Long focal length for deep sky astrophotography

[Tom33]

I have been imaging for over three years now, always with refractors and at the moment I have a roll off roof observatory, an EQ6-RPRO, a SW Esprit 100, an ASI294MC PRo and control it all with an ASI Air pro. I will move to NINA eventually but at the moment it just works! I would love a scope with a long focal length to go after some galaxies and the smaller deep sky objects but would appreciate some advice. I’ve looked at Newtonians, Cassegrain,s in all their formats but appreciate that collimation and other matters will come into play. If anyone has gone down this route I’d be pleased to hear from you.

Many thanks & clear skies

[SamAndrew]

I've had decent galaxy results from the Esprit 100; you don't really need that much focal length to get to a good sampling rate for galaxies: 1~1.5 arc seconds

I've also had imaging newtonians; entry level ones are a lot cheaper than RC scopes, you can't beat them on bang for your buck. 200pds or Quattro 8S can produce stunning results for the price.

The main thing RCs have going for them is the more compact design. Any entry level one will need a focuser upgrade.

[Tom33]

Thanks for the advice, much appreciated.

[Clarkey]

Despite really liking the RC8, I must admit that the images from the 115mm triplet are probably as good, if not better. Obviously the RC is binned and marginally quicker, but above about 800mm FL seeing becomes the predominant factor. I am currently considering a 250mm F4 Newtonian to give speed and FL (possibly even a Nexus to give F3).

My preference overall would be for a large refractor, but unfortunately my current mini observatory won't cope with the length of anything over the 115mm scope I now use. I did buy a 130mm triplet - but it was about 2" too long😭

[tomato]

Olly Penrice did an article a few years ago comparing the performance of large aperture moderate FL refractors with large aperture long FL reflectors for galaxy imaging, there wasn't a lot in it and the low maintenance advantage of the refractor won the day for me and I settled on an Esprit 150. As @Clarkey has stated, seeing is often the limiting factor, I would have made a different choice if my scope was going on top of a mountain in Atacama. 

[Sp@ce_d]

Yeah for minimum fuss a frac every time for me. I spent a few years with the SCT & RC but my goal was an fully automated Obsy and unless you enjoy constantly tinkering .. which I did at first but it soon wore off.. stick with a frac. Once I had the Esprit 150 I wondered why I had wasted so much imaging time. Of course a frac will come with the weight and length requirements.. (oh and cost!) but it’s pretty much set and forget.

[gorann]

A very popular one with a longer focal length that I cannot remember having read any complaints about is Edge HD8. It is easy to collimate (look at a de-focused star and play with three screws on the secondary mirror) and I expect it to hold collimation well, especially in an obsy.

[900SL]

I quite like the idea of the Vixen R200ss with the PH extender, giving f5.4 and 1150mm fl

[osbourne one-nil]

30 minutes ago, 900SL said:

I quite like the idea of the Vixen R200ss with the PH extender, giving f5.4 and 1150mm fl

I'm quite happy with the results I get from that setup. 

 

 

 

Edited February 19 by osbourne one-nil

[ONIKKINEN]

Im imaging with an 8" f/4.4 newtonian with a paracorr, bringing the effective focal length to 1018mm at roughly f/5. With my IMX571 and its 3.76 micron pixels i get 0.76"/px, which i have not been able to make use of fully even under very good seeing.

So for me its a bit too much resolution, but honestly im getting pretty close with BlurXterminator if the seeing was good so not a complete daydream to work at that resolution.

I think a 1 meter focal length newtonian would work very nicely as a galaxy scope with your camera. Need to tend to its needs though, particularly in the stability, collimation, and coma correction department.

[Bikingbill]

This is something I'm looking into as well. I have a heq5 so looking at lighter options. I have been considering the Stellamira 125 which seems a low maintenance option compared with a RC or newtonian. 

Or I could update my mount....

Lost in indecision 😁

[ollypenrice]

False click. Sorry.

Edited February 19 by ollypenrice

[vlaiv]

I wonder why people insist on "speed" of the telescope as being crucial thing when imaging galaxies?

Most galaxies are very small in angular size - maybe dozen of arc minutes at most. That is about 700px or less across the image of galaxy if one samples at highest practical sampling rates for amateur setups - which is 1"/px.

Now take any modern sensor that has more than 12MP - that is 4000x3000px or more, so you have 4000px/700px = ~x5 at least x5 larger sensor than you actually need in terms of pixels. You can bin x5 and you'll be still able to capture galaxy in its entirety + some surrounding space.

