Help with Imaging the smaller DSOs

[DougM43]

Ok guys could do with some help.

unlike most I suspect my journey into imaging wasn't to produce super high quality stunning images, but more to see better what I was looking at when observing, the image I've attached is a perfect example of what I mean. I've looked for M51 many times in the past with my scope (LX90) and never had a glimpse of any of the spiral arms and having to use my imagination rather than my eyes to see it, same goes for the Crab Nebula. So to my problem, although this image ticks all the boxes of what I wanted to achieve when I bought the lmages kit I've now got greedy and want more, particularly as my first image of the Orion Nebula was way better than I expected. I appreciate to get a better image of M51 I need a lot more subs and way better processing but there's one thing that won't improve,, the size, it's tiny. When I zoom in to make it bigger the stars look like bubbles and completely unnatural.

So what's the best way forward for me to go to to image smaller DSOs like this, the crab the horse head and the ring nebula are also on my hit list. I think for the larger objects I've got the right kit but not the small ones. 

Ive been using an ed80pro on a heq5 with a modified canon 40d, maybe a different scope or a different camera? Not sure, any help much appreciated.

cheers for reading. Doug

[Tinker1947]

This is NGC 891 with a 10" Reflector and a 414EX Mono no cropping, single image of 20 minutes...less pixels should mean a bigger image on the chip....i sure somebody can put it better...

 

[DougM43]

Thanks that's helped a lot, I've just put the Atik 414ex into the FOV calculator with my scope and it seems with that camera and the one I already own I would have most bases covered. £1000 though :-(

doubt I'd get that under the wife's radar. 

[SteveNickolls]

Hi Doug,

I share your reasons for starting imaging to see in more detail what is difficult to observe. I'm quite new to imaging but can't your processing software clip your image to reduce the empty space around M51 and make M51 appear larger, if you have enough detail that is in your original? Here is my capture of M51 using a SkyWatcher Startravel 102mm refractor and Canon 600D DSLR. I'm also limited by my alt-az mount and local light pollution. Your EQ mount should allow you to get longer indivudual exposures and much more detail if your skies are dark. This is an image using x113 forty second light frames plus x50 dark frames and x50 bias frames and it shows detail and colours I had never observed before in my binoculars. I use DSS to stack frames and StarTools to do subsequent processing.

Best regards,
Steve

[StuartJPP]

4 hours ago, Tinker1947 said:

This is NGC 891 with a 10" Reflector and a 414EX Mono no cropping, single image of 20 minutes...less pixels should mean a bigger image on the chip....i sure somebody can put it better...

 

Don't let Olly hear you say that! It is the combination of the focal length and the size of the pixels, not how many pixels there are or the size of the sensor.

Those FOV calculators can be deceiving as it only shows the FOV covered by various combinations, not how detailed the target will be. The pixel size of the Atik 414ex is 6.45µm and the Canon 40D is 5.7µm, so M57 will actually be smaller on the Atik than it would be on the Canon 40D using the same OTA...

The Canon 40D will have a lot more space around M51, but cropping 1:1 will produce a larger M51 than the Atik 414ex...

 

[Tinker1947]

45 minutes ago, StuartJPP said:

Don't let Olly hear you say that! It is the combination of the focal length and the size of the pixels, not how many pixels there are or the size of the sensor.

Those FOV calculators can be deceiving as it only shows the FOV covered by various combinations, not how detailed the target will be. The pixel size of the Atik 414ex is 6.45µm and the Canon 40D is 5.7µm, so M57 will actually be smaller on the Atik than it would be on the Canon 40D using the same OTA...

The Canon 40D will have a lot more space around M51, but cropping 1:1 will produce a larger M51 than the Atik 414ex...#

I knew there was something to do with pixel size, just needed it pointing out, thank you.....

 

 

 

[DougM43]

Think I need to do some more homework, I thought I was beginning to understand a little better,, smaller chip = larger image, but obviously it's not as simple as that. I won't pretend to understand the pixel size with focal length so a lot to learn and understand,,but Stuart I thank you for  your input and discouraging me from spending on another camera.

Steve I'm sure with time I can learn to take and process a much better image but would it not be the case that if I can do that with a small image then to do the same to a larger image it would result in a "better" image? Or am i as usual missing something in my thinking, that said your image is close to what I'd like to achieve, thanks.

 

[johnrt]

There is an absolute guaranteed 100% works every time trick that will get you a larger image on your camera chip - buy a telescope with longer focal length. Works every time.  

I have images at up to 900mm focal length on an heq5 so no worries about your mount. 

[ollypenrice]

To put this as simply as I think it can be put...

What matters is how many pixels you put on the object.

This is infallible. It works in all cases. There are two ways, and only two ways, in which you can put more pixels on the object.

