Astrophotography - DSLR? Webcam? Motor?

[NeilFawcett]

From having looked into this I'm a little confused about a couple of things. Hopefully someone can answer these newbie questions:-

DSLR focusing

If you attach your DSLR, switch it to manual focusing, I assume then you just use the camera viewfinder (or liveview) like the regular eyepiece and turn the focusing wheel on the telescope until things are in focus?

DSLR vs Webcam

There seems to be a lot of talk about using webcams over DSLRs? I'm a little confused why a webcam would be preferential? Surely the DSLR has far better optics (& control)?

I understand the power of the webcam is it (easily) takes films, which in effect are loads of individual frames that can be stacked (via software) but surely the DSLR can do this to? ie: Take 20 x 1 second exposures? Or some modern DSLRs (like mine) can take HD video. Surely stacking those outputs would be superior? Surely 10 seconds of HD video is going to contain some great information for a stacker (ie: 10x30 1080p images)?

Long exposures & motors

If you're trying to capture an object, long exposures are often discussed. For example a minute. Now, how do you achieve this with a regular dobson mount/telescope where the object will invariably move?

With a DSLR, for example, could you not take a dozen 5 second exposures which you then stack? This would surely equate to over a minute of exposure time, but without 'trails'?

Or is it simlpy the cases you have no choice but to have exposures that are minutes long, and lots of shorter individual exposures simply won't cut it?

[swag72]

DSLR and telescope - You attach the DSLR to the back of a scope using a T thread and focus using the telescope focuser. Imagine the telescope just as a big lens.

Can't help you on webcams I'm afraid.

Regarding long exposures - If you want them longer than a minute or 2 you are going to need a guiding scope and camera on your mount. The mount people generally use is a GEM (German equitorial) and with motors it will track the sky. You hook it all up to a PC and use a guiding programme that keeps the guide star in the same plac ein the frame for say 10 minutes. These are then stacked to increase the amount of signal you get. There are mega mounts for many £££££'s that can manage unguided subs for 10 minutes or more.

If you are thinking about a dabble in AP, DSO's in particular, the book 'Making every photon count' that you can buy from the book section of FLO is an excellent place to start and will answer most of your questions.

Hope that helps, I am sure that people will be around soon to fill in the gaps that I have left!!

[NeilFawcett]

DSLR and telescope - You attach the DSLR to the back of a scope using a T thread and focus using the telescope focuser. Imagine the telescope just as a big lens.

And as regards focusing? Was my assumption correct that basically you just turn the focusing wheel on the scope as per normal until the image/sky is in focus through the camera's viewfinder/liveview?

Regarding long exposures - If you want them longer than a minute or 2 you are going to need a guiding scope and camera on your mount.

Well this is where my brain doesn't compute, I suspect due to lack of knowledge...

Let's say we're photographing the ring nebula. Let's say we'd need your two minute exposure to achieve this. Without a motor this will result in 'trails'... Why can't I do a 20 second exposure with my DSLR, re-center the object in the telescope, and repeat 6 times... Then stack them? Won't that result in the same basic outcome, but without the need for a motor?

eg: How would someone with a Skytracker P200 dobson do a long exposure? Surely impossible?

[Capricorn]

Focusing: Seems the best is to get a Bahtinov mask and set the focus with that. Then LOCK everything down. Remove the mask and start collecting photons from the object. If you use just the camera and screen then pick a bright something or other to determine the best focus, then lock everything down.

DSLR/Webcam: Seems that we use a DSLR for the ability to get the longer exposures that the DSLR is capabile of producing. The chip is bigger, usually colour anf you probably own one. Disadvantage is the IR cut off filter fitted to many - it cuts off bits we would like to get to the chip.

Webcams basically take a movie, so you get a 2 minute movie of a planet, feed movie into a program, select the top 50% good frames, then stack one on top of the other. Bit more giggery pokery with the software and a good/decent image is produced of the planet. The movie approach is no good to DSO's. So Webcams are primarily used for planets, DLSR's for DSO's.

If you have a Dobsonian and want to do long exposure astrophotography then sell it and buy a motor driven equitorial. This comes up on a regular basis and for the last 3 or 4 years I have been on SGL no body seems to get it.

