour version would be like, but I have put that on the list too.

Thanks for the helpful advice.

Cheers,
Richard

[Decimus]

Thanks for your input and advice, Greg. I take your point about losing fine detail in nebulae with a colour CCD camera, but I suspect the QSI 683, magnificent as it is, will be a weighty set-up when laden with filters and could produce some tube flexure on my WO GT102? Hmm....Our weather down here is appalling and clear skies increasingly rare; hence my thinking about taking images quickly!:D

Have you any experience with Atik CCD cameras? Or Starlight Xpress Trius models? No one seems to talk about them and I am wondering how good they are.

Thanks again.
Cheers,
Richard

[Atmos]

I personally wouldn’t bother with the Atik11000. Compared to the newer CMOS sensors it an ancient piece of antiquated hardware. That’s not to say you couldn’t get great images with it but for the cost buying new these days it’s just not worth it.

I have an ASI094 which is the ASI version of the QHY367 and is the same sensor as a Nikon D810A.
If you’re not wanting to get too serious and not wanting to do narrowband at the moment I’d stick with a OSC. More than anything, they’re less fuss. Once you get everything even semi automated it makes little difference but if you want to go and shoot for a couple of hours periodically, OSC is the way to go.

The ASI1600MC is good but the ASI071 is better in virtually every way. ASI might have a 24 MP full frame, can’t remember. The QHY367 is probably the best astro FF OSC on the market at the moment.

[Decimus]

Thanks, Colin. You have persuaded me to maybe save up for the QHY367. ( believe the QHY12 is also very good?) Failing that, the ASI071 may be a cheaper alternative.

Meanwhile, back to some imaging with the D850.

Cheers,
Richard

[Camelopardalis]

And getting caught short with a mono sensor is easily avoidable...in SGP (for example) you can select “Rotate through events” so that it changes the filter between each exposure and you end up with a complete set of R/G/B images so long as you complete more than 3 exposures ;)

I have both a full frame DSLR and a tiny ASI(1600MM) and the results are...different. On the same scope, the tiny sensor with the tiny pixels will give you better resolution on a good night, obviously at the expense of total FOV. It really depends on your angle...

The mono sensor is more versatile in that it can be effective with narrowband too...not impossible with a DSLR, but not optimal.

[Decimus]

Thanks for your comments, Dunk. As you say and as I have been discovering, some small CCD sensors with small pixel size can indeed deliver better results than a full frame DSLR. Just makes choosing the right camera even more bewildering!

Cheers,
Richard

[Atmos]

I’d suggest it comes down to whether you want a really wide field or narrower and higher resolution. If you want the wider field then a FF OSC can be the way to go where as if you want resolution then a smaller sensor is on the cards. A mono ASI1600/QHY163M or even a colour ASI183/QHY183.

[Camelopardalis]

Richard, it might help you if you decide on some targets that you’d like to go after, and go from there.

For large, sprawling nebulae, you might want the larger chip size, but for galaxy chasing the a small sensor with tiny pixels might be a more appropriate tool for the job (besides a larger scope!)

Hint: if you look in the imaging forum, some of the “old faithful” are joining the CMOS party. Not that there’s anything wrong with some of last gen CCDs, they’re just more expensive and not inherently superior to a good CMOS sensor...

And for your quoted budget, you can have a full frame DSLR _and_ a cooler CMOS camera ;)

[Slawomir]

Hi Richard,

Some excellent advices were already given.

I used Atik 428ex and it was a nice camera, but I needed a more powerful cooling (and wanted a larger FOV), so I upgraded to QSI 690 WSG-8. It is quite heavy, with guide camera and filters the mass is about 1.6kg, but my standard focuser on TS 102 handled it well, so I suspect your WO should cope well too with this much mass hanging off it. Had my QSI for five years and although tempted by a larger real estate, I cannot see a worthy candidate for an upgrade at this stage. Low RE (measured 3.9e at -25C, 4e at -15C), high QE, no need for darks nor bias at all - when I experimented with my attempt at the Helix with 100+hrs of total exposure, darks + bias made measurable (in PixInsight) but not visible improvement, when stacking over 150 subs. These days I use 20min narrowband subs so I reach desirable for me SNR with about 30-40subs. With LRGB filters, a few minutes is plenty. I think this particular model of the camera really shines in narrowband when combined with a low to moderate focal lengths.

I would also check new offerings from Moravian cameras.

Regards
Suavi

[gregbradley]

Now you’re talking Richard. That qhy367 is the go mate.
If I were tempted by a osc it’d be that one. How different
would it be compared with a D850 though? Stock D850?
A modded D850 perhaps although I would
check reviews for its use for Astro. Nikon seems
hit or miss for Astro. Some models are great and others -hmmm.
Nikon does some spatial filtering and they also wont do proper flats
as well as not a proper long exposure noise reduction. Again depending in the model.

SX Trius, yep not bad. Atik? - dont know.

