The "No EQ" DSO Challenge!

[The Admiral]

37 minutes ago, rotatux said:

Hello, I'm new to SGL but having patiently read this thread from start to end over several weeks I decided to sign in and post some of my results to encourage others (if still needed :-) ).

FWIW I'm the proud owner of a SW130PDS so I've also read the entire "imaging with the 130pds" thread (over several months).

(sorry if I'm a bit long for this first post)

My first images were about 3 years ago on Nexstar SLT mount. Here the kind of thing I got:

Infos: Capture = 3 lights x 20s x 800iso, 2 darks, Olympus E-PM1 with Celestron MAK127 on Celestron SLT mount, dydimium filter; Processing = ImageJ + Gimp

Then I mostly turned to wide-field with old lens on self-made barn-door-tracker, had some nice images but always from only a few subs, due to my software pipeline of that time.

Then 1.5 year ago I bought a RA-motorized EQ3/4-class mount, in the hope to do unguided longer subs. I found it's good enough but is cumbersome to take to my pet dark spot (16kg+optics to take 50km and 45mn from home). I did some imaging with it and it's average, allowing good 30s subs but 60s is harder and depends on good polar alignment (which is too long for me and wastes rare good nights time).

So after reading this thread I thought "why not give my SLT a try again". In the meantime I have discovered Regim and it has changed my astro-imaging life ! Though it's not good at everything, it now allows me to properly align and stack many more subs. So stacking a quite high number of subs from an Alt-Az mount doesn't frighten my any more.

Here's one I did 2 months ago:

Infos: Capture = 103 lights x 8s x 1600iso, 21 darks, Olympus E-PM1 with Skywatcher 130PDS and CC on Celestron SLT mount, TS contrast filter; Processing = Regim, Fotoxx, Gimp

I am quite pleased with it since it's from Paris suburbs with huge light pollution, as this shows:

This shows possibilities are limited in such conditions without a good LP filter... so I had another try a few days ago on holidays from deep country and quite very dark spot:

Info: Capture = 86 good of 101 lights x 15s x 2500iso and 48 NG darks + 11 x 10s x 1250iso and scaled master dark, Olympus E-PM1 with Skywatcher 130PDS and CC on Celestron SLT mount, TS contrast filter; Processing = Regim, Fotoxx

My current masterpiece

Hi rotatux, welcome to the forum and SGL. Nothing wrong with that image, and you should be pleased with that. It's good to see that you've resolved the Trapezium too. Looking forward to seeing more of your work on the forum.

Ian

[Nigel G]

Hi rotatux, welcome to The No eq dso challenge  and SGL. 

As Ken said, your first image very much like my first image too. 

I look forward to seeing more of your DSOs 

Nige.

 

[rotatux]

« Welcome to the dark side »

Thanks ken.

« So keep pushing that mount harder! »

Err, for now my mount is quite reluctant to this :-/ : trying 30s gave a reject rate of about 85%, so I must be careful. I am working on M45 image where I pushed to 20s with only 26% rejects, so I try improving until I can find the root cause for this mis-behaviour.

« You also seem to be managing your light pollution very well. You will be surprised how well software can remove light pollution gradients »

Indeed I tried the background removal of Regim on the 2nd image, it's powerful but I find it not very precise, and I dont' like the cushion-like bumps it leaves on the background.

So I've opted for another technique...

« Have you considered using DeepSkyStacker to stack your images? »

Yes but I forgot to say I'm exclusively on Linux so my options are more limited than yours (and don't speak of Wine plz )

On the negative side, Regim only outputs 16-bit and no more, though it stacks on 32;
And is probably slower than DSS.

On some positive side, Regim has B-V color calibration, which gives me hassle-free color-correct images, while I read many here seem to struggle with the color module of ST.

« Nothing wrong with that image, and you should be pleased with that. »

Thanks Ian

« good to see that you've resolved the Trapezium too »

yes it is burnt in the 15s subs so I tried my first HDR with just a few lower exposure subs and it worked just like I expected.

