Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

Questions about astroimaging (or cosmoimaging?) with high end CCDs: Sony A7s or Mallincam Universe?


ramdom

Recommended Posts

I'm new to this so please forgive my ignorance. I just got my first telescope (a Celestron AVX 9.25" SCT) and while my passion is high, I'm also finding I keep having to do things to make my experience better. There are so many things I could write about but my main point is that if I knew some of the things I know now then I'd have made different choices. As it is, I got a 6SE first and then returned it since it wasn't fulfilling my expectations. But visually I'm happy with the 9.25" and it's manageable even though I have to do several trips to strip it down and build it up.

Anyways, my lack of satisfaction comes from not seeing the nice images people are taking with their CCDs visually. I realised why a human eye can't see images like that and so a CCD-based solution is the only answer particularly before winter comes again.

So having decided that, I'm trying to get something that can take great images but can also work in NRT. So both sensitivity and specificity (detail) is what I'm looking for, ideally in one package. I'm willing to compromise a bit one way or another: either with image quality or what "real time" means (waiting a minute is fine---waiting hours isn't).

The Mallincam Universe seems to fit this bill---it has a large sensor with large pixel size and produces a "live image" that can be monitored. Since my interest is in DSO imaging, this would probably work.

In another forum, there was a reference to the Sony A7s (which I'm unable to access right now) and I checked it out and I'm intrigued but I'm also ignorant about DSLRs having been content with my phone camera until now.  This camera has a larger sensor and more pixels and can output/record in utra HD. 

My questions are: if I get the Sony A7s, how do I hook it up to the SCT? Does it just go back to the bottom of the barrel? I believe I'll need a focal reducer and here Mallincam supplies a reducer which presumably fits its cameras as well as the scopes with either a 2" or 1.25" barrel (with appropriately spaced adapters). With the Sony apparently even though it is an alpha it uses a different ring and I'm not sure if I should get a 2" diagonal and then connect the Sony to that or is some other (hopefully easy) of how to a connect a mirrorless camera like the a7 series to an SCT? 

Here're a couple of sites that show some images and explain how it works:

http://www.mallincam.net/universe.html

http://petapixel.com/2014/07/30/sony-a7s-astrophotography-review/

The latter review is impressive but it is AP with a lens on the camera, not through a scope. 

--Ram

PS: I think knowing what I know now, and that I'd be considering the A7s, I probably should've gotten the EdgeHD.  I wonder if if a normal 9.25" OTA can be upgraded to have the EdgeHD mirror...

Link to comment
Share on other sites

  • Replies 28
  • Created
  • Last Reply

Hi, not to sure what you are trying to do. Do you want to produce images or use the camera as an observing aid to do video imaging?

I think your scope is F10 isn't it? Without a reducer that will be very slow for imaging. If you do get a reducer it should be suited to your scope and then you need to ensure the correct spacing to the camera.

It is not normal to use a diagonal for imaging.

Sent from my iPad using Tapatalk

Link to comment
Share on other sites

Hi D4N, both. I'd like to have an observing aid and get as high a quality as possible of the resulting image I see captured if I want to. Obversing aid doesn't necessarily translate to RT for me - somewhat NRT is okay.  I'm trying to figure out how to make a Sony A7s work with a Celestron AVX 9.25" SCT. I've googled and gotten some info but it seems people who are doing this already know a lot about AP so it's unclear to me how exactly they're hooking up the A7s to the scope.

Yeah, that's what I thought: the reducer + spacing is important. With the Mallincam Universe, you get Mallincam's reducers and filters which not only seem good, but seem to work with their cameras. With the Sony, unless someone has this specific setup and has done it and posted their results, it looks like trial and error.

Thanks.

--Ram

Link to comment
Share on other sites

...

So having decided that, I'm trying to get something that can take great images but can also work in NRT. So both sensitivity and specificity (detail) is what I'm looking for, ideally in one package. I'm willing to compromise a bit one way or another: either with image quality or what "real time" means (waiting a minute is fine---waiting hours isn't).

...