Greg

[Atmos]

Apparently the D850/Z7 has a lower QE than the D810A/QHY367/ASI094 but I get 14-bits of dynamic range which is better than the KAF16803 with 3.1e- read noise.
It’s a good chip

[Camelopardalis]

The troubling thing with the new Nikon mirrorless cameras are the reports of fixed pattern when pushing the shadows, that doesn’t exist in their DSLRs.

The spatial filtering is also a concern, if it can’t be disabled. This is less severe than Sony’s star eater, but it’s still monkeying with the signal...

[billdan]

If you don't mind buying a second hand OSC, there is a QHY10 (APSC 6 micron pixels CCD) for sale and a QHY168 (APSC 4.8 micron pixels CMOS) for sale in the IIS classifieds, both at reasonable prices.

[Decimus]

Thanks Greg, Bo, Dunk, Suavi and Colin. All of my imaging mates down here are die-hard monochrome CCD fans, and when you see what such gear can produce, it's persuasive. I like my Nikon D850, but I noticed last night that 40 seconds at 800 (moonless night) on a Zeiss 135mm (attached to a dovetail plate on my Vixen mount) was more than enough to saturate the sensor; 60 seconds was overkill. A ccd would, presumably, take longer before it reached saturation? The QHY 367 is full frame cooled cmos, so I guess that the cooling and maybe one or two other tech tweaks, are what diffentiates it from an unmodded Nikon D850?Don't know.

The QSI 683 or the QHY367 might be the way to go....two very different beasts!

Thanks everyone for your helpful advice. Much appreciated.

Cheers,
Richard

[Camelopardalis]

Can’t argue with the 367C, quite a beast, but both QHY and ZWO have announced a full-frame mono CMOS sensor for release this year which have pretty interesting specs.

While the 8300 is an old stalwart, there are other chips on the market now with better noise contributions (and thus SNR).

No reason for a mono sensor to take longer than a DSLR, it all depends on what gain is used (ISO is just gain). After all, a DSLR is just a mono sensor with a permanent filter array printed on it. Also, your 135mm is probably faster f-ratio than your scope, so will expose faster. But there’s no way of knowing exactly what the specs of Nikon’s sensor are...whereas astro and scientific manufacturers publish typical figures for their cameras.

[gregbradley]

You sure its lower QE? I don't think so.

The D850/Z7 Panasonic S1R/Sony A7rii/Sony A7riii is a slightly modified Sony A7rii and iii sensor which is still the highest rated sensor on DXO Mark. Its BSI, has copper wiring, has refined microlenses and dual gain analogue to digital on chip converters.

The QHY367 is the older 36mp full frame sensor and for sure less QE as the A7riii sensor is backside illuminated with a larger surface area for light to hit as the circuitry is underneath. It also does not have the copper wiring which was a later Sony development.

The latest you-beaut from Sony that is not even in their own cameras yet is the QHY600 which is the next gen full frame backside illuminated (BSI) 60mp super fast readout sensor from Sony. I would expect to see that sensor in the upcoming Sony A7Riv camera probably coming out towards the end of this year or early next year.

That sensor is also available as a stacked sensor meaning they weld a DRAM memory chip on the back of it like in the Sony A9 which gives very fast response time and allows various features none of which would be that useful for astro but perhaps for planetary (60mp though is too large for planetary I would think).

Its the direction Sony is going in.


Greg.

[poppasmurf]

Anothyer choice for CCD is Moravian, both OSC and Mono. I personally love my Moravian.

Shane

[Atmos]

The saturation point is all to do with well depth and at 40-60s it shouldn't even come close to saturation on anything but the brightest of objects even at ISO800.
What aperture were you using on the Zeiss 135mm?

At ISO800 the D850 likely has near a stop of dynamic range over the KAF8300 or ASI1600 variants.

Lagoon Nebula at ISO800 (https://www.astrobin.com/full/297031/0/) with 180s exposures at F/5.2.

[Atmos]

I went and re-read this thread on CN (https://www.cloudynights.com/topic/666514-a7riii-vs-d850/) and I think I remembered a few snippets but not the context. I do agree with you that the BSI sensor in the D850 should have a higher QE due to the architecture so possibly what was being suggested towards the end of the thread is that the D850 has a lower QE compared to some of the other BSI DSLR sensors on the market at the moment.

[Decimus]

Hi Colin. My D850 was simply mounted on a dovetail plate and the Vixen Mount(no telescope) on Friday and I found that with the Zeiss 135mm at f2.8 and 30 secs exposure, ISO 800, the image was overexposed (Rho Ophiuchi)Of course, at f2.8, the aperture would be a whopping 48mm....I looked at your link here to your image of M8. A stunner!

I am still having trouble finding the SCP let alone taking images like that...Maybe I should just hold on to my D850 for a while. I always thought that these CMOS cameras offered greater QE, better fidelity with colour frequencies, etc, but if all I am going to get in a camera like the QHY 367 is a CMOS that's cooled below ambient and is otherwise no better than my D850, why bother?