« Hi rotatux, welcome to The No eq dso challenge  and SGL »

Thanks Nige

« look forward to seeing more of your DSOs »

will post a few more in coming days, I don't want to crush the thread with my messages :-P

Edit: Quote seems to not work at all for me so I opted for this manually quoted form. Also I need to find where to set a signature... (not found in profile)

Edited January 6, 2017 by rotatux

[happy-kat]

Perhaps you have balance issue that is effecting your exposure length sucess or maybe it was a bit windy.

[Filroden]

14 minutes ago, rotatux said:

Err, for now my mount is quite reluctant to this :-/ : trying 30s gave a reject rate of about 85%, so I must be careful. I am working on M45 image where I pushed to 20s with only 26% rejects, so I try improving until I can find the root cause for this mis-behaviour.

M45 is almost due south in the early evening so has the worst rotation. I normally avoid targets close to the meridian or reduce my exposures to 15-30s depending on altitude.

[rotatux]

« Perhaps you have balance issue that is effecting your exposure length sucess or maybe it was a bit windy. »

That's what I suspect. (the balance, not the wind I had zero)
I attached a 500g weight to the front to test, assuming the OTA would be pulled by the mount gears (rather than the reverse) when tracking east.
That's how my last M42 and my next M45 were taken, then 20s didn't work on M42 at ~35-40° alt (<33% good) but was fine 2 nights later on M45 at ~50-55° alt.

So I'm quite confused currently about this. Maybe I will try balancing the reverse way (put the weight at the back) and see what comes.

[Nigel G]

3 hours ago, rotatux said:

 

Edit: Quote seems to not work at all for me so I opted for this manually quoted form. Also I need to find where to set a signature... (not found in profile)

If you click on your own name, top right  of home page, next to your user picture, a drop down list shows settings,  click that then on left will be signature setting. 

Nige.

[The Admiral]

OK, after a few more hours of messing about, I cannot fathom what deconvolution is doing in ST. ST seems to mask stars with a rounded square and as the centre of M42 is 'busy' there is a collection of star squares there. That means that part of the nebula is affected by decon, and an adjacent part not, and this shows up as a clear demarcation. I tried painting over the mask in this region to get decon to work on it, but it wasn't successful. I gave up on decon and continued as before (which is a bit of a shame really as decon did bring out detail in the nebula, where it worked!), but hopefully better refining the image compared to last time. So I've finally got to the image below which I think is an improvement on the previous one; it's hard to be objective after sitting in front of the PC for hours . Anyway, that's it for a while, time to give the bum a rest! I've got colour (aka gaudy) versions still to blend, but that will keep. I think I'm just about satisfied with this one (law of diminishing returns here).

Ian

Edited January 7, 2017 by The Admiral
Re-cropped image

[Filroden]

54 minutes ago, The Admiral said:

I cannot fathom what deconvolution is doing in ST. ST seems to mask stars with a rounded square

In Pixinsight, I create a star mask to protect the stars from getting rings and I also mask the background to prevent it being sharpened. The mask is usually always a little bigger than the star. I also have the option of calculating the point spread of stars by selecting them (it was recommended to select about 50-60 stars to get a good average). Alternatively you can just use a default star shape. As I got similar results from both, I chose the easier option! I guess the star masks/deconcolution in ST are doing the same thing. It does sharpen and reduce the stars, making them much nicer shapes. I haven't noticed much improvement in other areas and I sometimes skip deconvolution entirely if I'm already happy with star shape.

[Nigel G]

I think Decon needs at least 50% binning to have best results.

Auto mask then, radius 2.0 pixel, mask fuzz 14.5 seems to work well for my images most of the time.

Nige.

[imballinger]

Hey guys Great thread and some amazing pictures!

I have a Celestron nexstar evoulution 8 with it's original mount, any suggestions on a camera to get?

[Filroden]

Hi and welcome to the thread. I hope you'll post your results here once you're into imaging

In reply to your question (and the question you've posted elsewhere as they are all related): what do you need to get into imaging?