Someone will put up an image of the moon they managed once and they says here is proof that a dob can image. If you want a poorish image of the moon once every 6 months then keep the dob, if you want to get into astrophotography then buy an solid motor driven equitorial. It is really that simple.

You can set up an equitorial to track reasonable for 30-60 seconds so to get a DSO image you may need to get say 20 60s exposures (I have no real idea)

To get longer exposures then a guide scope and camera system is required, scope, camera, laptop say £1000.

[Freddie]

A webcam or specialist cam such as the DMK can take more frames more quickly than a DSLR. You may be on 1/30 sec exposure and you want to capture a few thousand images quickly. The short exposure will capture the moments of good seeing. The best frames are then stacked together in something like Registax.

[StuW]

From having looked into this I'm a little confused about a couple of things. Hopefully someone can answer these newbie questions:-

DSLR focusing

If you attach your DSLR, switch it to manual focusing, I assume then you just use the camera viewfinder (or liveview) like the regular eyepiece and turn the focusing wheel on the telescope until things are in focus?

Correct.

DSLR vs Webcam

There seems to be a lot of talk about using webcams over DSLRs? I'm a little confused why a webcam would be preferential? Surely the DSLR has far better optics (& control)?

I understand the power of the webcam is it (easily) takes films, which in effect are loads of individual frames that can be stacked (via software) but surely the DSLR can do this to? ie: Take 20 x 1 second exposures? Or some modern DSLRs (like mine) can take HD video. Surely stacking those outputs would be superior? Surely 10 seconds of HD video is going to contain some great information for a stacker (ie: 10x30 1080p images)?

For brighter objects such as planets or the moon, the advantages of a webcam are two-fold, the smaller sensor gives higher effective magnification along with the ability to take video that can be split into frames and stacked to reduce noise and increase detail. This is the thing that confused me at first, but if you realize that the size of the object hitting the sensor is the same and you change the size of the sensor, then it is this that determines image magnification (over simplistic because i'm not including pixel sizes, but easier to understand)

Long exposures & motors

If you're trying to capture an object, long exposures are often discussed. For example a minute. Now, how do you achieve this with a regular dobson mount/telescope where the object will invariably move?

With a DSLR, for example, could you not take a dozen 5 second exposures which you then stack? This would surely equate to over a minute of exposure time, but without 'trails'?

Or is it simlpy the cases you have no choice but to have exposures that are minutes long, and lots of shorter individual exposures simply won't cut it?

This is one I'm not 100% on, but as I understand it, nothing can replace stacking lots of long exposures, including stacking even more shorter exposures and therefore you need a tracking mount at the least and ideally some sort of guiding setup.

[NeilFawcett]

Focusing: Seems the best is to get a Bahtinov mask and set the focus with that. Then LOCK everything down. Remove the mask and start collecting photons from the object. If you use just the camera and screen then pick a bright something or other to determine the best focus, then lock everything down.

And by "set focus" there is nothing to do on the camera at all. Simply the telescopes focusing wheel is beign turned out the outcome viewed on the camera's viewfinder/liveview?

If you have a Dobsonian and want to do long exposure astrophotography then sell it and buy a motor driven equitorial. This comes up on a regular basis and for the last 3 or 4 years I have been on SGL no body seems to get it.

Someone will put up an image of the moon they managed once and they says here is proof that a dob can image. If you want a poorish image of the moon once every 6 months then keep the dob, if you want to get into astrophotography then buy an solid motor driven equitorial. It is really that simple.

You can set up an equitorial to track reasonable for 30-60 seconds so to get a DSO image you may need to get say 20 60s exposures (I have no real idea)

Well, I think we're getting to the nub of it.

If we're talking about photographing a planet, would it be realistic to suggest that can be done with a dobson? eg: 50 shots taken with a DSLR then stacked? Surely there'd be no "trail" issue with that?

And if we are talking about photographing a nebula, I'm still confused why say 20x10 second exposures - ensuring the object is centered between shots - wouldn't still give a reason outcome when stacked? Or is there simply no choice but to do a single X minute exposure when photographing just an object?

[NeilFawcett]

A webcam or specialist cam such as the DMK can take more frames more quickly than a DSLR. You may be on 1/30 sec exposure and you want to capture a few thousand images quickly. The short exposure will capture the moments of good seeing. The best frames are then stacked together in something like Registax.

