585mc or a6000 for DSO with 8" F5 unguided?

[MichaelBibby]

Zwo asi585mc or Sony a6000?

I have a 8" F5 (1000mm) Newt on a belt modded Heq5 pro and am interested primarily in DSO.
Given the fact that I'm pushing the weight limits of my mount, and I aren't using a guidescope, I will be limited to short exposures (between 1 - 30 sec). Which camera will make the most of these limitations: the Zwo asi585mc/player one Uranus C or the Sony a6000? And why? I know the 585 has 2.9um pixels and the a6000 has 3.88um which would be more suited to my focal length, gather light faster and be more forgiving of tracking problems, but what other factors/specs should I consider?

Edited October 2, 2022 by MichaelBibby

[Adam J]

On 02/10/2022 at 05:28, MichaelBibby said:

Zwo asi585mc or Sony a6000?

I have a 8" F5 (1000mm) Newt on a belt modded Heq5 pro and am interested primarily in DSO.
Given the fact that I'm pushing the weight limits of my mount, and I aren't using a guidescope, I will be limited to short exposures (between 1 - 30 sec). Which camera will make the most of these limitations: the Zwo asi585mc/player one Uranus C or the Sony a6000? And why? I know the 585 has 2.9um pixels and the a6000 has 3.88um which would be more suited to my focal length, gather light faster and be more forgiving of tracking problems, but what other factors/specs should I consider?

You want low read noise, note that the ASI533mc is now available uncooled. But I believe you will get lower read noise with the Uranus C. 

[MichaelBibby]

17 minutes ago, Adam J said:

You want low read noise, note that the ASI533mc is now available uncooled. But I believe you will get lower read noise with the Uranus C. 

I can't see any information about an uncooled version of the 533 anywhere, I imagine that would be big news if true.

Something else I only just learnt is that the a6000 doesn't work with SharpCap, which is a problem...
 

[vlaiv]

11 minutes ago, MichaelBibby said:

I can't see any information about an uncooled version of the 533 anywhere, I imagine that would be big news if true.

https://www.teleskop-express.de/shop/product_info.php/language/de/info/p15858_.html

Not translated into English yet.

There is also official ASI version, but won't be available from TS for next 12 days:

https://www.teleskop-express.de/shop/product_info.php/language/en/info/p15856_ZWO-Farb-Astrokamera-ASI533MC-ungekuehlt--Chip-D--16-mm---3-76--m-Pixel.html

[Adam J]

13 minutes ago, MichaelBibby said:

I can't see any information about an uncooled version of the 533 anywhere, I imagine that would be big news if true.

 

and yet here it is:

https://www.firstlightoptics.com/zwo-cameras/zwo-asi-533mc-colour-usb-30-camera.html

[vlaiv]

15 minutes ago, MichaelBibby said:

Something else I only just learnt is that the a6000 doesn't work with SharpCap, which is a problem...

Check to see if Nina or other long exposure capture software is capable of using Sony cameras.

I think that APT certainly is (at least Canon DSLRs).

Also look up to see if there is ASCOM driver for Sony camera - then you can use it with any application supporting ASCOM cameras.

SharpCap is usually not the best for long exposure astrophotography. Planetary and live stacking is where it is at its best.

[MichaelBibby]

44 minutes ago, vlaiv said:

https://www.teleskop-express.de/shop/product_info.php/language/de/info/p15858_.html

Not translated into English yet.

There is also official ASI version, but won't be available from TS for next 12 days:

https://www.teleskop-express.de/shop/product_info.php/language/en/info/p15856_ZWO-Farb-Astrokamera-ASI533MC-ungekuehlt--Chip-D--16-mm---3-76--m-Pixel.html

Thats great! I know that a lot of people would be interested in an uncooled 533. Given my FL of 1000mm the 3.76um pixel size would be perfect.

So now I wonder: would the larger pixel size of the 533 offset the lower read noise of the 585?