Mount

You have a good altaz mount. Mine will keep my target centred for many hours once aligned. This is great for visual use but not actually all that important for imaging (even though it's helpful) because of other limitations linked to the altaz nature of the mount which mean you will be restricted to individual exposures of under a minute or two. Altaz mounts track differently to EQ mounts and suffer an effect called field rotation. The mounts hold the camera in the same absolute orientation related to ground, so as the stars move across the sky, they will rotate within the image. The easiest way to understand this is to look at Orion when it rises in the early evening and check it again a few hours later - he starts leaning to the left in the evening, he is upright when he passes south and he leans to the right when he sets. The same thing happens in your images. This can be worked around by taking shorter images so there is not enough time for the rotation to be seen. However, you still see rotation between each image and when you integrate the images you will see this. We have to crop our images to remove these stacking artifacts around the borders. I can lose up to 30% of the image if I have been taking images over a very extended period.

The goto accuracy of your mount is also very good and it's very easy to align (even easier with the StarSense accessary but that's just a time saver - you can get as accurate an alignment using the two or three star methods). I found the wifi useful and used to control the mount from my iPad but it can be temperamental. I think my issues with wifi were that my iPad would try reconnecting to my stronger home wifi over the scope's weaker wifi signal. Once it drops like this, I found it impossible to reconnect and had to get the handontroller out and do a new alignment (you cannot align with one and use the other - which even method you use for alignment needs to be the method you use to send goto commands).

Scope

Again, a great scope and ideal for visual. It's also very good for imaging lunar and planetary targets, especially when combined with a barlow, because they are so bright and you only need exposures of a second or less. It has a very long focal length at just over 2 metres. As a reflector, it doesn't suffer chromatic aberation (where light of different colour is focused at slightly different points) so you will get good colour once you reach sharp focus. You'll probably need a dew shield as a minimum. The corrector plate will mist up quite quickly without one. I use a shield but no heater.

Now about that focal length...this is where you'll run into problems with imaging. The first is the field of view (which will depend on your choice of camera). With a standard DSLR you won't be able to fit much of the biggest DSO's into the field of view, e.g. M31, M42 or M45. You could just about fit M33 into the field but with field rotation you would lose much of its arms when you had to crop. Realistically, at 2m focal length, you'll be targetting DSOs like M1 or M51. You could add a focal reducer. The most common would reduce the focal length by 1/3rd but I don't think that would bring any of the big DSOs within range.

So you will need to decide what type of targets you want to image. The big (and bright) targets would probably be better suited to scopes with focal lengths of 800mm or less. My refractor has a 400mm focal length and gives me nice wide fields with both my cameras. My 9.25" evolution is reserved for planetary targets and small galaxies.

The second is that longer focal lengths need very accurate tracking. If not, you are limited to shorter and shorter exposures to avoid star trails. I found I could only get a maximum of 15s exposures with the SCT (2350mm focal length) and more often would actually only get 5-10s. My refractor, on the same mount, lets me take exposures of 30-120s.

So with your scope, you'll be probably hunting fainter targets which would prefer longer exposures but you'll actually have to take shorter exposures. Even this can be worked around with the right camera or filters or a good, dark location.

Work arounds

You'll here this a lot on this thread. We know we're not imaging the best possible way. But we're imaging with the tools we already have, can afford or choose for other reasons. We don't expect the same results but we like to get the best results we can with what we have.

If you were just starting out with no equipment, you will get different recommendations that if you are trying to use equipment you already have. Never forget the advice here is always a compromise between what would be better/best and what you could more easily achieve so long as you accept the limitations.

Camera

This is the other half of the equation for imaging and it's chip size also defines the field of view when combined with the scope's focal length. There is a really useful tool on the First Light Optics website for showing field of view with different DSO targets and combinations of scope/camera at http://astronomy.tools/calculators/field_of_view/

If you already own a DSLR I would recommend trying with this first. All you need is an adaptor that fits directly onto the DSLR body and ends with a nose that fits into the scope in exactly the same way as an eyepiece. The better way of attaching the camera is a fully threaded method but this requires a few extra adaptors (though the cost difference isn't that large). Using an existing camera will give you the chance to see if you're happy with the images you can get from the mount/scope. DSLR's have some of the widest fields of view so any camera upgrade is likely to be similar or have a smaller field of view.