If I have learnt anything in astrophotography then it is that good images and impatience exclude each other. The example image for the Mallincam is M42, one of the brightest nebulas in the sky, a f/4.5 scope and about 14 minutes exposure. With a slower scope and a fainter object this camera will require hours of exposure time. The Sony A7s is a full frame sensor camera and I am not sure whether the SCT will illuminate the whole sensor. There is a thread somewhere in this forum about this camera. By all accounts it is a beast :smiley: .  Attachment would be, like with any other DSLR, through a T-ring/adapter suitable for the Sony A7s (I think it's a E-mount). But even with this camera I doubt that the one-minute time frame is realistic.

Link to comment
Share on other sites

The two requirements of NLT and high quality DSO imaging are mutually exclusive I'm afraid.

You could use your existing telescope but please bear in mind that longer focal length instruments like SCTs put extra demands on your mount with regard to tracking during longer exposures.

Sent from my iPhone from somewhere dark .....

Link to comment
Share on other sites

IMO focus on the scope strengths rather than its hard to do ability's, pop a web cam in and start imaging the moon / planets, this will be rewarding after you have got it mastered and will also introduce you to the requirements of looking at DSO imaging, to do DSO's with a 9.25 you do need a capable mount, I know Sara got good results with hers on a avalon. (although she was not happy with them ;) )

Link to comment
Share on other sites

I think you need to detach long exposure, deep sky, stacked, calibrated and carefully processed images from anything at all that is 'semi instanateous.' We wouldn't spend 22 hours capturing, and then a further few days processing, images like the one below if we could just point and shoot and get the same result.

M51%20DEC%20VERSION%20clip-M.jpg

The video cameras which integrate a few minutes' worth of data and present it in an auto stretch in semi real time can indeed show some depth and detail but if you have been disappointed not to see laboriously crafted images at the eyepiece don't run the risk of being disappointed by the video camera a second time. 20 hour images use twenty hours' worth of data for a reason!

I'm just urging the management of expectation, really.

Olly

Link to comment
Share on other sites

You have a very nice visual scope in the form of the 9.25, unfortunately that does not make it a very suitable astro imaging scope.

For the astro imaging, especially the DSO items you really need a short fast scope - there is a post somewhere that asks why is the 80mm apo the chosen/preferred scope for DSO imaging. The answer is simply that it does the job very well with few negative aspects.

A 9.25" SCT is far removed from an 80mm apo.

The mount you have is well suited to imaging, and so the answer is a different/additional scope, at least for the DSO imaging aspect.

The 9.25" SCT will be well suited to imaging Jupiter and Saturn. Basically scope, barlow (2x or 3x) and a webcam. Get a video of the planet and process the best 20% of the frames. Limit Jupiter to 90 seconds.

Would guess an A7 is not ideal, if as someone has said the sensor is full frame I am not sure how good the final image will come out, the usual problem is field curvature and the bigger the sensor the worse it is. The other aspect is that the A7 is not a dedicated astro imaging camera, so although good there will be compromises. First one is that many images you have seen will be via a mono camera not a colour one, and the A7 is colour, it cannot be cooled is another.

Least present costly route would be to go and try planetary imaging, you likely have all the bits required, that will occupy you for some time.

Debate a more suitable AP scope, maybe look at the used market for something like an ED 80, likely someone will advertise one in the next 3 months.

Camera, not sure about, going to duck that one.

I say the next 3 months as the summer and the shorter nights are prevalent and you likely will not get much chance to go DSO imaging for a while.

Any clubs near to you to visit?

If so I suspect that looking at imaging set ups will help a lot, theree can be a vast range but that is mainly in cost.

Link to comment
Share on other sites

I agree with Olly that it is about managing expectations. At the same time I'm constantly surprised at just what is visible in near real-time, in my case with an 8" f/4 Newtonian and Lodestar X2 mono camera (mainly posted on the video forum on SGL). Perhaps I'm easy to please.  :smiley: 

But seriously, with this setup you can observe objects whose light set off when the universe was only a sixth of its current age; you can pick up 100s of faint galaxies in a single frame; you can see the amazing effects of gravitational interactions between pairs of galaxies; and for brighter galaxies and nebula the amount of detail in star forming regions is just remarkable. See the central star in M57 in one second. The trapezium in M42 blows out in a similar time and the view of dark nebulae is also stunning. OK, some of what I look at (well, most…) is rather faint, but the back story is what interests me.