Edited October 3, 2022 by MichaelBibby

[Adam J]

7 minutes ago, MichaelBibby said:

Thats great! I know that a lot of people would be interested in an uncooled 585. Given my FL of 1000mm the 3.76um would be perfect.

So now I wonder: would the larger pixel size of the 533 offset the lower read noise of the 585?

If you are doing really short exposures (1s) then it would not be unreasonable to use the 533 at gain 400 so about 1e read noise. SO I think yes its would offset. What you probably want to work out though it read noise per pixel area. Its worth noting that you would not be restricted to 1s and that it would easily do 30s exposures without the thermal noise being an issue on the 533. Cooling is really not required for your use case. 

Adam

Edited October 3, 2022 by Adam J

[vlaiv]

3 minutes ago, MichaelBibby said:

Given my FL of 1000mm the 3.76um pixel size would be perfect.

Don't think so, but that is my opinion.

Since you are starting out, I would advise you to go for 1.8"/px resolution.

With focal length of 1000mm that translates into ~8.72um pixel size.

That would mean that you should get 4.3um pixel camera and bin it x2 or maybe 2.9um camera and bin that x3.

Now ASI585 might seem to fit the bill - with its 2.9um pixel size, but have you checked field of view on that camera with 1000mm of focal length?

First - since it has 3840 x 2160 with bin x3 you'll actually get 1280 x 720 image size. Second, and probably more important - you'll get very small field of view - only 0.64 x 0.36 degrees.

Ideally when starting you want something that is middle ground like 2x1.5 degrees or similar.

[vlaiv]

11 minutes ago, MichaelBibby said:

So now I wonder: would the larger pixel size of the 533 offset the lower read noise of the 585?

Neither are very important as you can handle either with the way you image.

Read noise is only important when determining minimum exposure length that you should go for and that also depends on your light pollution or in case of camera that is not cooled - dark current at ambient temperature.

Pixel size can be (somewhat) changed by binning. Smaller pixels have certain flexibility in this regard as they offer more options when binning.

3.75um pixel size - next step is 7.5um and next is 10.25um

But something like 2.4um will have more steps in same range

2.4, 4.8, 7.2, 9.6, 11

(native, bin x2, bin x3, bin x4 and so on)

However in order to exploit that - you really need to know how to treat your data and do proper software binning.

[Adam J]

6 minutes ago, vlaiv said:

Don't think so, but that is my opinion.

Since you are starting out, I would advise you to go for 1.8"/px resolution.

With focal length of 1000mm that translates into ~8.72um pixel size.

That would mean that you should get 4.3um pixel camera and bin it x2 or maybe 2.9um camera and bin that x3.

Now ASI585 might seem to fit the bill - with its 2.9um pixel size, but have you checked field of view on that camera with 1000mm of focal length?

First - since it has 3840 x 2160 with bin x3 you'll actually get 1280 x 720 image size. Second, and probably more important - you'll get very small field of view - only 0.64 x 0.36 degrees.

Ideally when starting you want something that is middle ground like 2x1.5 degrees or similar.

I think this is a shoot off of another thread, short exposure technique (1000s of short exposures) was discussed by OP in relation to the performance of his mount / scope combination. Personally I would not recommend this route for a beginner either, but in reality he has the equipment he has and if he is dedicated then what he is proposing is not entirely unreasonable. 

Adam

Edited October 3, 2022 by Adam J

[vlaiv]

1 minute ago, Adam J said:

I think this is a shoot off of another thread, short exposure technique was discussed by OP in relation to the performance of his mount / scope combination. Personally I would not recommend this route for a beginner either, but in reality he has the equipment he has and if he is dedicated then what he is proposing is not entirely unreasonable. 

Adam

Oh I see.

I was under impression that we are talking about regular deep exposure imaging.

In any case, recommendation still stands - even when using short exposure, all above apply unless approach is lucky type imaging where exposures are in range of milliseconds rather than seconds.