I started with a Canon EOS60d. I had one for daytime photography so I wanted to use it. Eventually its downsides put me off enough to upgrade. My camera had a lot of noise. You cannot cool them without a lot of modification so the chip ran at over 20C which causes a lot of noise. You cannot easily use darks to calibrate because you cannot match temperature. I also didn't like the sensitivity of the camera. one-shot colour cameras have to divide the chip up to particular colours so you're not using the full power of the camera. Half of the chip detects green light (great in the daytime, not so great for us) and only a quarter is dedicated to reds and to blues (much more interesting for us). Also, one of the main sources of red, Hydrogen Alpha, falls in one of the least sensitive parts of the chip. I think my camera could only capture about a quarter of the actual Ha data because of its efficiency - so I was only getting a quarter of the data on a quarter of the chip! Again, you can modify the camera to improve its red sensitivity.

I went for an upgrade. I went to the ZWO ASI1600MM-C. It uses the same chip technology as a DSLR (CMOS) but in a astro-dedicated package. It's a mono camera so I use it's full capacity. It's more sensitive to reds. It's cooled so much less noisy. And it has a very low read noise. So shorter exposures work well with it. It does however only have a bit depth of 12 (so it can only measure light strength on a scale up to 4k). My DSLR had a 14bit depth (so four times bigger) and most CCDs have 16bits (so measure light on a scale up to 64k). Bit depth is important as it limits the amount of contrast you can achieve. You can overcome this by taking lots and lots of exposures. If I take 4 exposures, I can get back to the bit depth of a single exposure from my DSLR. If I take 16 exposures, I can get back to the bit depth of a single exposure from a CCD.

The ZWO camera came with everything I needed to attach it to my scope other than spacers to get the back focus spot on (something I still haven't fixed yet).

Filters

Light pollution: if you live in an area with lots of light pollution from sodium and mercury lights (bright orange glow) then light pollution filters will help you take longer exposures without overexposing because of the background light. However, these don't work for LEDs because they shine with broad spectrum white light (boo).

Broadband colour filters: if you go the mono route, you need RGB (and probably L) filters so you can combine them to get colour images. I use LRGB filters in an electronic filter wheel so that I can automatically swap filters during imaging. My L filter is just an IR/UV filter to cut off non-visible light (which stops star bloat). You can use a light pollution filter as an L filter too.

Narrowband filters: you can get filters which are limited to very small ranges of light that match up with the main sources of light in emission nebula such as H alpha, SII and OIII. You can also get them for N, H beta, but these are less common. They come in various "widths" which shows how tight they are. The narrower the filter the more expensive it will be. Because they only pass through a limited range of light they usually need longer exposures (or very bright targets). Again, my camera has some nice capabilities that mean you can still get reasonable results with exposures in the 30-120s region - in range for my refractor but no good for my SCT.

Capture method

With a DSLR you can just attach it and click away (remembering to allow pause time for the vibrations to settle). No other equipment required. Or you could hook you kit up to a laptop and control it from there. You'll also want to consider processing software both for integrating the images and then developing them.

Second health warning

Your mileage might differ to mine  I've found solutions that I like working with and give me results I'm happy with. I've thrown a lot of money at making it easy rather than making it good (though it has made it much much better). With the same budget I could have got an EQ mount and got better results but at the cost of ease.

As I said before, give it a try first before investing a lot. Use (or borrow) and existing DSLR and give it a test. See if you like/enjoy the results and then see what you want to achieve. The more you know what you want to acheive (choice of targets, field of view, etc) the easier it is to give more specific advice.

Edited January 8, 2017 by Filroden

[Nigel G]

Hi imballinger welcome to the thread.