Btw - bin x5 will make F/15 scope work as if it was F/3 scope - so what is the point in going for F/4 Newtonian scope when you can comfortably use compact Cass type - be that SCT, MCT, RC or CC and produce excellent galaxy image.

 

My take on this would be - get largest aperture that you can afford and comfortably mount and use and adjust your working resolution to range of 1-1.2"/px for best small galaxy captures.

[ollypenrice]

1 minute ago, vlaiv said:

I wonder why people insist on "speed" of the telescope as being crucial thing when imaging galaxies?

Most galaxies are very small in angular size - maybe dozen of arc minutes at most. That is about 700px or less across the image of galaxy if one samples at highest practical sampling rates for amateur setups - which is 1"/px.

Now take any modern sensor that has more than 12MP - that is 4000x3000px or more, so you have 4000px/700px = ~x5 at least x5 larger sensor than you actually need in terms of pixels. You can bin x5 and you'll be still able to capture galaxy in its entirety + some surrounding space.

Btw - bin x5 will make F/15 scope work as if it was F/3 scope - so what is the point in going for F/4 Newtonian scope when you can comfortably use compact Cass type - be that SCT, MCT, RC or CC and produce excellent galaxy image.

 

My take on this would be - get largest aperture that you can afford and comfortably mount and use and adjust your working resolution to range of 1-1.2"/px for best small galaxy captures.

I take this point but have little experience of binning with CMOS cameras. What's your view on how this should be done and how effective it is?

Olly

[Guest]

On 17/02/2024 at 14:08, Clarkey said:

Despite really liking the RC8, I must admit that the images from the 115mm triplet are probably as good, if not better. Obviously the RC is binned and marginally quicker, but above about 800mm FL seeing becomes the predominant factor. I am currently considering a 250mm F4 Newtonian to give speed and FL (possibly even a Nexus to give F3).

My preference overall would be for a large refractor, but unfortunately my current mini observatory won't cope with the length of anything over the 115mm scope I now use. I did buy a 130mm triplet - but it was about 2" too long😭

I will second this, I have the at115edt and compared it with my rc6 at 1370mm on M33 and the at115 actually out resolves it.  I get pretty decent seeing conditions here in SE Arizona since I'm at fairly high altitude, 4200ft (about 1300m).  John Hayes who I very much respect as an astronomer, had his setup near Tucson and observed similar things, even when using pretty large aperture long fl scopes.  Eventually he sent his kit to Chile to get full benefit from his scope.

My take is imaging anything at over 1500mm won't gain you anything.  Oversampling can help a little.  The best case forward for large aperture scopes is you can reduce them so you can end up with relative light buckets.  Refractors tend to be slow.  Something like a sharpstar sca360 would be ideal although not cheap, or a fast newt.  Edgehd scopes with reducers are also a good option especially since they can be converted to low fl light buckets via hyperstar.  That's kinda the path I'm looking at myself down the road.

 

[vlaiv]

3 minutes ago, ollypenrice said:

how effective it is?

100% effective

Only difference between software and hardware binning is level of read noise. With CCDs and hardware binning - you have the same read noise regardless if you bin or not, but with CMOS sensors "effective" read noise is increased by bin factor.

If you bin x2 - read noise is x2 larger, if you bin x3 - read noise is x3 larger and so on.

However - this does not make any difference on final result if you already expose to swamp the read noise at bin x1. When you bin and increase read noise - you also increase other noise source in the same manner so their ratio - or "by how much you swamp" the read noise with say sky noise - remains the same.

Best way to bin is to actually not bin at all - best way is to split your subs so that different pixels end up in different sub subs. This way you avoid any mathematical operations with pixels - you reduce sampling rate (because you leave every other or third pixel) but you end up with multiples of subs - as you've imaged for longer. This shows that there is really nothing is lost - it is pure trade off between sampling rate and SNR,

Just to clarify what I'm saying - you take one sub and you split it into 4 smaller subs - first containing odd, odd pixels (in x and y), second odd, even, third even, odd and fourth even, even (a bit like bayer matrix splitting). In both axis - in X and Y you have twice as few pixels so each new sub is half the height and half the width - sampled at twice smaller sampling rate - but you have x4 more sub to stack - which improves total SNR x2.

Ok, but no software in reality implements (that I know of) above approach, although I'm sure that PI script can be written, so next best thing is to simply bin each sub after calibration and before stacking - either average or sum will do, but take care to save each sub in 32bit floating point format to avoid loosing precision (you should do this anyway when calibrating).