Ahhh! I think I understand. So if we trying to capture Jupiter, then we want lots of quick shots to try and get a number where atmospherics have least messed up the image?

if we simply did lots of 1 second exposures of Jupiter, each would be 'mushed' up by atmospherics differently. Instead we want lots of 1/30th of a second exposures to try and capture a number of "good seeing" moments?

I assume the stacker determines as it goes through a thousand images/frames which are the "good seeing" images/frames? It's not done manually by the user before stacking?

The reason I ask is in your example, I'm trying to see the difference between a webcam, and a DSLR taking a HD movie (at 30fps in 1080p)? Surely 20 seconds of video from a DSLR would be superior to 20 seconds of footage from a webcam?

[swag72]

Yes you can take a 20 second exposure of the ring nebula and you can take 10 images and stack them. Whether or not the detail will be the same in this as in 1 x 200s exposure is difficult to say. Certainly, when you start taking a 20 minute exposure for example, you will get far more of the faint detail showing. Many of the DSO's have very faint outlying detail and much of that will not be captured in a very short exposure.

[StuW]

The reason I ask is in your example, I'm trying to see the difference between a webcam, and a DSLR taking a HD movie (at 30fps in 1080p)? Surely 20 seconds of video from a DSLR would be superior to 20 seconds of footage from a webcam?

Due to the image scale on the different sized sensors. Jupiter at the focal point of a scope may be 1mm in size, so on a webcam with a 5x5mm sensor it takes up 1/5th of the sensor, but a DSLR with a 20x30mm sensor, the image would be tiny.

[x6gas]

In theory (i.e. assuming a perfect sensor) there would be no difference between a single 20 second exposure and 20 x 1 second exposures stacked using an 'add' algorithm. My understanding is, though, that things like the sensor's Quantum Efficiency and read-out noise (I say "things like" because I am clearly out of my depth here!) mean that you are better off with longer subs where the signal-to-noise ratio is better before you stack the subs...

[NeilFawcett]

Due to the image scale on the different sized sensors. Jupiter at the focal point of a scope may be 1mm in size, so on a webcam with a 5x5mm sensor it takes up 1/5th of the sensor, but a DSLR with a 20x30mm sensor, the image would be tiny.

Ahhh! So the DSLR's superior optics all go to waste! ie: Most of that lovely sensor isn't even used?

So if I was looking for a first telescope, with the opportunity to do some basic photography as well, should I set my eyes on webcam/laptop instead of DSLR for ease of use and general results? And would people suggest a motorised scope such as a Skywatcher Explorer 130P SynScan AZ Goto, or a Skywatcher 200P dobson where you get more light gathering for your money instead (but no motor and ease of locating objects)?

[Tinker1947]

You will need a EQ mount for tracking, there good when aligned for around a minute, then longer exposures will require autoguiding, EQ mounts have weight limits you need to be on the safe side for a stable image depending on you chosen imagine path, different scopes for Planets, Nebulars and DSO's, one type won't really do justice to all....

[swag72]

Depends on what you want to image - DSO's require a different setup to plants and solar for example. You will not get a one size fits all telescope and camera combo.

[NeilFawcett]

Depends on what you want to image - DSO's require a different setup to plants and solar for example. You will not get a one size fits all telescope and camera combo.

I'm fascinated by why a telescope that's good a looking at say Jupiter, quarter of a billion miles away, isn't just as good as looking at the light from a nebula that's passing by it?

OK, there's far more light from Jupiter, but surely in bother cases (Jupiter and a nebula) it's about capturing as much light as possible? Can you point me to something that explains why they're such different battles?

Is there not a solution that is a best-of-both-worlds solution. ie: Isn't the best, but adequate enough so you can dabble in both? Is there a spec for a £200-400 scope, with say £100 of webcam additions that would allow reasonably good views and digital capture of planets, whilst also allowing some nebulas and galaxies to also be captured to reasonable degree?

[JamesF]

Someone will put up an image of the moon they managed once and they says here is proof that a dob can image. If you want a poorish image of the moon once every 6 months then keep the dob, if you want to get into astrophotography then buy an solid motor driven equitorial. It is really that simple.