Ken said it all

look forward to seeing some of your images.

[The Admiral]

4 hours ago, imballinger said:

Hey guys Great thread and some amazing pictures!

I have a Celestron nexstar evoulution 8 with it's original mount, any suggestions on a camera to get?

Hi and welcome to the thread. Yes, I think Ken has summed it up very succinctly. I would certainly commend you to try astrophotography with what you already have, as it is certainly possible. It can be a long journey and sometimes a frustrating one, so it's best to see how you get on with it before splashing the cash on all the fancy gizmos oft recommended.

Don't get the impression that all DSLRs are inherently bad at capturing the deep reds of H-alpha, some are better than others in this regard. That response is not a priority for manufacturers so it's not a given. And DSLRs can be modified if that becomes a priority, though as Ken says, when you need to start spending serious cash it's best to be clear on what you plan on imaging.

Looking forward to seeing the results of your endeavours, and feel free to ask any questions along the way.

Ian

[The Admiral]

16 hours ago, Nigel G said:

I think Decon needs at least 50% binning to have best results.

Auto mask then, radius 2.0 pixel, mask fuzz 14.5 seems to work well for my images most of the time.

Nige.

Did you get that from trial and error Nige?

Ian

Edited January 8, 2017 by The Admiral

[happy-kat]

AltAz mounts move in tiny left right up down movements hence they keep the object in the field of view but the object is rotating. This is the exposure length limitation.

[Nigel G]

2 hours ago, The Admiral said:

Did you get that from trial and error Nige?

Ian

Ian, trial and error with a little guidance from Ivo down under.

Nige.

[Nigel G]

6 hours ago, imballinger said:

Hey guys Great thread and some amazing pictures!

I have a Celestron nexstar evoulution 8 with it's original mount, any suggestions on a camera to get?

I have a self modified Canon 1200d an unmodified 1300d and an Orion star shoot.

The 1300d is pretty good, the modified is great with Ha nebula but won't improve Galaxy DSO's. The difference between the 2 is around 40-50 % improvement with red's Hydrogen alpha, well worth the effort.

The Orion star shoot is great for planetary and solar system photography, ok for DSO's (but DSLR better) and can also be used for guiding if I ever change to Eq mounts.

A lot will come down to what you want to image and how much you want to spend. New cameras between £300 and £ 5000. Second hand modded from £ 250 ish

Nige.

[The Admiral]

12 minutes ago, Nigel G said:

I have ..... an unmodified 1300d.....

Nige.

Now if I remember correctly, that's your wife's camera

Ian

[Nigel G]

11 minutes ago, The Admiral said:

Now if I remember correctly, that's your wife's camera

Ian

That's correct Ian, She asked if she could use it when we were away in December. LOL. I had to remind her who's it is but she prefers here Samsung A5 quick easy and in her pocket 

So it conveniently fits in my new camera case with all the lenses and other camera keeping it safe from damage. Now that's a result. ( predictable result )

Nige.

Edited January 8, 2017 by Nigel G

[Filroden]

One step forward, seventeen steps back...

I now remember why I never liked trying to image with the 9.25" SCT. I bought myself a new finder base for the SCT so I could easily swap the finder scope and StarSense between my two scopes. Tonight was supposed to be cloudy but it had been crystal clear all day and as the sun set it was still clear. I thought I'd give the SCT a whirl as I'd not used it since buying the refractor.

I got set up and did my first StarSense alignment. There was already a few clouds starting to form. After aligning I did a goto to Capella to focus and see how good the uncalibrated alignment was (it was currently calibrated for the refractor). No bright star in the field of view but I could at least see three or four nice big doughnuts of out of focus stars. Previously I think I would be lucky to see even that it was often so far from focus so it was pretty easy to a. figure which direction to focus and b. bring it to a close focus. Then the fun began and I had my first reminder of why a 2350mm focal length scope is a beast. Any slight movement of the focuser and the whole image would wobble, taking a good few seconds to settle. By which time a small gust of wind would have set it moving again and another wait, by which time a kid playing in the street would have made the ground tremor enough to set it wobbling again.