Third option is to simply take your stack and bin it before you start to process it.

In principle - above three are equivalent bar some minute differences that have to do with interpolation when aligning the subs for stacking (mostly academic arguments - no practical difference).

 

 

[Guest]

12 minutes ago, vlaiv said:

I wonder why people insist on "speed" of the telescope as being crucial thing when imaging galaxies?

Most galaxies are very small in angular size - maybe dozen of arc minutes at most. That is about 700px or less across the image of galaxy if one samples at highest practical sampling rates for amateur setups - which is 1"/px.

Now take any modern sensor that has more than 12MP - that is 4000x3000px or more, so you have 4000px/700px = ~x5 at least x5 larger sensor than you actually need in terms of pixels. You can bin x5 and you'll be still able to capture galaxy in its entirety + some surrounding space.

Btw - bin x5 will make F/15 scope work as if it was F/3 scope - so what is the point in going for F/4 Newtonian scope when you can comfortably use compact Cass type - be that SCT, MCT, RC or CC and produce excellent galaxy image.

 

My take on this would be - get largest aperture that you can afford and comfortably mount and use and adjust your working resolution to range of 1-1.2"/px for best small galaxy captures.

I had the same thoughts until recently.  But, a lot of galaxies have very dim outer arms, dust other stuff like jet streams that require either a lot of integration or fast optics.  So a fast say 1300mm fl scopes is a great idea imo.

Cmos binning is not the same as cc'd binning as it's all software based.  It basically averages the pixel data or sums it up.  Problem with that, the device still reads the pixels individually, so each pixel readout still has its own read noise.  With ccds, all the binned pixels are read at once with only one read noise involved.  So binning CMOS only really saves you disk space, because you can do it any time later after data acquisition instead of in camera.

There is some advantage in oversampling beyond the Nyquist limit which is why I don't recommend binning normally.

[vlaiv]

Just now, dciobota said:

So binning CMOS only really saves you disk space, because you can do it any time later after data acquisition instead of in camera.

Nope.

It improves SNR by bin factor from recorded image - regardless of how read noise is treated. Once you have image - no matter how it was acquired - was it CMOS or CCD - software binning will improve its SNR by bin factor (if you bin x2 - you will get x2 improvement, x3 - x3 SNR improvement and so on).

Binning is the same underlying procedure as stacking - which is in turn the same underlying procedure as longer integration.

You effectively trade spatial resolution for integration time when you bin.

 

[vlaiv]

Just for anyone doubting software binning - I'll give an example.

Can you capture M51 tidal tail with 1 minute exposure with 8" telescope from Bortle 8 light pollution?

Most will agree that it is impossible, right?

Ok, so this is single 1 minute sub at native resolution:

It looks like one might expect, right?

Look what the same sub looks like - if I bin it quite a bit:

That is one sub - one minute, no stacking, no monkey business - enough SNR to show tidal trails!

However, I had to trade in large amount of resolution to get it - image of galaxy is now tiny

 

[ONIKKINEN]

46 minutes ago, vlaiv said:

However - this does not make any difference on final result if you already expose to swamp the read noise at bin x1.

I think you kind of skimmed over this point as if its nothing, but lets be serious here this is a major issue if we are aiming to use an f/15 scope.

Say we are imaging under decent skies with an f/5 scope and we swamp read noise x5 with a 2 minute sub, well now with the f/15 scope that turns into an 18 minute sub. How many off the shelf mounts can do that reliably even if we know how to make the most out of our guiding gear?

I agree with you on binning making up for speed and do that with mosaiced images every chance i get, but taking the focal ratio to an extreme is not at all as easy as just using a "faster" scope.

[vlaiv]

3 minutes ago, ONIKKINEN said:

I think you kind of skimmed over this point as if its nothing, but lets be serious here this is a major issue if we are aiming to use an f/15 scope.

Say we are imaging under decent skies with an f/5 scope and we swamp read noise x5 with a 2 minute sub, well now with the f/15 scope that turns into an 18 minute sub. How many off the shelf mounts can do that reliably even if we know how to make the most out of our guiding gear?

I agree with you on binning making up for speed and do that with mosaiced images every chance i get, but taking the focal ratio to an extreme is not at all as easy as just using a "faster" scope.

I indeed skimmed over that because there is no issue there really.

Let me explain by using two points.

First is - read noise in terms of CCD vs CMOS.

CCDs used to have very large read noise - like 7-8e and sometimes even more (very few models had read noise as low as 5-6e).