1) In a given scope you choose a camera with smaller pixels. (If you image M51 with 8 micron pixels you will put 'x' number of pixels onto it. If you image it with 4 micron pixels you will put four times 'x' number of pixels on it. Twice as many across and twice as many up and down. When this is displayed on your screen it will mean the object will be four times larger, twice as wide and twice as high.

2) With a given camera you can use a longer focal length telescope.  (If you double the focal length of the scope it will project an image four times larger - twice as wide and twice as high - onto the chip so the object will cover four times as many pixels. When shown on your PC screen it will will be four times larger.

The Atik 4000 has a 15x15 mm chip and the SXVH36 has a 24x36mm chip. They have exactly the same 7.3 micron Kodak pixels. In any given scope the image they take of M51 will be exactly the same size. They have both put the same number of pixels onto the object. All you get with the big chip in this case is more empty sky around the object. You can either crop this sky off or zoom in to 100% on your PC and M51 will be identical in both images.

So I'm going to stick with this mantra:

What matters is how many pixels you put on the object. Not how many you put on the sky around the object, just the object. 

M51 is precisely the same size in both these images.

Olly

 

 

[StuartJPP]

3 hours ago, ollypenrice said:

What matters is how many pixels you put on the object.

Well said...no ambiguity there (although you will still have to calculate it for each combination).

[ollypenrice]

15 minutes ago, StuartJPP said:

Well said...no ambiguity there (although you will still have to calculate it for each combination).

Gawd, yes, it has taken me a while to come up with this but the teacher in me is always on the lookout for the shortest, simplest way tp express an idea. I'm relieved that it strikes you as OK!

Olly

[gnomus]

Doug

[Whisper] Be very careful of that Olly guy.  He starts off all helpful like this and the next thing you know he's talking you into purchasing all sorts of ludicrously expensive hardware.  

[ollypenrice]

1 hour ago, gnomus said:

Doug

[Whisper] Be very careful of that Olly guy.  He starts off all helpful like this and the next thing you know he's talking you into purchasing all sorts of ludicrously expensive hardware.  

Who me Gov? Anyone would think I was on commission.

Any vendors interested in systematic corruption and giving me backhanders should contact me via my regular email address but, for Gawd's sake, don't mention it on 'ere. They're red hot. Schtum! Schtum!!

Olly

[DougM43]

On 26 April 2016 at 15:59, ollypenrice said:

To put this as simply as I think it can be put...

What matters is how many pixels you put on the object.

This is infallible. It works in all cases. There are two ways, and only two ways, in which you can put more pixels on the object.

1) In a given scope you choose a camera with smaller pixels. (If you image M51 with 8 micron pixels you will put 'x' number of pixels onto it. If you image it with 4 micron pixels you will put four times 'x' number of pixels on it. Twice as many across and twice as many up and down. When this is displayed on your screen it will mean the object will be four times larger, twice as wide and twice as high.

2) With a given camera you can use a longer focal length telescope.  (If you double the focal length of the scope it will project an image four times larger - twice as wide and twice as high - onto the chip so the object will cover four times as many pixels. When shown on your PC screen it will will be four times larger.

The Atik 4000 has a 15x15 mm chip and the SXVH36 has a 24x36mm chip. They have exactly the same 7.3 micron Kodak pixels. In any given scope the image they take of M51 will be exactly the same size. They have both put the same number of pixels onto the object. All you get with the big chip in this case is more empty sky around the object. You can either crop this sky off or zoom in to 100% on your PC and M51 will be identical in both images.

So I'm going to stick with this mantra:

What matters is how many pixels you put on the object. Not how many you put on the sky around the object, just the object. 

M51 is precisely the same size in both these images.

Olly

 

 

Superb information, thank you very much.

I have started making up a spreadsheet to help me decide on the best combination of scope and camera, bit Ott maybe but flicking between web pages looking at various cameras and scopes was getting quite confusing. One thing I've discovered so far is to get an image even a third bigger with a camera alone is expensive,, very expensive. I was aware that a longer focal length scope would produce a larger image as it dose visually, but all I have learned so far has said shorter focal length = faster image acquisition, so I had not considered that as an option, but I am now thinking that may be the way to go. "But" if I understand correctly if I was to double the focal length it would take 4x longer to achieve the same image brightness which for my image above would be 200 x 120 second subs, over 6 hrs,, :-( 

So to my next question to add to this invaluable help,, 

If I was to use a telescope with twice the aperture would that reduce the time needed to image the object back by a factor of 4? If that makes any sense.

thanks in advance,

Cheers Doug. 

 

[ollypenrice]

1 hour ago, DougM43 said:

Superb information, thank you very much.