Martin

Link to comment
Share on other sites

You guys have given me great advice. Yeah,  I think starting small and then learning the ropes that way may be a good idea. If my initial outlay is low and I get a year or so of use out of it then I won't feel bad about discarding a year after when I get the high end version (I'm not catholic, but I already feel guilty about spending so much on this hobby when there are so many poor starving people, etc.). I wish I could try them all out - then I'd know. 

The problem with planets is that they're not always visible. I think initially focussed on the biggest and brightest DSOs would be fine and as I said, I'm looking for something to complement my visual observation that can take still images also.  Eventually I realise know that I'll have to get a small refractor 80-100mm to complement my C9.25" if I remain serious about this hobby---I'm thinking carbon fibre explore scientific triplet or a takahashi.

ollypenrice: that's an awesome image (as an aside I'd like to know how you did it exactly---beautiful colours). I am not expecting that level of quality with a point and shoot approach. My expectations are the images I've seen on the Web and nightskynetwork.com and many are using a c9.25" and a Mallincam (usually). 

--Ram

Link to comment
Share on other sites

To illustrate my point, I use Slooh also and I've captured some great images there by 10 minutes of observation and essentially a point and shoot approach. I'm okay with that outcome. Of course, the telescopes are slooh are much larger and they're at a dark site, etc.  



Martin, the way you're using your scope is the way I'd like to do it.  I just looked at the milky way last night (without a scope) in relation to ourselves and what I thought was wisps of cloud is really the dust clouds that you can see when you look edge on the galactic plane... the Sony A7s images that reviewer did without a scope are amazing in this regard.



Olly:  I'm okay with doing 20 hour exposures and stacking and processing, etc. for a final image. I'm looking for an option where I can roughly see what I am taking an image of. A 20 hour video and then processing all the good frames to get a great image like yours would be the very ultimate goal probably after a few years.... :)



But the idea of starting small and then going big has appeal: I've actually spent most of my time in the couple of weeks since I've got my scope to look at the Jupiter. I could occupy myself with the sun too. I'm afraid to see the sun since I have only one functional eye (lost one to a bottle rocket when I was 10) and if something goes wrong then I'll be blind. So imaging sounds like something that would be fun...

Link to comment
Share on other sites

To give you an idea of what things look like: The following picture is taken with a Pentax K30, a SW 80ED (f/7.5) on a HEQ5 Pro mount. 60sec exposure at ISO6400 - no processing just resampled to make the size manageable.

post-39098-0-65014300-1432093231_thumb.j

You can see the smudge of M83 in the centre of the image so you know you are on target and you can zoom in on a star to see whether you are in focus... Then you start taking frames (in this case 148) and you do all the post-processing you can think of and you get something like this:

post-39098-0-19654500-1432093974_thumb.j

And after all this, you are still a long way off from where Olly is, just because he has the experience (and the better equipment) which I don't have (experience and equipment that is).

And yet, the journey is half the fun if not more.

Good luck!

Link to comment
Share on other sites

Hi Ramdom

To continue with M51, with a sensitive camera like the Lodestar this is the kind of result you can expect. After waiting 30s for a single frame:

post-11492-0-64283200-1432103901.png

Then just let the excellent free LodestarLive software perform live stacking, and 30s later you have

post-11492-0-72835400-1432103960.png

Let this go on for as long as you want to observe. After 4 x 30s you can get rid of half of the noise to obtain a reasonable image.

post-11492-0-32155300-1432104003.png

Then its a case of diminishing returns

post-11492-0-35563700-1432104144.png

This is with the mono camera. With the colour version you'll get some colour but nothing like what Olly has!