Second to two is enough for seeing to average out at whatever FWHM it is going to have so there won't be any real benefit that lucky imaging provides (with discarding poor frames) - but yes, if tracking is an issue and that wants to be handled with short exposures then read noise is important factor - but you are absolutely right there - one should be looking at read noise per unit pixel area as read noise grows when binning (kind of binning we are discussing here - software).

 

[MichaelBibby]

Thanks guys, appreciate the time your taking helping me to understand all the variables. I'm a technician so capable of understanding all the details, just need to dedicate more time to it. And now you've given me some more lines of inquiry to follow up.

[Adam J]

2 hours ago, MichaelBibby said:

Thanks guys, appreciate the time your taking helping me to understand all the variables. I'm a technician so capable of understanding all the details, just need to dedicate more time to it. And now you've given me some more lines of inquiry to follow up.

Honestly a better option across the board unless you want to dedicate yourself to detailed imaging of small bright galaxies is to buy a 130PDS or 150PDS probably the former which are both well within the capacity for the HEQ5 and will serve you better for imaging than the 200PDS, you can always go back to it in the future. Then I would consider the 533mc to pair with it. 

As an technician am sure you will appreciate the concept of the right tool for the right job. 

Adam

[MichaelBibby]

Thanks Adam, I understand that would have made a more reliable and versatile combination. I knew what I was getting myself into putting this hulking scope on this mount, and took it as a challenge to see what I can get out of this setup. It will be mounted on a pier, shielded from the wind, under Bortal 4 skies, so I have some things going for me. I'll only explore guiding once I've pushed the limits of my tracking setup first. I'd love to have a second telescope for widefield imaging, and longer exposures, eventually too.

Edited October 3, 2022 by MichaelBibby

[MichaelBibby]

21 hours ago, vlaiv said:

Don't think so, but that is my opinion.

Since you are starting out, I would advise you to go for 1.8"/px resolution.

With focal length of 1000mm that translates into ~8.72um pixel size.

That would mean that you should get 4.3um pixel camera and bin it x2 or maybe 2.9um camera and bin that x3.

Now ASI585 might seem to fit the bill - with its 2.9um pixel size, but have you checked field of view on that camera with 1000mm of focal length?

First - since it has 3840 x 2160 with bin x3 you'll actually get 1280 x 720 image size. Second, and probably more important - you'll get very small field of view - only 0.64 x 0.36 degrees.

Ideally when starting you want something that is middle ground like 2x1.5 degrees or similar.

Thanks for your feedback vlaiv. I know the field of view of the 585 at FL 1000mm (I've been using this tool as well as Stellarium to get a sense of all that: https://astronomy.tools/calculators/field_of_view ). I'm most interested in imaging galaxies, which is why I opted for a 8" 1000mm over the more convenient 6" 750mm. I knew I was going to involve myself in difficulties, but I like challenges. I know that with guiding and careful balancing the Heq5 pro and 8" f5 can work quite well for exposures of a couple of minutes, but a lot of things have to go right for it to work out. I'll 'hypertune' my mount before I make any serious attempt at long exposures, within the limitations of an uncooled camera. And I do plan on exploring off axis guiding down the road too. Right now, I just want to play with some EAA with sub 30s exposures, and only intend on imaging when seeing conditions are quite good.

Money is a limitation, and right now I can't see a better camera option than the 585 for that price point. My main concern is that oversampling will be my biggest limitation, but my understanding is that under very good seeing conditions, using short exposures, this setup could work quite well.  ? But I am open to other camera suggestions (the uncooled 533 is probably a little too expensive for me, nearly twice that of the 585).

 

[happy-kat]

Was your other suggestion the sony because you already own it?

