Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Stargazing Telescope guiding techniques

  1. Oct 20, 2017 #1

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    My star pictures are improving and I am at the next stage of needing some money. My tracking seems to be limiting my exposure times to only a couple of minutes and there is a hint of star trails. So I need to do some auto guiding.
    I have been doing some homework. In the systems I have seen advertised, this involves a camera that is looking at a chosen Guide Star and there is a servo system to keep the guide and main scopes pointing at that star.
    Bearing in mind the powerful imaging processing that is available, I am surprised that a system is not for sale using detection of movement of the whole sky. Signal to noise ratio would be vastly better and pointing accuracy would go up accordingly. Are the present guiding systems good enough not to need improvement? FFTs of the present and previous images would show the spatial phase change in 2D and that would give the necessary correction signal. I would not be surprised if chips were already available to do this. Perhaps it's just a matter of cost.
     
  2. jcsd
  3. Oct 20, 2017 #2
    Exciting!:smile:
    That already sounds quite good. What focal length are you imaging at? What mount are you using?
    Correct. One important choice that is left up to the astronomer is how to attach the guide camera to the telescope. I would strongly encourage doing this using an off axis guider. Using a separate guidescope never really worked for me because of flexure between the telescope and guide scope.
    One problem is that you generally want guiding accuracy around an arcsecond or so and an "all-sky" camera cannot resolve individual stars at that small of an angle. You need to be looking at the guide star through the telescope.
     
  4. Oct 20, 2017 #3

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I am not sure that my Wife is "excited" by the notion of spending 'loadsa munny' haha.
    I have a 500mm SW Equinox and, so far, I am using a DSLR. I have read similar comments about the disadvantages of guidescopes and it is interesting that you confirm them. It's too early for me to be buying a dedicated astro camera but I want to be reasonably future proof. I have a 1200mm Newtonian, too and there are reported issues with focus when there are filter wheels and OAGs up at the focuser.
    An OAG would be a lot more elegant than another whole scope, hanging on top of the main scope. Perhaps I should go the way you did.
    Ahh. You mean that there may not be more than a very few stars available on the area of the guide camera sensor. But, if there were a reasonable number of stars on the sensor, then suitable processing can include all their contributions to position information. That processing needs to be more than just measuring a distance and a direction between images as they come in.
    This topic is a bit along the lines of Camera autofocus which has progressed a long way since the early methods. So far that it's pretty well seamless for most regular photos and it's all down to some pretty clever processing.
    PS The mount is an NEQ6, which is very firm and it sits on three large slate blocks, on concrete in the lawn.
     
  5. Oct 20, 2017 #4

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    so this all assumes that your telescope mount has an input for an auto guider ??

    no, only one star is chosen and used as the guide star

    Auto guiding, I have the gear to do that

    this is the sort of things you need to get up and running
    Orion Starshoot auto guider
    this is the camera ......
    https://www.telescope.com/Orion-StarShoot-AutoGuider/p/52064.uts

    this page has the list of software drivers and operating program
    https://www.telescope.com/catalog/product.jsp?productId=99565

    this is the small 50mm scope that it goes into

    https://www.bintel.com.au/product/orion-mini-50mm-guide-scope/


    As you can see the camera has 2 ports, one port connects the camera to the computer to the guiding control software,
    PHD, and the other port connects to the telescope mount to sending corrections data to the mount drive control


    Dave
     
  6. Oct 20, 2017 #5

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

  7. Oct 20, 2017 #6

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    My SW NEQ6 has a phone - style socket on the side, marked "auto guider" and the guy who I bought it from used to auto guide it.
    Yes, that is what they do at the moment. But general principles would suggest that there is a better way if the whole field of the scope is used for positional information. The software would be more advanced than the present system but detecting the motion of the whole of what you see has the potential to improve performance. Processors will do all that sort of thing with images; it's done with digital TV at a much higher rate than is needed in a slow feedback loop and I was putting it up as a suggestion. The only reason that it might not be an improvement if the total input to the sensor is dominated by just one star. But performance would not be worse than the present system, in any case. It would eliminate the need to find just one guide star that's bright enough and would eliminate the problem of double stars; two close stars would just be part of the whole field. I can't see it being done by a tethered laptop; it could need a dedicated processor sited right on the camera, I think.
    Watch this space; a company could sweep the market with this. :smile:
    And you can see all the other stars which could go together to provide more positional information.
    You obviously have a good working system. How well does it perform on the finder bracket? People talk of slop between main and guide scope affecting the guiding.
    From what I have read, an off axis system sounds attractive. But that also has its problems, apparently.
     