If you do go own this route, do check out images from different cameras and software. Not all are equal in this respect and I hope the above demonstrates how powerful live stacking is. I should also mention the ease of use of LodestarLive e.g. the ability to do nonlinear stretches in real time, seeing the effect immediately, which is also I think an essential component of near-live viewing. See my recent thread on the effect this has on globulars on the video forum if you're interested. It also performs live dark subtraction (though not yet flats: my images above could benefit from this!)

cheers

Martin

Link to comment
Share on other sites

Hi Martin and hjw, thanks for illustrating how imaging works. Yeah, I don't expect such amazing pictures as what Olly has without a lot of effort and skill but what you have looks very nice.  I'd be satisfied if I could view whatever my telescope pointed to within 30 seconds and got a picture like Martin's first one, or even doing it for a few to several minutes to get the end result of hjw's picture. 

I'd like anything I buy to do more but for initial use that looks very good.  

You know, when I look at the milky way axis through my telescope it's very faint but I can see the zillions of points of light in the background which you can't see with your naked eye.  It's just opened up a new world and I realise how weak our eyes are - plus I lost one eye to a bottle rocket accident when I was 10 and since then my night vision has been terrible.  So initially pointing to the Milky Way axis and exposing and 

Olly, yep, I agree with regards to Martin's phrase - I do research in complex biological systems (bioinformatics and computational biology) so the physics and science behind what I'm seeing is what attracts me to cosmology.  I also like your phrase about getting close to the universe - we already are physically with only a thin atmosphere separating us from space. 

--Ram

http://ram.org

http://compbio.org

Link to comment
Share on other sites

You should also be aware that Martins images are at f4, you would need something in the region of over 3 minutes at f10 to get a similar level of data to his single 30 second sub.

A reducer can help with this but you can get other problems then and will need to check that the light cone will still cover the whole sensor.

Sent from my iPad using Tapatalk

Link to comment
Share on other sites

The Sony A7S will work well with the Celestron in real time - I am able to just make out the spiral arms of M51 in liveview on my C11, from a darkish site.
 
A 90 sec exposure from an unmodified A7S on a C11 with 0.8 reducer looks like this:

post-19658-0-87385600-1432158694_thumb.j

This is a crop at full 1:1 scale at ISO 2000.  I've subtracted the light pollution, so from a properly dark site you would get a much better result than this. It's uncalibrated, which is why you can see hot pixels.  But the image will not be sharp out to the edges of the full-frame sensor because, as far as I know, you cannot get a large enough corrected image circle from an SCT.

I think for near-instantaneous continuous capture and image integration it could produce excellent results, when cropped.

Unfortunately, I don't know anything about the Mallincam so I'm not able to say if the noise levels are as incredibly low as the Sony.

Mark

Link to comment
Share on other sites

sharkmelley, thanks! I think this is the first time I've seen images of an A7s via a longer focal length scope (I assume yours is also f/10 unreduced). That really makes a point about any doubts about the a7s I might have. I was thinking of getting the Mallincam MFR-5 II reducer which is a 0.33 reducer which is a two cell unit. How did the image look on the LCD screen?

So that's great for 90s live image - now if you take a video of that or if you do multiple 90s snaps can you stack and process them can you refine the image? 

In other words, is taking a long video with short exposures for each frame and stacking frames the same as doing multiple snaps with longer exposures and stacking?

Thanks!

--Ram

Link to comment
Share on other sites

Hi Ram

On the question of whether multiple short subs (N x M seconds) are equivalent to a single M*N second exposure, the answer is…. hotly debated. But here's my understanding based on trawling around for information (and working in audio signal processing in a professional capacity). Leaving aside read noise (which you'll have N times more of in the short subs scenario), and everything else being equal, they are theoretically equivalent. The dominating noise source varies with the square root of the exposure length, and it doesn't matter if that is captured in one long exposure or several shorter exposures.  

The all other things being equal (which should precede all discussions in astronomy kit…) is almost never true, of course, so its worth listing some of the factors to take into account when comparing short vs long exposures. Short exposures make lower demands on tracking/guiding. Short exposures enable the occasional poor sub to be rejected. Thermal noise (dark current) is proportional to length of exposure. Shorter exposures are less likely to lead to saturation. (Most of these issues can be addressed  by calibration etc for those taking longer exposures though.)