[Adam J]

43 minutes ago, MichaelBibby said:

Thanks for your feedback vlaiv. I know the field of view of the 585 at FL 1000mm (I've been using this tool as well as Stellarium to get a sense of all that: https://astronomy.tools/calculators/field_of_view ). I'm most interested in imaging galaxies, which is why I opted for a 8" 1000mm over the more convenient 6" 750mm. I knew I was going to involve myself in difficulties, but I like challenges. I know that with guiding and careful balancing the Heq5 pro and 8" f5 can work quite well for exposures of a couple of minutes, but a lot of things have to go right for it to work out. I'll 'hypertune' my mount before I make any serious attempt at long exposures, within the limitations of an uncooled camera. And I do plan on exploring off axis guiding down the road too. Right now, I just want to play with some EAA with sub 30s exposures, and only intend on imaging when seeing conditions are quite good.

Money is a limitation, and right now I can't see a better camera option than the 585 for that price point. My main concern is that oversampling will be my biggest limitation, but my understanding is that under very good seeing conditions, using short exposures, this setup could work quite well.  ? But I am open to other camera suggestions (the uncooled 533 is probably a little too expensive for me, nearly twice that of the 585).

 

The 585 will work well for short exposure imaging. honestly i would still recommend guiding though even with shot exposures, you will also need to be able to use plate solving without question as with such a small sensor at 1000mm focal length I don't like the odds of you reliably finding targets without it. My recommended starting point for such a setup is M82 as its bright and will work out well with short exposures. The Player one 585 is without a question in my mind superior in your use case due to lower read noise for higher maintained dynamic range. I believe that this is essentially the same camera as TS optics are selling under their own brand name. 

There are also some real used bargains about at the moment. Such as this:

https://www.astrobuysell.com/uk/propview.php?view=191449

You could image Galaxies in mono for a while before adding RGB data later when you can afford a filter wheel. You would bin it 2x2 in software. 

Adam

[MichaelBibby]

15 minutes ago, happy-kat said:

Was your other suggestion the sony because you already own it?

No, I could get a pretty good deal on one, thats why.

[vlaiv]

5 hours ago, MichaelBibby said:

I'm most interested in imaging galaxies, which is why I opted for a 8" 1000mm over the more convenient 6" 750mm. I knew I was going to involve myself in difficulties, but I like challenges. I know that with guiding and careful balancing the Heq5 pro and 8" f5 can work quite well for exposures of a couple of minutes, but a lot of things have to go right for it to work out. I'll 'hypertune' my mount before I make any serious attempt at long exposures, within the limitations of an uncooled camera.

Ok, so here are a few guidelines then. You mentioned that you are a technician, so I won't shy away from technical stuff then. I'll also be very brief and to the point as there is a lot to cover.

Regardless the fact you'll be using short exposures - anything expressed in seconds rather than milliseconds is long exposure as far as seeing is concerned. For long exposure, here is breakdown of what you can expect in terms of resolving the target.

There are three main components to the blur affecting the image.

1. Seeing - expressed in FWHM in arc seconds, and represents full width at half maximum of Gaussian profile approximation to seeing blur. It is measured with very large aperture on very good mount in course of 2 seconds for exposure (see - as soon as we step into seconds - seeing averages out).

2. Mount tracking / guiding performance. If you don't guide - this is nothing more than a guess. I'll give you some guides on what you can expect from HEQ5 type mount later on. If you guide - then you have measure of how good your tracking is and is expressed in RMS error in arc seconds.

Two are related by simple equation FWHM = 2.355 * RMS (for Gaussian profile). We always use Gaussian profile for approximation as it is fairly accurate approximation (central theorem) and easy to work with.

3. Aperture size. Here we approximate Airy disk with Gaussian profile. It holds true for perfect aperture but in reality, especially when using correctors (which correct over whole field but deteriorate on axis performance) this blur is somewhat bigger. When available spot diagram RMS is good alternative (and often more precise if one is using correctors or reducers).

Once we have all three values (measured or estimated) - total blur is square root of sum of their squares. This lets us get estimate for expected FWHM of stars in our image - which is in turn tightly related to sampling rate (as stars are point sources and their profile is directly representing PSF of blur).