  8. Oct 20, 2017 #7

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    awesome :)

    yeah, but using just one star makes the system must easier to computer process :)

    It really only needs a single reference, cuz all it is doing is measuring that single star's drift caused by inaccuracies in the main scope drive/
    polar alignment and correcting for that


    will be interesting to see developments at the years roll by
    Guide scope systems and the PHD software is in high usage throughout the maging community
    one of the main reasons I suggested it :smile:


    as long as it is mounted properly, there isn't really an issue

    gosh, I haven't used an off-axis guider in 25 yrs.... way back when my mate and I were doing
    film astrophotography when I was back in NZ.

    It's main disadvantage is that it's more difficult to find a star you can guide on
    It also takes a little bit of light out of your main optical path going to your imaging sensor
    A separate guide-scope has neither of those issues



    D
     
  9. Oct 20, 2017 #8

    Drakkith

    User Avatar

    Staff: Mentor

    I think the issue is that there's just no need for the kind of processing you're talking about, Sophie. Current methods work just fine and extra accuracy isn't needed. Any extra accuracy provided by your method would be overwhelmed by the blur in the picture introduced by the atmosphere, delay between processing and correction by the mount, imperfections in the mount gears, and a few other things. My own camera and software can already provide sub-arcsecond tracking using a single guide star when everything is set up correctly.

    The difficulty with guiding is to provide a stable platform for the guiding camera. An unstable platform, such as what sometimes happens when the guide camera is attached to a small telescope mounted on top of the imaging telescope, will introduce "slop" or "drift" during exposures. This is one reason why off-axis guiders are used. The off-axis guide camera isn't susceptible to slop or drift, but its FOV and light-gathering ability are heavily limited by the size of the pick-off mirror that transfers the light from the telescope to the camera. It only picks up a portion of the incoming light, so it can be difficult to find guide stars bright enough to use.

    My own method is to use a camera which has a small guide chip mounted just underneath the imaging chip. While the FOV of this guide chip is much smaller than the imaging chip, it is probably larger than using an off-axis guider and its light-gathering ability is equal to the imaging chip.
     
  10. Oct 20, 2017 #9
    This will be good to start learning with. Longer focal lengths are more sensitive to pointing accuracy since the camera is sampling a smaller portion of the sky. In the future however, you may want to increase the focal length to better resolve small objects like planetary nebulas.
    There are of course disadvantages too, such as the limited field of view to find a guide star as Drakkith has already mentioned. If you have a focuser which is able to rotate the field of view, this is less of an issue since you can just rotate the whole imaging train until a suitable guide star enters the FOV.
    Yes, most Newtonians have very little back focus which can prevent you from reaching prime focus if the distance to the imaging chip is too large. Newtonians are also prone to severe flexure issues so a separate guide scope may not work well with this telescope.
    Does the hand control have a port which can be directly connected to a computer? If so, it may be preferential to use this rather than the dedicated autoguider port.

    As far as purchasing an autoguider, the Orion star-shoot camera davenn mentioned is good. I use an ASI120 which is very similar to the star-shoot camera. Pretty much any monochrome camera which has very small pixels can be used for autoguiding.
     
    Last edited: Oct 20, 2017
  11. Oct 20, 2017 #10

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I was thinking about your comment and my response afterwards
    There's no point in tracking multiple stars .... what's the point ?, they are all moving by the same amount
     
  12. Oct 21, 2017 #11

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Firstly, computer processes are not something for the user to worry about (we don't know precisely how our computer works, do we?). Secondly, having to look for a suitable guide star is not necessary when there are nearly always a number of high mag stars in any slice of sky. Having to find yourself a guide star is because the signal to noise of a random bit of space is not high enough. " they are all moving by the same amount" is the reason that you should use them all at once for guiding. You wouldn't need to select one at all and that would be one less operation. You may have had the double star problem? That wouldn't exist. We would just turn on FFG (Full Field Guiding), sit and watch.
    The argument "we don't need' is never valid, particularly in a high tech business like AP. You will remember the days of having to find a star and to track by hand. Many manual users would have said that they didn't need automatic tracking but Goto and tracking are taken for granted these days. I'm only suggesting something that 'can ' be done and which would make life easier. There can't be any objection to that, surely.
    It won't make any difference to my AP because it would cost the first customers many hundreds of quid to be proud owners but that doesn't mean I wouldn't like to have the system.
    No, it (Synscan) has a screw-on connector. I think I would be aiming to dispense with the synscan and have direct control from the computer (when I have one).