I have started making up a spreadsheet to help me decide on the best combination of scope and camera, bit Ott maybe but flicking between web pages looking at various cameras and scopes was getting quite confusing. One thing I've discovered so far is to get an image even a third bigger with a camera alone is expensive,, very expensive. I was aware that a longer focal length scope would produce a larger image as it dose visually, but all I have learned so far has said shorter focal length = faster image acquisition, so I had not considered that as an option, but I am now thinking that may be the way to go. "But" if I understand correctly if I was to double the focal length it would take 4x longer to achieve the same image brightness which for my image above would be 200 x 120 second subs, over 6 hrs,, :-( 

So to my next question to add to this invaluable help,, 

If I was to use a telescope with twice the aperture would that reduce the time needed to image the object back by a factor of 4? If that makes any sense.

thanks in advance,

Cheers Doug. 

 

1) Is this so? The Atik 460, 490 or 383L have assorted sizes of pixel, all of which I'd call small.

2) Careful. What follows in your post suggests that we need to unravel focal length from focal ratio.

FR is the relationship between aperture and FL. It is simply the focal length divided by the aperture. If the focal ratio remains constant then the exposure time remains constant.*

If we increase the focal length by a factor of two, say from 500mm to 1000mm, and we want to preserve the focal ratio, then we have to increae the aperture by a factor of two. To preserve F5 at 500mm we need an aperture of 100mm and to preserve F5 at 1000mm we need an aperture of 200mm. Now doubling the aperture increases the area of a single optical surface by a factor of four, (Pi r squared) and the cost tends to go by that ratio. In a naive calculation of cost a single surface 200mm Newtonian will be 4x as expensive as a 100mm one. The bad news comes in with a triplet refractor. Here you have six optical surfaces which quadruple in area. Ouch, ouch and treble ouch. (A TEC 140mm triplet apo costs $5,900. If you less-than-double that aperture to 250mm the cost rises to $49,000. Guess which model I have!)

So I'd have expected the 'small pixel' solutuion to be cheaper than the 'more aperture' solution. The single surface or dual surface reflector solution also makes things cheaper. This is why lots of us enjoy the quality and simplicity of refractors in small aperutres but are obliged by price to look to reflectors for larger apertures and longer focal lengths. Fast F ratios are also more expensive than slow and, again, this goes by surface area so many of us put up with slower f ratios in bigger scopes. We simply can't afford big and fast and we know that big and fast and cheap means a lot of work to make it work. (Don't expect the manufacturers of big and fast and cheap to tell you this.)

3) Yes! A nice easy one. Pity about the cost implications attributable to point 2 though!! Don't shoot the messenger...

I begin by deciding what I want to image. I already have cameras so that makes it easy. It tells me what focal length I need to fill the chip with what I want to image. Now, how much aperture can I afford at that focal length? At short FL, quite a lot. At long FL, not a lot. And how much hassle will it be? I'm a 'provider' to visiting guests so the answer to that has to be 'very little hassle' or we won't be up and running when we need to be. If you have more time to tinker than you have more options.

Olly

*What I haven't touched on here is the F ratio myth in regard to focal reducers, which do not have any significant effect on reducing exposure times for objects which fit on the chip without them. For the purposes of your post I don't think this matters much.

[pipnina]

For regular photography, bigger image sensor combined with a proportionally longer fl lens = same field of view but a sharper image over that field of view. This is why Imax films are shot on 70mm film (genuine negatives) (golly).

For astro, I'm not sure this matters so much. We know exactly what our scopes are capable of resolution-wise and can buy cameras to suit the field of view and angular resolution we want. I, personally, would still want the biggest image sensor paired with the biggest telescope... But that's just me being fussy probably. 

 

[ollypenrice]

8 hours ago, pipnina said:

For regular photography, bigger image sensor combined with a proportionally longer fl lens = same field of view but a sharper image over that field of view. This is why Imax films are shot on 70mm film (genuine negatives) (golly).

For astro, I'm not sure this matters so much. We know exactly what our scopes are capable of resolution-wise and can buy cameras to suit the field of view and angular resolution we want. I, personally, would still want the biggest image sensor paired with the biggest telescope... But that's just me being fussy probably. 

 

But you'd also want the pixel size to give the right kind of resolution and you simply can't have it at moderate FL because the camera doesn't exist. The popular Tak FSQ106 can cover an 88mm circle. You can buy a 36x36mm chip, no problem. But only with 9 micron pixels, which are far too big to allow you to resolve to the scope's optical limit. The art of the possible!

Olly

[pipnina]

9 hours ago, ollypenrice said:

But you'd also want the pixel size to give the right kind of resolution and you simply can't have it at moderate FL because the camera doesn't exist. The popular Tak FSQ106 can cover an 88mm circle. You can buy a 36x36mm chip, no problem. But only with 9 micron pixels, which are far too big to allow you to resolve to the scope's optical limit. The art of the possible!