Very simple relationship that you can use (but math behind it is not so simple) - is sampling_rate = FWHM / 1.6

Good seeing is 1" or below - that does not happen often. You have several websites that offer seeing forecast - and it is usually fairly accurate if you make sure you don't have any local influences (that can be very detrimental) - like properly cooled optics, no seeing disturbances around - like hot roads or large bodies of water or houses with heating and chimneys and so on.

More often seeing will be around 1.5"-2.0". That could be taken as average.

As far as mount performance goes - unguided performance usually depends on two things

a) periodic error

b) poor polar alignment

People always seem to blame poor polar alignment for star trailing - but in my view and experience - periodic error is much more responsible for star streaks and mount poor performance.

With either of the two you must estimate drift rate and limit your unguided exposure depending on wanted resolution (one you are aiming at).

For HEQ5 you can safely estimate that periodic error is about 35"-40" peak to peak. In fact, I once did recording of unguided performance of my HEQ5 and here is what it looks like:

Up down motion is periodic error (and you can clearly see how it periodically repeats - hence the name) - right to left drift is due to polar alignment. You can clearly see that PE is much larger in magnitude over shorter periods of time than PA error.

HEQ5 mount has period of 638s and if you have say 35" P2P periodic error - that means that mount will need to trail / lead - or drift in general for 70" (there and back again by Bilbo Baggins ). If drift is uniform (and it never is as seen from recording) then you would have drift of 70" / 638s = ~0.11"/s

This is important number - as drift due to periodic error will sometimes be more than this and sometimes less than this. If you take this to be reference then in 30s exposure - you will have ~3" of trailing on average. Half of frames will be less than this but half will be more.

You will probably discard worse than this (or even at 3" trailing). In fact in above recording you can clearly see how in some subs stars get elongated while in other are round. I think I used 1 minute exposures there on my HEQ5 (1200mm FL and 3.75um pixel size).

This is so that you can understand that there is percentage of subs that you will have to throw away if you don't guide and that percentage will depend on your tolerance for elongated stars and exposure length that you'll be using.

Back to resolution.

We have seeing that is around 2", when you start guiding - you can expect stock HEQ5 to guide at about 1 RMS, and we have 8" of aperture. If you use simple coma corrector - you'll get spherical aberration on axis and star bloat, but for sake of simplicity lets go with diffraction limited scope.

In those conditions - your final SNR will be ~3.14" FWHM or that will be about supporting 1.96"/px resolution. That is about 9.7um pixel size (so you know how much you'll have to bin based on initial pixel size - at least x3 if using 2.9um).

Further - most galaxies are rather small in size. Someone mentioned trying M82 - which is about 11' long or 660".

With 2"/px - that is only 330px. I'm just letting you know what you can expect. And that is with guiding (stock mount). Just tracked - you probably won't achieve 2"/px resolution due to additional blur.

I'm not saying this to put you off - but rather to prepare you.

If budget is tight - you might consider using simple web camera for guiding and modifying your finder scope for that role. Any sort of guiding will be better than no guiding at all.

On the other side of spectrum - when you tune and mod your HEQ5 - best you can hope to achieve is around 0.5" RMS guide error.

I once managed to go as low as 0.36" RMS and have a screen shot to prove it

But list of modifications that I did to my mount is:

1. All bearings replaced for SKF high quality ones

2. Mount cleaned, regreased, tuned

3. Periodic error correction recorded and applied (in EQMOD)

4. Saddle replaced for Geoptik dual Vixen/Losmandy variant

5. Rowan belt mod

6. Berlebach planet tripod

If we have diffraction limited 8" scope, 0.5" RMS guiding and happen to image in 1.5" FWHM seeing - we can hope to achieve about 2" FWHM or about 1.24"/px.

Even at this resolution - most galaxies will be just few hundred pixels across, and that is about as good as you can get (maybe down to 1"/px - in ideal conditions and with better mount and larger aperture).