    Did you (and Dave and anyone else) decide against off axis guiding or was the system just not available when you started guiding in the dim and distant past? It does sound attractive in many ways and I would be starting from scratch. This sort of information is very useful for me.
     
  13. Oct 21, 2017 #12

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    read what I wrote above :wink:
     
  14. Oct 21, 2017 #13

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Oh yes - 25 years ago (with fewer DSLRs around for AP, I imagine). I guess you have not identified any problems so there's no reason for you to do different now. But there would be less clutter on the scope without an extra tube there. I'd also like as many other opinions as possible. I am so damned 'careful' when spending hobby money.
    I did wonder about using a guider with my 250mm camera lens but the mount is very good and, with f4 aperture, exposures would be shortish. It could always sit on the refractor for guiding, in any case. It was a surprise for me to see photos of astro set ups with many telescopes bolted together on the same mount.
     
  15. Oct 21, 2017 #14

    Drakkith

    User Avatar

    Staff: Mentor

    You're talking about 2 different things I think. One is the tracking of multiple guide stars to increase accuracy. The other is making it easier for the user to guide. Computerized guiding systems have overtaken manual guiding less because they are more accurate and more because they are far easier and less time consuming to use. Having a small system that I could simply place on my telescope, plug into the mount, and then hit a button to have it start autoguiding with no tedious setup like current methods have would be wonderful. But it's mostly a pipe dream unless it comes built-in to the mount, of which I've only seen a single mount by Meade that has this feature, and the mount costs 5000+ I think.
     
  16. Oct 21, 2017 #15

    Nidum

    User Avatar
    Science Advisor
    Gold Member

    Are there computerised star maps available that give you a true view from any point on Earth and at a particular GMT ?

    If there are then would it be possible to locate the area of sky or specific star of interest on the computerised map and then use the computer to calculate the viewing angles for the telescope both to set them initially and then continuously update them as the Earth rotates ?
     
  17. Oct 21, 2017 #16

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    My idea is not just multiple guide stars as much as using the whole image ( whatever it happens to be) in the field of the scope. Not the same process as tracking a single star.
    If you bear in mind that the electronics inside of most electromechanical systems gets ever cheaper, the processor would be a small fraction of the cost, once it’s an established system.
    There is always a time during which new things are ‘objects to aspire to’ (an Apple phrase). After a while they are cheap and everyone has one.
    A bit of a pipe dream, I agree but I am putting forward an idea not a piece of kit (yet). It would increase accuracy and make life easier.
     
  18. Oct 21, 2017 #17
    I tried using a separate guide scope when I first started imaging. It just wasn't for me. I couldn't get the results I wanted due to flexure between the imaging scope and the guide scope. I tried making the connection between the two scopes as solid as I could, but it still didn't fix the problem. I eventually switched to using an off-axis guider and my images dramatically improved. I now use off-axis guiding almost exclusively for exposures 10+ min long.
     
  19. Oct 21, 2017 #18

    Drakkith

    User Avatar

    Staff: Mentor

    Using the whole image is the same as using multiple guide stars. That's what it reduces down to at least. One reason only one guide star is used is that it takes considerably less time to download and process a 50 x 50 pixel section of the image around the guide star compared to a multi-megapixel image. When doing image processing on my 2 megapixel images on my desktop computer, it can take 5-10 seconds just to make a single change to the image. And those changes are often a simple addition or multiplication of each pixel, nothing close to finding a thousand stars, storing their locations, then doing the same thing on the next image and comparing the two to compute a tracking correction factor to send to the mount.