Olly

This is true, although with a big sensor the option is always there to stick a barlow/powermate between the camera and scope to get the pixels "smaller". With a small-sensor camera the scenario cannot be reversed.

I believe drizzle is a feature in PixInsight, too, which would effectively allow a camera with 9 micron pixels to become a camera with 4.5 or 3 micron pixels. But I don't think there's much point in having a 2.5 degree field of view while trying to image a 0.4 degree target.

I just think that a larger sensor is more flexible than a smaller one. But if Doug43 already has a wide setup... Maybe the smaller sensor wouldn't be so restrictive.

[DougM43]

Again many thanks for the input

, rarther than buying a new camera I was thinking along the lines of the 150PDS or the Quattro 200, they both have longer focal lengths and a shorter focal ratio than my ED80 so if I'm understanding correctly would give a larger image in less time, although by how much and how long I'm not so sure,, the spreadsheet I'm working on to work this out has got far too complicated so I need a rest from it for now to preserve my sanity.

there must be something on the internet where you can input pixel size, FL, FR and aperture to show image size , not just FOV that someone much cleverer than me has already done. 

I was just thinking it was only a couple or three months ago when I did my first polar alignment (which was difficult enough to grasp) and now I'm trying to work this stuff out,, this is not a learning curve,, it's an inverted vertical twisting loop the loop.

Again many thanks.

Doug.

[pipnina]

4 minutes ago, DougM43 said:

Again many thanks for the input

, rarther than buying a new camera I was thinking along the lines of the 150PDS or the Quattro 200, they both have longer focal lengths and a shorter focal ratio than my ED80 so if I'm understanding correctly would give a larger image in less time, although by how much and how long I'm not so sure,, the spreadsheet I'm working on to work this out has got far too complicated so I need a rest from it for now to preserve my sanity.

there must be something on the internet where you can input pixel size, FL, FR and aperture to show image size , not just FOV that someone much cleverer than me has already done. 

I was just thinking it was only a couple or three months ago when I did my first polar alignment (which was difficult enough to grasp) and now I'm trying to work this stuff out,, this is not a learning curve,, it's an inverted vertical twisting loop the loop.

Again many thanks.

Doug.

Stellarium will show you the size of objects in the image:

It requires a plugin, I believe it is not activated by default but is installed with the program.

Launch stellarium, press F2 on your keyboard (on a laptop this might require pressing the FN key as well?) and look for a plugin in the plugins tab called "Field of view".

 

There was a website that did the same but with better information, but alas I can't find it any more. (It was book marked on my chrome... on a previous windows installation!!!)

 

I hope that's what you're after

    ~pip

[DougM43]

1 hour ago, pipnina said:

Stellarium will show you the size of objects in the image:

It requires a plugin, I believe it is not activated by default but is installed with the program.

Launch stellarium, press F2 on your keyboard (on a laptop this might require pressing the FN key as well?) and look for a plugin in the plugins tab called "Field of view".

 

There was a website that did the same but with better information, but alas I can't find it any more. (It was book marked on my chrome... on a previous windows installation!!!)

 

I hope that's what you're after

    ~pip

That is exactly what I'm after,, thank you very much

Doug. 

[DougM43]

Ok I've been adding various scopes and cameras to the stellarium plugin and of the cameras I've tried so far the Atik 420L is coming out as a good bet both for image scale and price, supprisingly (to me anyway) apperture with the various PDS imaging scopes I've put in isn't making that much diffrence, (yes I know I was told)  

So one more question if you folks don't mind. Is there any reason I can't fit the Atik 420L to my ED80 to increase the image size over my 40D? Or to put it another way, would the two be a good combination.

again, many thanks.

Doug

[pipnina]

7 minutes ago, DougM43 said:

Ok I've been adding various scopes and cameras to the stellarium plugin and of the cameras I've tried so far the Atik 420L is coming out as a good bet both for image scale and price, supprisingly (to me anyway) apperture with the various PDS imaging scopes I've put in isn't making that much diffrence, (yes I know I was told)  

So one more question if you folks don't mind. Is there any reason I can't fit the Atik 420L to my ED80 to increase the image size over my 40D? Or to put it another way, would the two be a good combination.

again, many thanks.

Doug

I'm not sure, the 420L is only 1600x1200 pixels.. That's less than a 1080p screen. I think the sensor is smaller than the D40, too, and by quite a way.

It has smaller pixels, though, so what you would get is a higher-sampled image of whatever you pointed the scope at. It being a CCD also means there's likely a lot less noise.

I don't think it's in any way suitable for wide views, but for smaller targets I'd say it's a goer.