For DSO imaging I absolutely agree that a dob isn't the way to go. You can see just how bad they are for solar system images here:

http://stargazerslounge.com/topic/155275-upgrading-from-spc900-to-dfk-or-similar-worth-it/#entry1605194

for example.

James

[swag72]

If you are wanting to do AP, then you will benefit from different scopes and camera's for the 2. If you ae interested in visual, then I understand that a good all rounder is considered to be a 120 refractor.

There's many 'what telescope' threads in the beginners section that can explain far better than me what and why the difference exists. I'm not a technical kind of girl, just remember bits that I read!!

[JamesF]

Is there not a solution that is a best-of-both-worlds solution. ie: Isn't the best, but adequate enough so you can dabble in both? Is there a spec for a £200-400 scope, with say £100 of webcam additions that would allow reasonably good views and digital capture of planets, whilst also allowing some nebulas and galaxies to also be captured to reasonable degree?

The issue is that you're trying to solve two completely different problems.

Jupiter, for instance, as well as being the largest planet in the solar system also has the largest apparent size from Earth (other than the Moon and Sun). It is still only 50 arcseconds at its biggest though. That's 1/72nd of one degree. So for a good image of Jupiter you need as little area as possible per pixel (known as plate scale, or image scale). Because plate scale is inversely proportional to focal length that means you need as long a focal length as you can handle. For some of my images of Mars earlier this year I was using an effective focal length of 5 metres or more.

By comparison, M13 is 1/3rd of a degree -- 24 times wider than Jupiter, and something like the North America Nebula is over 200 times wider than Jupiter, so magnification is not what you're after. You want something that "sees" a wide area of sky and for that you need a "fast" focal ratio. That is, the aperture needs to be as large a proportion of the focal length as you can manage. That's not really compatible with a scope that works on Jupiter. The same scope as I was using above has a "normal" focal length of 1.5m (before I start throwing barlows etc. at it) and if you want a reasonably fast focal ratio for that (f/5 say, which is good but not blindingly fast) you need large optics and they start to get expensive at an astonishing rate. You don't actually need lots of aperture though because you have the benefit of time to collect all the photons you need. You can use a small aperture scope, keep the cost down, achieve your desired focal ratio more easily and still get decent images. Small telescopes have other advantages for DSO imaging too, because long exposures need guiding, and small, short, light telescopes are less hassle to guide than long, heavy ones.

There are other factors that all need to be balanced against each other such as the camera pixel size and the resolution limit of the telescope, but hopefully you can see where things are going already

I really struggle to think of a scope in your price bracket that will work reasonably well for both. I'd probably say the closest you'd get would be something like a 200P on a HEQ5 or NEQ6, but you're way over £1000 there when you've factored in the cost of the mount, let alone all the other bits you'd need. It can be done for less by using a cheaper mount if you want to put in the work, but it's not an easy path to walk.

James

[RikM]

For DSO imaging I absolutely agree that a dob isn't the way to go. You can see just how bad they are for solar system images here:

http://stargazerslou...t/#entry1605194

for example.

James

To be fair James, these were taken with a motor driven Dob rather than a manual one.

I'm fascinated by why a telescope that's good a looking at say Jupiter, quarter of a billion miles away, isn't just as good as looking at the light from a nebula that's passing by it?

OK, there's far more light from Jupiter, but surely in bother cases (Jupiter and a nebula) it's about capturing as much light as possible? Can you point me to something that explains why they're such different battles?

Is there not a solution that is a best-of-both-worlds solution. ie: Isn't the best, but adequate enough so you can dabble in both? Is there a spec for a £200-400 scope, with say £100 of webcam additions that would allow reasonably good views and digital capture of planets, whilst also allowing some nebulas and galaxies to also be captured to reasonable degree?

With planetary imaging it not about capturing as much light as possible. It is about getting a good image scale on the chip relative to the size of the pixels so you get a nice amount of detail. Then it's about taking loads of frames so you get the best chance to get a reasonable number of frames that aren't smudged due to atmospherics. To do that you need a long focal length and a smallish sensor. Remember that if you have a large sensor you get large image files and the stacking software has to work that much harder. (There are ways round that though). A Phillips web cam and a Canon DSLR both have pixels about 5.5µm so they will give the same resolution and detail if used on the same scope.