Modern CMOS sensors have read noise in 1-2e range. That is at least x4 less then CCD sensor - so one would need to expose for x16 loner with CCD to reach the same level of "overwhelm" with sky noise. Indeed, back in the day, exposures of 20 or more minutes very fairly common (even half an hour or longer for NB imaging).

Now onto mounts and guiding. Most mounts have periodic error that is order of up to 10 minutes or there about. That is full period, and half period - where mount takes to go from peak to peak is half that. We could argue that "road" from peak to peak is either a) smooth - making RA drift same for first two and a half minutes as for second two and a half minutes - then if you can image/guide for 2.5 minutes - you should be able to image whole 5 minutes without issues and by extension whole worm cycle as it is the same road in other direction or b) one of two parts is significantly steeper - so it can't be guided - then you would loose every other sub to not being able to guide.  If that is not the case - and you don't loose subs - then you should be able to guide whole RA period - and if you can guide whole RA period - what stops you from guiding 2 consecutive periods?

In any case - I don't think that sub duration is very important issue. If one can't guide for 10-15 minutes, one should sort out that bit first before attempting to do close up galaxies.

[ONIKKINEN]

41 minutes ago, vlaiv said:

I indeed skimmed over that because there is no issue there really.

Let me explain by using two points.

First is - read noise in terms of CCD vs CMOS.

CCDs used to have very large read noise - like 7-8e and sometimes even more (very few models had read noise as low as 5-6e).

Modern CMOS sensors have read noise in 1-2e range. That is at least x4 less then CCD sensor - so one would need to expose for x16 loner with CCD to reach the same level of "overwhelm" with sky noise. Indeed, back in the day, exposures of 20 or more minutes very fairly common (even half an hour or longer for NB imaging).

Now onto mounts and guiding. Most mounts have periodic error that is order of up to 10 minutes or there about. That is full period, and half period - where mount takes to go from peak to peak is half that. We could argue that "road" from peak to peak is either a) smooth - making RA drift same for first two and a half minutes as for second two and a half minutes - then if you can image/guide for 2.5 minutes - you should be able to image whole 5 minutes without issues and by extension whole worm cycle as it is the same road in other direction or b) one of two parts is significantly steeper - so it can't be guided - then you would loose every other sub to not being able to guide.  If that is not the case - and you don't loose subs - then you should be able to guide whole RA period - and if you can guide whole RA period - what stops you from guiding 2 consecutive periods?

In any case - I don't think that sub duration is very important issue. If one can't guide for 10-15 minutes, one should sort out that bit first before attempting to do close up galaxies.

I was thinking more about things we cant help, like a gust of wind which will now scrap 18 minutes of data instead of 2.

Agree with your reasoning with the other things though, there is no difference in a 3 minute or 30 minute guide duration if we dont have external disturbances. Getting quickly off topic now though.

[vlaiv]

1 minute ago, ONIKKINEN said:

I was thinking more about things we cant help, like a gust of wind which will now scrap 18 minutes of data instead of 2.

Valid point - but you can always choose to let read noise have a bit larger impact by shortening individual subs.

You might choose swamp factor of 3 over 5 or something like that and simply go with shorter subs if wind gusts are real concern. Alternatively - everyone likes lower read noise camera, so maybe they will keep reducing the read noise further

[osbourne one-nil]

I have no self respect to lose, so let me ask what might be a stupid question. 

If I use my 8" Newt and its dedicated coma corrector, with my ASI2600 camera, it gives me a resolution of 1.02" per pixel. If I was to use 100mm refractor with a similar focal length, and therefore a similar pixel scale, would the bigger aperture of the Newtonian give me any benefit?

My brain is struggling. 

[vlaiv]

28 minutes ago, osbourne one-nil said:

I have no self respect to lose, so let me ask what might be a stupid question. 

If I use my 8" Newt and its dedicated coma corrector, with my ASI2600 camera, it gives me a resolution of 1.02" per pixel. If I was to use 100mm refractor with a similar focal length, and therefore a similar pixel scale, would the bigger aperture of the Newtonian give me any benefit?

My brain is struggling. 

Yes - you would produce the same quality image (in terms of SNR) in 1/4 of the time with Newtonian as you would with 100mm refractor. This is because you have x4 more light gathering area with 200mm of aperture versus 100mm of aperture.

Second benefit is that given same sky, same mount, same conditions - 8" would produce very slightly sharper image than 4" - if both scopes are diffraction limited (which might not be the case if you use CC for newtonian or field flattener for refractor).