Thanks again for the comments.

Cheers,
Richard

[Decimus]

Hi Colin,
one example of an over-exposed image of Rho Ophiuchi (over saturated) with the 135mm lens at F2.8 is attached here, (apologies it was ISO 1600, not 800 as I claimed); so too, one of my images of the Galactic centre with the Zeiss Milvus 35mm at F2.8 30 secs and ISO 800 (and a feeble attempt at processing the image in LR). My camera was just sitting on my mount which was not polar aligned (no scope and it's hard to polar align with a camera alone; the polar scope was useless). Star trailing is obvious and I guess the white balance setting in the D850 was wrong (I forgot whether it was Auto or something else), hence the ghastly sepia tone of all 107 images I took. Not much to salvage here - but my point is if I were tracking accurately, how much better would any of these images look with 3 or 4 minute subs? Is it cooling the CMOS that is critical? As I have a full frame bsi sensor, would it be better to buy a CCD camera as I cannot imagine the Nikon in its unmodded state, delivering much more image detail, and certainly not the quality of your ASI094 image of Rho Ophiuchi....

[Camelopardalis]

I’d speculate that there’d be more similarities than differences between the 094 and the D850, especially in winter. While Nikon undoubtedly had a hand in it themselves, technologically it would have drawn on a lot of the Sony tech.

The cooling isn’t going to make a lot of difference on a cool winter evening. On a hot summer one, it’ll certainly help keep the thermal noise in check (but the influence of this would be limited in 40 second subs). Down in Tassie, you have to consider how much that is worth to you.

Of course, the other advantage to the cooled camera is that you can create a proper set of calibration frames. But again, how much is that really worth for some widefield shots?

So long as you captured RAW data with your Nikon, then the white balance would be irrelevant since it’s usually just a profile for viewing with.

If you had the camera on the tracking mount, the quality of your images are going to be largely dictated by the aberrations of the lens the sensor is peering through.

[Atmos]

What you’re dealing with there is the setting of the black point and possibly light pollution. I’ve quickly run it through Lightroom on my phone.

Dropping the ISO down to maybe 200 and using some tracking for longer exposures will make a big difference. As Dunk says, there won’t be much of a difference in winter but during the summer having cooling can make a big noise difference.

[Decimus]

Hi Dunk,
These photos were taken inner city Hobart, with the city glow directly underneath (east), and facing south, my neighbours' lights were all on too, so I knew it was going to be bad, but the tracking was very poor without proper polar alignment. The Zeiss 35mm is perfection from F3.5 (half a stop down from the 2.8 I used); the Zeiss 135 is close to perfect at 2.8, but at extreme enlargement, some edge aberrations are present. Again, F3.5 is stunning. White balance/temperature can be altered easily in LR, thank goodness, as I hate the uniform dull brown cast of the raw files.

So are you saying that the 084 would not really offer significant gains over my unmodded D850 except perhaps during summer? Food for thought.

Thanks again for the comments.

Cheers,
Richard

[Decimus]

Hi Colin,
You are spot-on about light pollution (as I have just written in reply to Dunk). I had a crack at Rho Ophiuchi in LR but without real success...The 'dehaze' tool is by far the most dramatic in its impact but it adds some sort of surrealistic sharpness here, I guess because, as you intimate, there is not enough data here to work with in the first place. I'll go to our dark site next time and try longer exposures (with correct PA).

Thanks again.
Cheers,
Richard

[Atmos]

The biggest improvement you’ll get with any astro OSC is the higher Ha sensitivity over an unmodded DSLR. It’s the main reason I bought the ASI094 while having a D810. The D810 is my every day camera but lacks the Ha sensitivity.

[Decimus]

Thanks again, Colin. Will wait a bit longer (ZWO have a new sensor coming out apparently) before I take the plunge.

Have been looking at your amazing images on Astrobin. Fantastic!

Cheers,
Richard

[gregbradley]

In my experience mono cameras with filtered imaging deal with light pollution better than OSC. Something to consider.

Plus mono gives the option of narrowband as Andy has shown so well. Narrowband is not really an option for OSC. Ha perhaps but not that deep.

Greg.

[Slawomir]

I would also go for a mono - and to keep things simple for a start one Luminance filter and one Ha filter - mono images of DSOs can be as beautiful as colour ones :thumbsup:

[Decimus]

Thanks again, Suavi Greg and Colin.

The points you all make are also featured in a video I watched this afternoon from the Astro Imaging Channel : https://www.youtube.com/watch?v=d43vhG5R8xo - which, in my case is appropriately called 'Moving from a DSLR to Dedicated Astro Camera'.

Cheers,
Richard

[Decimus]

Thanks for this suggestion, Shane. Will add Moravian models into my look-into list.

Cheers,
Richard