With deep sky imaging it is all about capturng as much light as possible, and building up a good signal to noise ratio. The reason you can't just centre the object with a manual dob and then take an exposure is that you will start to see star trailing after only a second or so at that focal length and a second isn't long enough to get much signal to work with. Most nice deep sky image exposure times run to several hours, rather than a couple of minutes. Just like with camera lenses, you need the fastest f ratio to get the most light in the shortest time, which is the exact opposite of what you need for planetary imaging. The best advice you could get here is to read Steve's book first, as it could save you making a costly mistake buying the wrong kit.

Edit: my typing is too slow

[NeilFawcett]

Due to the image scale on the different sized sensors. Jupiter at the focal point of a scope may be 1mm in size, so on a webcam with a 5x5mm sensor it takes up 1/5th of the sensor, but a DSLR with a 20x30mm sensor, the image would be tiny.

Sorry, thinking about this I don't understand!?

Are you actually suggesting the image is only projected onto part of the DSLR's sensor? eg: The image is only over a tiny square of it, far smaller than the 23x15mm area?

If so then in reality what size are is actually used by these censors? ie: 1cm square? Less? More?

In my head I simply thought the camera would be places such that the image from the telescope was projected over the majority of the its sensor. eg: 15x15mm in the above (Nikon D90) case. And in the case of a webcam, it would be moved so the telescope's image was projected over say the 5x5mm area of its sensor?

[Cornelius Varley]

Heres an example. If you image the moon with a 1000mm focal length telescope the imageof the moon formed at the focal plain is about 9.2mm in diameter. A dslr has a sensor about 22x14mm in size so the full lunar disk would fit on to the sensor. A webcam such as the SPC900 has a sensor about 3.5x2.7mm in size, in this case the disk will be bigger than the sensor. But, because both cameras are at the focal plain the webcam only covers a smaller area of the disk than the bigger dslr sensor.

Peter

[JamesF]

Sorry, thinking about this I don't understand!?

Are you actually suggesting the image is only projected onto part of the DSLR's sensor? eg: The image is only over a tiny square of it, far smaller than the 23x15mm area?

If so then in reality what size are is actually used by these censors? ie: 1cm square? Less? More?

In my head I simply thought the camera would be places such that the image from the telescope was projected over the majority of the its sensor. eg: 15x15mm in the above (Nikon D90) case. And in the case of a webcam, it would be moved so the telescope's image was projected over say the 5x5mm area of its sensor?

Hmmm... Where to start...

For the sake of simplicity I'll talk about refractors or reflectors here. Much the same applies to SCTs and Maks, but they're just a bit more complicated internally.

The primary lens or mirror of the telescope creates an image that is in focus at a plane that is a distance equal to the focal length away from the lens or mirror. If you like, you can imagine there's a screen at that point. This is the "image plane". Given that the mirror or lens is round, it will be a round image. When you use an eyepiece, the focuser moves the eyepiece back and forth until its point of focus coincides with the image plane and at that point you'll see an image that's in focus. The image here doesn't move and never changes size.

When you focus with a camera (at prime focus rather than afocal), you have to move the camera sensor back and forth so that the sensor actually coincides with the image plane. That's the only way to get it in focus. Generally the circular image will overlap the camera sensor by some margin. If it doesn't, you'll see vignetting (darkening) around the corners of the image. A 640x480 webcam sensor, being only about 5mm by 4mm, doesn't have much of the image at the image plane projected onto it at all, so whatever is central in the image plane (your planet, for instance) appears to cover most of the camera sensor. On the other hand, a DSLR sensor which is four or five times bigger has much more of the image covering it, so what appeared to fill the webcam image will only fill a small amount of the DSLR image. If the pixels of the webcam and DSLR are the same size then the actual diameter of the image of the planet will be the same in both cases, but with the DSLR it will look smaller relative to the overall frame size.

James

[NeilFawcett]

The primary lens or mirror of the telescope creates an image that is in focus at a plane that is a distance equal to the focal length away from the lens or mirror. If you like, you can imagine there's a screen at that point. This is the "image plane". Given that the mirror or lens is round, it will be a round image. When you use an eyepiece, the focuser moves the eyepiece back and forth until its point of focus coincides with the image plane and at that point you'll see an image that's in focus. The image here doesn't move and never changes size

Thanks James, I think it's starting to make sense?