My experience in practice suggests that stacking very short exposures (of 1 second, say) is not worthwhile due to read noise. But by the time I get to 10-15 seconds then stacking these seems to produce similar results to what would be obtained by stacking longer exposures. The reason I tend to go for 30s is that in live stacking mode if I do need to reject subs this needs to be done on the fly, and 10-15s is really too quick to both be rejecting subs and enjoying the image as it builds up. I don't go longer than 30s is simply because I use my mount in alt-az mode and field rotation can start to appear beyond 30s at my current focal length.

Craig Stark has a series of tutorials on signal and noise that are worth digging out if you're interested in the topic. Berry and Burnell's Handbook of Astronomical Image Processing, though dated in some areas, is good for principles.

cheers

Martin

Link to comment
Share on other sites

I used a 0.8 reducer giving F8.

I stacked 2 hours worth. See the result in Deep Sky section - can't easily do a link from my phone.

I could have used 2 hours of 30sec subs with no perceptible difference because the A7S read noise is very low indeed. But 2 hours of 30sec subs is a lot of data.

However stacking 30 sec subs in real time would be a great public outreach gimmick.

Mark

Link to comment
Share on other sites

Here are the links to the stacked versions of M51 before and after modding - raw stacks with only a stretch and colour balance applied in post-processing.

http://stargazerslounge.com/topic/242439-m51-whirlpool-first-light-with-sony-a7s/

http://stargazerslounge.com/topic/244140-m51-whirlpool-with-newly-modded-sony-a7s/

Please read my review before jumping in and buying one because there are plenty of negatives that could put you off:

http://stargazerslounge.com/topic/243671-review-of-the-sony-a7s/

If you can work around the issues listed then you'll be working with an incredibly performant camera.

Mark

Link to comment
Share on other sites

I'm one of the regular 'hot debaters' on the matter of long subs versus short. As a deep sky imager, often in search of the faint detail, I am, like everyone else who does this kind of imaging, convinced that you need very long sub exposures to go deep. I know of no practising deep sky imagers who would assert that lots of short subs will match a smaller (if respectable) number of long ones. That said, you are not (or were not originally) asking about this. If you want the outer halo of M51, or the tidal tail in the Leo Triplet, or the Integrated Flux Nebula, or the Squid Nebula, or the Soap Bubble, etc etc etc, you are going to have to buckle down to some serious DS work. But the posters on this thread have shown what can be done in near real time and I'm impressed.

However, my regular diet of 30 minute subs won't be going anywhere soon.

Olly

Link to comment
Share on other sites

Yeah, my original question was different and I think that was answered in more ways than one so I am grateful to everyone, including you, who provided illumination in this regard. I just ordered a bunch of stuff (StarSense, which I hope to try out tonight, a couple of filters, vibration pads, dew shield, extending and rolling chair, etc.) so I want to slow down a bit (and booked our vacation package to Disney World for a week).  I've been told I can borrow a Mallincam Junior Pro so I think that may be a good way to get my feet wet into imaging/VA/EAA.

But ultimately, hopefully before the next winter arrives, I think I will get the a7s, which has other uses besides astrophotography and given what has been posted here. I like it for a lot of reasons and I've seen some good deals on it. If I can get it hooked up then I hopefully can see the winter skies via wifi through it.

Thanks again!

Link to comment
Share on other sites

I'm one of the regular 'hot debaters' on the matter of long subs versus short. As a deep sky imager, often in search of the faint detail, I am, like everyone else who does this kind of imaging, convinced that you need very long sub exposures to go deep. I know of no practising deep sky imagers who would assert that lots of short subs will match a smaller (if respectable) number of long ones.

Low read noise changes the equation - big time.  The A7S has read noise below 1e at some high ISOs.  Compare that with the read noise of your Atik 11000.  You are absolutely right to keep to long exposures on astro-CCDs.  But it's no longer true for newer generation DSLR/Mirrorless cameras.

Mark

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.