I thought I'd try the Bahtinov mask. Not a chance. So went back to minimising HFR and patiently got it to a score I was somewhat happy with. Now to find Capella. I noted the position of a star in the centre of the field of view then manually slewed it to each corner in the hope Capella would appear. Not a chance. So I did a plate solve and finally found I was about three fields of view away. It didn't take me long to find it. And boy was it huge. I thought I'd hit the moon or something. So I calibrated on Capella and did a fresh alignment.

A quick goto to M1 later and I found the stars I was looking for to know I had M1 close to the frame. I was running continuous 1s exposures at this point so no sign of M1 even with a light stretch. Checked outside. That nice grey even background I was seeing between the stars...more clouds...but some gaps, some of them big.

So I upped the exposure to 5s and just watched each image. Slowly, as a cloud moved, I'd see a cloud in the centre of the field. Could it be? Yep, it looked the right shape to be M1. So I thought this was a goer. I tried to improve focus a little more. Think I'm going to need a new focuser if I am going to persevere with the SCT!!

I also tested a 10s exposure which seemed to work but not as well as 5s in terms of good frames, so I decided on 5s.

I set up a sequence of 12x5s each for LRGB in the hope of doing short runs I could get something between the clouds. I ran it twice by which time the clouds were totally obscuring the target and they were so dense on the horizon I packed up.

Of the 96 images most were affected by clouds. I decided not to even care and ran a star alignment on them using a good L frame as reference. Only 2 of the 24 B frames had enough stars to match. More of the G, L and R frames gave star matches and in total I had 42 images that could be aligned.

Now normally I'd really check each frame and ditch any that had clouds. But with only 42 frames I just didn't care! I threw the lot (LRB and G) into a single integration in the hope enough signal would climb above the clouds. There was no chance of integrating them separately though I did test with just the L frames and it was similar in result but just a little noisier. So all I had was the certainty that I could align my SCT still, I could just about focus it, and it would just about take 5-10s images.

Here's a very poor mono M1 with just over 3 minutes of LRGB data all combined. Oh, and did I mention the Moon was just out of frame?

[SteveNickolls]

Thanks Ken for showing what perseverance can bring, or is it a warning not to repeat as it is so difficult? I have an old 9.25" stored away and often wondered if it will ever see the light of day. Is your 'scope able to take a Hyperstar?

Cheers,
Steve

[Filroden]

4 hours ago, SteveNickolls said:

Thanks Ken for showing what perseverance can bring, or is it a warning not to repeat as it is so difficult? I have an old 9.25" stored away and often wondered if it will ever see the light of day. Is your 'scope able to take a Hyperstar?

I think it was a little of both. To be fair, it was the clouds which caused the main imaging issue but it's certainly a much harder scope to work with than the refractor. 

I think it is compatible with Hyperstar but I think when I looked into it, it was not much more to buy the fast refractor. This gave me a 400mm scope compared to the 540mm that Hyperstar would give me. Given the difficulties I have focusing, I don't know how Hyperstar would work. I can at least add a new focused to the visual back that will be better than moving the primary mirror and could be automated. I do t know how Hyperstar could be focused other than moving the primary. 

Its supposed to be clear tonight but very windy. I'd still like to have ago at M1 just to see if it's worth doing other targets. I reckon I need at least 240x5s subs for each of LRGB to even start getting something reasonable. Ideally I would test 10s and 15s subs but if their fail rate is too high I'm safer staying with short subs and increasing the camera gain.

[The Admiral]

Forecast here is for clearing late tonight, but 'when?' is a matter of conjecture as the weather sites differ. But the Moon is getting on for full, in my preferred direction (and near to M1, Ken), so I don't think I'll be trying.

Ian

Edited January 9, 2017 by The Admiral

[Filroden]

6 minutes ago, The Admiral said:

and near to M1

I'd just noticed that. I'd probably still have given something a go but the wind will make it impossible.