In the end - I want to explain one more thing - when I say that you should aim for say 2"/px because that what your setup / sky can support - that does not mean that you can't image at 1"/px or even 0.5"/px. Sure you can, but two things will happen:

1. you will record image that is very devoid of detail when viewed at 100% zoom - since you are over sampling

2. As soon as you start over sampling - you are starting to have slower system than if sampling properly. Light is spread over more area and each pixel receives less photons / less signal. Less signal means less SNR, and astrophotography is all about SNR.

If you want large galaxy image (devoid of detail) - then I advise you to sample it properly to get best SNR and them simply enlarge image in software - result will be the same as far as detail goes - neither can pull detail out of thin air.

 

[Adam J]

Not my image a link to one posted on cloudy nights. 0.5s exposures at F5.3 with a 12 inch scope. I doubt he is resolving to the pixel scale but to be frank it's the highest resolution image of m82 I have seen from amateurs so the technique is doing something beneficial. 

[vlaiv]

6 minutes ago, Adam J said:

Not my image a link to one posted on cloudy nights. 0.5s exposures at F5.3 with a 12 inch scope. I doubt he is resolving to the pixel scale but to be frank it's the highest resolution image of m82 I have seen from amateurs so the technique is doing something beneficial. 

Quick inspection suggests that he is about x2.1 over sampled there.

Here is frequency response of that image:

As you can see - all signal is concentrated in central part, rest is just noise. Inspecting "edge" of frequency signal gives ~4.3 px per cycle (instead of x2 needed for proper sampling).

However, most of exposures are in 0.5s range, and I wonder how many are discarded?

 

[Adam J]

5 minutes ago, vlaiv said:

Quick inspection suggests that he is about x2.1 over sampled there.

Here is frequency response of that image:

As you can see - all signal is concentrated in central part, rest is just noise. Inspecting "edge" of frequency signal gives ~4.3 px per cycle (instead of x2 needed for proper sampling).

However, most of exposures are in 0.5s range, and I wonder how many are discarded?

 

No idea how many got discarded. All I am saying is that I doubt it would be as detailed as it is if he was doing even 30s exposures. 

[vlaiv]

9 hours ago, Adam J said:

No idea how many got discarded. All I am saying is that I doubt it would be as detailed as it is if he was doing even 30s exposures. 

Indeed, and I think it is a shame it is over sampled.

With such short exposures, there is not much signal in the image and every little bit of SNR helps to establish if sub is viable and should be used in stack and to help with stars used for alignment when stacking.

[RolandKol]

On 02/10/2022 at 05:28, MichaelBibby said:

Zwo asi585mc or Sony a6000?

I have a 8" F5 (1000mm) Newt on a belt modded Heq5 pro and am interested primarily in DSO.
Given the fact that I'm pushing the weight limits of my mount, and I aren't using a guidescope, I will be limited to short exposures (between 1 - 30 sec). Which camera will make the most of these limitations: the Zwo asi585mc/player one Uranus C or the Sony a6000? And why? I know the 585 has 2.9um pixels and the a6000 has 3.88um which would be more suited to my focal length, gather light faster and be more forgiving of tracking problems, but what other factors/specs should I consider?

I just want to insert my 2 pence...
Short exposure DSO imaging is a real headache mate...

Just in 2 hours you may end up with the amount of 5k or more subs (16-32mb each), later stacking will take ages and software may even freeze or crash...
(not even talking about selecting best images, - blinking is not the option... Any other solution will also take ages, so you will end up dropping all images into the stack blindly).

Unless you have lots of patience and good hardware to process all of that.

You can read my experiences of a similar imaging below (process was done on MSI gaming laptop, with i7 9gen CPU, 16RAM and GPU GeForece 1050).

My suggestion, - go for the second hand cam which you will be able to use for guiding later, play with it on some galaxies and planets and if you decide to stay in hoby, later get the better mount or smaller scope and move from there.

 

 

Edited October 5, 2022 by RolandKol