If we even forget about cameras and webcams for the moment, and just concentrate on what you've just described, with the focal plane is where the image from the telescope is in focus...

So:-

Question/point 1 - Different eye pieces (eg 10mm or 25mm) are all working with that exact same image (data) at the image plane (or focal point?) from the telescope, but simply magnifying the center part of it? The important thing here is, the eye pieces are using the same source image/data, and just simply magnifying the center of it to different degrees?

Question/point 2 - So when we use a sensor (webcam or DSLR) we're trying to get that image plane onto the sensor? If this is the case, then how do we work out the actual size (in mm diameter) of the image plane for a telescope? If the image at the image plane for example is 10mm in diameter, then a 5mm square sensor is going to only be capturing 31% of (the center) of the image, with the other 69% falling off the edges of the sensor!?

...or is my noddy understanding totally wrong!? Your following comment seems to imply I have understood it!?

A 640x480 webcam sensor, being only about 5mm by 4mm, doesn't have much of the image at the image plane projected onto it at all, so whatever is central in the image plane (your planet, for instance) appears to cover most of the camera sensor. On the other hand, a DSLR sensor which is four or five times bigger has much more of the image covering it, so what appeared to fill the webcam image will only fill a small amount of the DSLR image. If the pixels of the webcam and DSLR are the same size then the actual diameter of the image of the planet will be the same in both cases, but with the DSLR it will look smaller relative to the overall frame size.

Yes, but in this example both the webcam and DSLR have caught exactly the same data (at the center) of the image (assuming the pixels per mm are the same). Just the DSLR has also caught more of the surrounding image, which surely can do no harm? And maybe the DSLR sensor is higher quality so the data is better?

[JamesF]

I think your understanding is pretty much there

On point 1 yes, that's pretty much what happens. You'll then find different ranges of eyepieces work with different amounts of that image data, varying from around 45 degrees to as much as 100 degrees and there are formulae for working out the true field of view from the field of view of the eyepiece and the focal lengths of the telescope and eyepiece (eyepiece field of view x eyepiece focal length / telescope focal length, if you're interested).

For point 2, you're correct about the focusing. The position of the camera sensor and the image plane must coincide to get an image that is in focus. In reality, the image plane isn't always a flat plane, but slightly curved. That can mean that you can't get the entire image in focus at the same time which is why people sometimes use flatteners, but for long focal lengths it's not such an issue.

The image at the focal plane will in fact fill the entire focuser drawtube as long as the physical design (lenses and mechanism) of the telescope allows it to. Towards the edges it may be very distorted because the optics will probably not be perfect over the entire potential field of view, just as the view through a magnifying class can be distorted when viewing at sharp angles to the face of the lens. Most people I think just worry about what area of the sky they can cover with the camera sensor and whether they can fit the object they want to image onto it and not how much of the available image plane they could be using.

In the case of the DSLR and webcam there are other issues to consider as well. If the sensor pixels are the same size then for a single shot it's entirely possible the DSLR may give a better image. It's not always the case that the pixels are the same size though, and whilst smaller pixels may look like they're going to give you a more detailed image they may also need a lot more light coming into the scope to function correctly because of their smaller area. Sometimes you just can't get that amount of light from the target you're interested in. And smaller pixels may not help at all if they're smaller than the detail that a given telescope can resolve.

Where the planets are concerned all sorts of image processing can be done if you can collect a lot of images of the same thing and can produce a much better image than might be possible with a single shot, but the planets are rotating and you need to get all of those (several thousand, sometimes) images before the rotation causes blurring in the result. In the main, DSLRs just can't do that whereas a webcam can. For my images of Mars earlier this year I was capturing 5000 to 6000 images over a period of nine to ten minutes, for instance.

On the other hand, DSLRs win over webcams on DSOs because if they need lots more light to get a good image, you just take longer exposures. In the main, webcams can't take arbitrary length exposures (though the SPC900 can be modified to do this).

Because they're very bright objects, DSLRs work quite nicely on the Sun and Moon too, though it's quite common to use webcams for imaging, say, individual craters on the Moon, just as would be done for planets. They are probably the two objects where you can pick and choose which camera you want to use depending on what you want to achieve.

James

[RikM]

Very nicely explained