    Tracking a single star is extremely easy. The program just looks at the small image, computes the point of greatest intensity, does the same thing to the next image, and compares the two. Having multiple stars to track greatly increases the complexity of the program and the resulting computational time. Instead of just finding the point of greatest intensity, the program would have to do some very complicated stuff to keep track of a huge number of stars whose relative intensities can vary from image to image. You'd essentially be plate-solving every single image. Not only that, it would have to keep track of the differential rate at which the stars in the image move when the mount is moved (not a trivial issue when the FOV is large enough).

    As I said before, greater accuracy isn't needed. Every method available can have sub-arcsecond accuracy when done correctly. What's needed, in my opinion, is to make it easier on the user. Which is one of the main things that makes Apple products so popular.
     
  20. Oct 21, 2017 #19

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    yes, programs like Stellarium. you can set it up to where-ever your location is

    you can hook up Stellarium to your scope and use it for finding objects etc

    There's only really 2 purposes for doing guided tracking
    1) for correcting errors with the scope drive accuracy ( gears in the drive chain ... slop/backlash)
    2) for correcting errors in the polar alignment done when setting up the mount



    I agree with Drakkith's comments

    I still see no purpose in tracking multiple stars .... what's the point ?
    Since all the stars in the FOV of the guider are all moving the same, then you only need to track one of them to see the inaccuracies
    that need to be corrected for that I commented earlier on in this post



    Dave
     
  21. Oct 22, 2017 #20

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I was confused at first by the conflicting views here and elsewhere about OAGs or separate scopes but I have concluded that it's as broad as it's long. People like the system they use, which sort of indicates that it's fairly straightforward as long as the pixel size is small enough for the main telescope focal length - which makes sense.
    I am probably going to need to get a Laptop (PC) and not use Apple for talking to the scope equipment.
     
  22. Oct 22, 2017 #21

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    It's not, really. I am suggesting using a totally different approach to processing the guide scope image.
    You are talking about the limitations of your general purpose processor and your programming and doing the process 'your way'. Your TV is doing the sort of video processing (with dedicated hardware) that would take many seconds per frame on your computer. All I am suggesting can be achieved with video speed processing. If you reject the idea on the grounds of what you could do with your system, you are not giving it a chance. There would be no need to download the camera image all the way to a processor on your computer, which would also take some time unless you used Gigabyte Ethernet speeds. The processor could to be connected directly (or even integrated) with the guide camera (not a problem). The only problem with realisation of this could be the cost of low volume production but technology now allows low volume chips to be produced so it could arrive sooner than you know it. Tracking a star is a lot easier than what a driverless car needs to achieve (multiple sub images with different motion vectors) and they are out on the road.
    Whether or not all this is worth while is certainly a question but the convenience factor would always be there. And you should bear in mind that it won't be too long before you will be able to tell your guider "Show me M42 please and give me an hour's exposure with telescope A." (Maybe not in the spirit of the amateur astronomer)
     
  23. Oct 22, 2017 #22

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    To get myself an idea of the sort of video process speeds that are achievable, I was searching for some idea of the time for a typical 2D FFT. Most sources don't seem to commit on actual times but this link seems to suggest that a 2D Double Precision FFT on a 1k X 1k array can be done in around 1ms on a 3GHZ Intel Duo Core. Plenty of time to do a few of those and some other jiggery pokery to drive a control loop with a second or two time constant. That figure is based on a general purpose processor and could surely be exceeded significantly with a purpose built processor.
     
  24. Oct 22, 2017 #23

    Nidum

    User Avatar
    Science Advisor
    Gold Member

    Graphics card ? These can process image data at extremely high speeds .
     
  25. Oct 22, 2017 #24

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Yes - a purpose built processor for commonly required graphics operations. For guiding, the functions would be different but the execution could be similar.
    Shifting large files about with standard comms protocols is very sluggish and a screenful of image data is big enough to make it worth while sticking to the faster transfer speeds that a computer bus will handle (parallel data too!) An integrated system would easily go fast enough.
    This could bring on the pains of Power Supply Drain which can add to the problems of using this system in the field but look at the computing power available on your smart phone. No real problem, I think.
     
  26. Oct 22, 2017 #25

    Drakkith

    User Avatar

    Staff: Mentor

    You're just taking the processing and moving it to the camera. You're still going to be reduced to tracking one or more stars. And I can't see any benefit to tracking more than one. Tracking accuracy isn't even increased.
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook