27 August 2013

TUTORIAL (IMAGING): Setting up and Using AstroTortilla for Plate Solving

AstroTortilla is one of those programs that may seem overly complicated to set up and start using, hence this short tutorial for those interested. It is however a program that once you start using, you will not regret having spent the time to install it and set it up. Best of all, AstroTortilla is completely free and works wonders for plate solving. 

For those unaware, plate solving is a process by which software checks your astrophotograph for star patterns and determines what it is you are actually looking at. Why is this in any way important, though? Simple - when you tell your telescope to go to a deep space object, it will go to it but will probably miss and not have your chosen object centred in view. Of course, a decent three-star goto alignment prior to slewing to your target removes quite a lot of the error but:

1. The goto alignment will very seldom provide you with perfect goto accuracy. 
2. Why would you want to bother picking multiple stars and aligning them perfectly when you would rather spend the time actually imaging your target?

It is here that AstroTortilla excels. You set up your telescope and CCD camera, you send your telescope to your target and then tell AstroTortilla to plate solve. It will capture an image with your CCD camera, analyse it to check where you are actually aiming and then slew your telescope to remove the goto error and centre your target. It can even iterate through this process until your target is very precisely centred within a chosen margin of error (default is 1 arcminute). I tend to only need a single iteration to get my target perfectly centred. 

Interested? Well, you will need to sort out a couple of things beforehand:

1. Your mount must be controlled via a laptop and the likes of EQMod or other such ASCOM software, through an EQDirect USB interface. This is paramount as AstroTortilla needs to interface with your mount directly via ASCOM
2. Download the latest version of AstroTortilla for free from this link

With these things sorted, we proceed.

Yes, it may sound silly going through the installation process but there is a bit we need to pay close attention to. Choose to install AstroTortilla to the default directory. 

As we are installing everything from scratch, leave the default option "Install AstroTortilla, Cygwin, astrometry.net and indexes" selected with everything checked. 

If it were all installed but we simply wanted additional astrometric index files (these are the reference files that contain the star patterns to match up against your captured images and astrometric index files are different for different fields of view), we would select "Install additional astrometric index files"

It is important to install Cygwin (part of AstroTortilla's package) to its default directory. 

We now reach the important stage that we need to carefully consider. 

You may have multiple telescopes and multiple CCD cameras. Matching these up will naturally produce different fields of view and if we want to install AstroTortilla armed with all required astrometric index files to work on ALL your combinations, we will need to figure out the fields of view produced by your narrowest-field setup and your widest-field setup. For example, I have two telescopes and one CCD camera. One produces a wide-field and another a narrow-field, so it is easy to figure out the range of astrometric index files required. 

In order to calculate the fields of view, we can use the free, online Sky at Night Field of View Calculator

You may find your telescope and CCD camera listed and therefore you can simply select these (do verify the details are correct!). My CCD camera, the QSI 660i, was not on the list and neither were my telescopes. So, I therefore verified the details of my CCD camera with the manufacturer, QSI, and then manually entered my first telescope's details (having measured the exact focal length beforehand). 

I note that my Borg 77EDII with my QSI 660i yields a field of view of 1.95 degrees x 1.56 degrees. Now I enter my second telescope's details. 

Finally, I note that my Altair Astro 8" RC with my QSI 660i yields a field of view of 39.31 arcminutes x 31.47 arcminutes

Now we return to the AstroTortilla installation process. First I select my widest field of view. We always use the biggest number of the two. My widest field of view is 1.95 degrees x 1.56 degrees, so I take the field of view number to be 1.95 degrees (the largest). From the astrometric index files list on Widest level, I select:
index 4212, 4.0MB, 2-2.83 deg (120-170 arcmin)
The range of 2 - 2.83 degrees clearly covers my 1.95 degrees. Now for the narrowest field of view. The installation suggests we select about 20% of our narrowest field of view, so let us do that. My narrowest field of view is 39.31 arcminutes x 31.47 arcminutes, so I take the field of view number to be 31.47 arcminutes (the smallest). 20% of it is 6.294 arcminutes. From the astrometric index files list on Narrowest level, I select:
index 4203, 2.5GB, 5.6-8 arcmin
This range clearly covers 20% of my narrowest field of view, 6.294 arcminutes

Please note the installation will actually download these two astrometric index files and every other one in-between them. Checking the file sizes, the astrometric index files for narrower fields of view are quite large and the entire download probably sums up to about 4.9GB in total for my telescopes and CCD camera. Your astrometric index files will probably differ from mine. 

If indeed you only have one telescope and one CCD camera, then it is simpler. For example, let us say that I only have the Borg 77EDII telescope and QSI 660i CCD camera. We calculated the field of view to be 1.95 degrees x 1.56 degrees for this setup. We therefore take our widest field of view here to be 1.95 degrees and the narrowest to be 1.56 degrees. 20% of the narrowest is then 0.312 degrees (or 18.72 arcminutes, multiplying by 60) and we have our numbers for selecting appropriate astrometric index files.

Once you have adequately selected your needed astrometric index files, proceed with the installation and finish it. No doubt the downloading of astrometric index files will take a while, depending on your Internet connection with the server. 

STEP 2. Setting up AstroTortilla with optimum settings for fast plate solving

When you first run AstroTortilla, it might look intimidating and perhaps even unrefined. It is however very powerful and you need not do much to configure it and get it running. 

Let us first summarise the sections of this window. Under Telescope, we connect to the telescope mount and once connected, you are shown your current telescope position in the night sky as well as your intended target. Basically when you issue a goto command for the telescope to slew to your target, it will most assuredly miss and you will notice the difference under Current and Target

Under Camera, we connect to the CCD camera via compatible software such as Nebulosity, MaxIm DLBackyard EOS, etc. We can also ask AstroTortilla to control the CCD camera directly via ASCOM or in fact, select File Open dialog and select an image that we have taken by our own means (without asking AstroTortilla to do it itself automatically). Here we also set the desired exposure time. Unless you are currently using a narrowband filter, 10 seconds usually suffices. If you are using a narrowband filter, it may be a good idea to switch to a regular LRGB filter. If you are using a refracting telescope and are already focused with say, Hydrogen-Alpha (or Sulphur-II), using Red would be a close match for focus (Green is a good match for Oxygen-III). AstroTortilla will work even if you are slightly out of focus anyway. Using a regular LRGB filter, which is broadband, will allow a lot more light through and will produce more stars for plate solving in a shorter exposure time (this is particularly true for Luminance). Once plate solved, you can always switch back to your narrowband filter for imaging without ever touching the focus. 

Under Solver, we can customise a couple of settings, and this is what we will do first and foremost. These settings tell AstroTortilla what field of view we are actually dealing with (to give it a head-start in plate solving and allowing it to take much less time), how many degrees off from your target you wish to look at, the noise tolerance, etc. 

Under Actions, we set off AstroTortilla to do its job. It is of utmost importance that we always check Sync scope and Re-slew to target as these will make sure AstroTortilla does the job of centering your target in view and adding a sync point to EQMod (like a goto alignment point) to correct future slewing accordingly. You could also of course check the third option Repeat until within and enter an amount of arcminute tolerance. Please note this will cause AstroTortilla to iterate through the process at least twice (first time to perform the initial error correction and second time to check resulting accuracy), with more iterations possible depending on how far off you are from your target after the first correction. The Capture and Solve button of course, sets it all off to do its magic. 

The first thing we will do is change some settings in AstroTortilla, under the Solver section. Since I have two telescopes and one CCD camera, I have two setups I can possibly use. I will therefore create two settings files, one for each setup, so that in the future I can load up the one I need and set it off quickly. 

The first settings we will change are Scale minimum and Scale maximum. These are where you tell AstroTortilla the minimum and maximum field of view possible. Let us say I am using my Altair Astro 8" RC telescope first. This provided a field of view of 39.31 arcminutes x 31.47 arcminutes with my CCD camera. We will work with degrees rather than arcminutes and therefore dividing these numbers by 60 gives us the field of view to be 0.655 degrees x 0.525 degrees. We note then that the maximum is 0.655 degrees and the minimum is 0.525 degrees. We need not be 100% accurate (these numbers are a little rounded off) and so I will enter 0.525 in Scale minimum and 0.655 in Scale maximum

I will leave degwidth in Scale units as we are entering everything in degrees. Now we must edit Search radius. The default value of 180 is way too high. This means AstroTortilla will search 180 degrees around your target. This is quite simply savage as though we will miss the target, we will not miss it by 180 degrees! To be safe though, we will set 15 degrees, so I enter 15 here. Please keep in mind this value if it turns out AstroTortilla is not successfully plate solving your images. You may want to first try increasing it to 30 or even 45 just in case. You should never need to go higher than 45. Similarly, you may never be as far off as 15 degrees and so you may decrease it and test it out. As long as it plate solves, the lower this value is, the faster it will work. 

Finally, we will alter Custom options. The first option that is there by default is particularly important, --sigma 1. This defines the noise tolerance in analysing your captured image for plate solving. The default value of 1 is very, very small and pretty much any speck on your image may be considered a star. Even with a cooled CCD camera, this is certainly not the case and in fact, we do not want TOO many stars to be considered for plate solving as it would take ages. For a 10 second exposure (without a narrowband filter in place!), a much better value is something like 50 to 70. In fact, any value between 50 and 100 is good depending on the number and brightness of the stars in your field of view. This value is balanced out with exposure time, as you would imagine. I would set it to 70 and then just alter exposure time to suit. We leave the --no-plots and -N none options there but add two more. We will reiterate to AstroTortilla the field of view we expect. For this, we use the options -H and -L, which are widest field of view and narrowest field of view, respectively. I will stick to the same values as with Scale minimum and Scale maximum and therefore enter -H 0.655 -L 0.525 to the end of Custom options, after adding a space.

There are two more Custom options that are particularly interesting, one of which I was reminded by a comment in this tutorial by Blackwater Skies and the other of which I only found out through this very same comment. My thanks to Blackwater Skies for the contributions. The two options that help speed up plate solving are in some way related to one another and are -r and --objs. The former, -r, need only be added as that. This command instructs AstroTortilla to sort the stars in order of brightest first, therefore allowing AstroTortilla to optimise the plate solving process as it starts with the brightest stars in your image. The second option to help speed up plate solving is rather to ensure that if it is going to fail, that it fails quicker rather than keeping you waiting. The --objs command must be accompanied by a number, e.g. --objs 100. What this command does is instruct AstroTortilla to use a maximum of 100 stars for the plate solving (the developers generally recommend between 100 and 200 stars). After having sorted them in brightest-first, it seems like a good idea to restrict how many AstroTortilla should use as if it fails after the first and brightest 100, for example, there is a very, very good chance it will simply fail altogether. At that point of course you would adjust other parameters such as your exposure time. Please be aware however that these options are not always best. Sometimes you may want to include more stars than 100 or you may not want to restrict the number of stars used. This could be due to persistent plate solving failures. Feel free to remove the --objs command if you feel it necessary and though the -r command tends to be a good idea to leave active, feel free to test by removing it also.

Due to the length, you cannot see the entire Custom options line in the above screenshot. For reference, it reads:
--sigma 70 --no-plots -N none -H 0.655 -L 0.525 -r --objs 100
Before saving the settings file, I note that I will be personally using Nebulosity 3 for my imaging, with my QSI 660i CCD camera. Also, I will be binning my images in 2x2 when it comes to plate solving as this will brighten the image significantly in the exposure time set and it will bring out more stars as well. Binning has the additional benefit that the file size and resolution is reduced by a factor of four, meaning that AstroTortilla will be much faster at processing the image and plate solving it. To set the software, camera and binning mode I will use, I simply select Nebulosity 2/3 from under Camera and then click Setup. Here, I select QSI 500/600 from the list, verify the directory for Nebulosity 3 is correct and select 2 in Binning. Users of a DSLR, which cannot bin images, would benefit from setting Downscaling to 1 under Solver. This will at least downscale the image in order to make it faster to process and plate solve. Binning is most definitely preferable to downscaling simple because binning downscales by its very nature but at the same time, produces an image with a lot more brightness and more stars picked up. Downscaling alone simply makes the image smaller, with no added benefit to the stars detected.

AstroTortilla is now customised for my Altair Astro 8" RC and QSI 660i so I save the settings to a new file, naming it appropriately for future reference. 

Now to customise AstroTortilla to my Borg 77EDII telescope with the QSI 660i CCD camera. We note the field of view was calculated to be 1.95 degrees x 1.56 degrees. This will again define our values of Scale minimum and Scale maximum. We therefore enter 1.56 for Scale minimum and 1.95 for Scale maximum

We will leave Search radius at 15, like before, and --sigma 70 in Custom options (for noise tolerance). I of course will alter the -H and -L options to enter the same values as Scale minimum and Scale maximum. -r and --objs 100 are added on as well. 

Due to the length, you cannot see the entire Custom options line in the above screenshot. For reference, it reads:
--sigma 70 --no-plots -N none -H 1.95 -L 1.56 -r --objs 100
The same settings under Camera are kept, using Nebulosity 2/3, the QSI 500/600 camera and Binning at 2 (for me personally, of course). I now save these settings to a new settings file, naming it appropriately. 

I have now customised AstroTortilla for both my telescopes and their resulting fields of view. Depending on my imaging setup on a particular night, I can load up the appropriate settings file. These settings optimise the plate solving process to make it significantly faster and always yield a solution. You may need to alter exposure time to suit if a solution is not found when plate solving or it seems to be taking too long. For reference, plate solving should normally take under 3 minutes to perform in total.

If for any reason the stars in your images are not coming out round and plate solving is therefore failing (optical defects, collimation, focus, etc), you can add another Custom option to relax AstroTortilla's restriction on what constitutes a star. The command for this is -c, which needs to be followed by a value. As a default, AstroTortilla uses -c 0.01 but if you wish to relax this a little, you can add -c 0.02 to your Custom options. If you increase it, do so by very small amounts, like -c 0.02.

STEP 3. Using AstroTortilla for plate solving

With the appropriate settings file loaded (for the imaging setup being used), the first thing we need to do is connect to our telescope mount via ASCOM. So, we select ASCOM Telescope from the list at the top and then confirm via the ASCOM dialog that pops up (I am using the simulator for this tutorial!). 

Once connected to the telescope, I issued a goto command in Stellarium to slew the telescope to the Veil Nebula

With the telescope on target, you can gauge an idea of the error under Telescope within AstroTortilla (comparing Current to Target co-ordinates). In a real-life situation, the difference here will be much bigger (remember I am using the simulator for this tutorial!). With Tracking displayed under Target, we are ready for plate solving. 

Now we will need to connect to the CCD camera. You have a choice of what to do here. For example, I use Nebulosity 3 for astrophotography so open your program and connect to the CCD camera within the program. Once connected to the CCD camera, select the program from the list in AstroTortilla if you have not done so beforehand (when you created your settings file). 

It is important to start the program and connect to your CCD camera beforehand as AstroTortilla will issue commands to the program itself to start capture, after which it will download the resulting image and start the plate solving. If your program is not listed on AstroTortilla or for some reason is not working as intended, you can always capture the image yourself within your program and save it somewhere. Then in AstroTortilla, select File Open dialog for Camera and later when we start plate solving, it will ask you to open the image file. This ensures AstroTortilla is very universal. You can also of course, connect to your CCD camera directly by selecting ASCOM Camera and then selecting the appropriate driver from the list, which will be installed if you have used your CCD camera in the past. If you are having AstroTortilla capture and download the image itself (and not providing one via the File Open dialog), set an appropriate exposure time as well. 10 seconds tends to be very suitable and you are advised not to make exposures too long, though you can also do 20 seconds or so. Under one minute is always good.

Before starting plate solving, check Sync scope and Re-slew to target under Actions. The first ensures an alignment point is added to the likes of EQMod (this acts like a star alignment, which makes future goto commands more precise). The second ensures the target is physically centred after AstroTortilla plate solves, which is the main idea behind us even using AstroTortilla to begin with. 

To start plate solving, all we need to do now is click the Capture and Solve button and watch it happen. In my experience, one plate solving iteration centers my objects perfectly but if you want to be very precise about it, check the third option Repeat until within and enter a respective accuracy in arcminutes. Please note that this will cause the plate solving process to be repeated again and again until the desired accuracy is achieved. At the very minimum, two plate solving processes will occur - the first to perform the initial centering to your target and the second to check that it is within desired accuracy. Speaking from personal experience, I find my target is always precisely centred in the field of view of my camera without using the Repeat until within option, with a single plate solve. 

In any case, when happy, click Capture and Solve. AstroTortilla will first capture your image, download it and start plate solving (if you chose to provide an image file yourself via File Open dialog, it will just ask you for the image file). The plate solving process starts by searching its astrometric index files (the ones you downloaded for installation) for a match, calculating how far off you are, slewing your telescope appropriately and syncing the difference as a goto alignment point in EQMod. Subsequent iterations of this process are up to you (dependent on your chosen accuracy for repetition). 

This concludes how to use AstroTortilla. It really is that easy and after having customised it for your possible telescope setups, plate solving should be a fast process. Do make sure you play around with exposure time if a solution is not yielded with a 10 second exposure but remember that under one minute is good. Too long an exposure time will basically take longer to capture and will not really be necessary for plate solving. You can of course play around with the --sigma value here as well. Decrease it to include more stars and increase it to reject dimmer stars.

STEP 4. Further optimising plate solving speed

The following technique was pointed out to me by Stargazer's Lounge user Thalestris24 in this thread, with due credit to the author of this tutorial. It is so useful, I thought it deserved its own sub-section on this tutorial.

After you have used AstroTortilla to plate solve successfully a couple of times, you will notice a pattern. Since you have a large number of astrometric index files downloaded and stored for plate solving, AstroTortilla will look through them all until it finds a match to your exposure and then does its magic. If you look closely, it will seem that it will always find a match for your particular telescope-camera pair in one astrometric index file, or two or three, but never seems to make use of the other astrometric index files. This is normal. During the installation, a lot of astrometric index files were downloaded as per your field of view limitations. Most of these however, are not needed (as evidenced by your plate solving successes). We can use this fact to produce a further speed optimisation to plate solving.

To accurately determine which astrometric index files are useful, one must use the Log Viewer by going to Tools and clicking Log Viewer. This will open a new, small window on the side.

Keep Info selected from the list and now carry out your plate solving as usual. The events that appear and disappear quickly on the very bottom of AstroTortilla will be listed in sequence in the the Log Viewer. You can scroll through them to see what has happened over time. You can also scroll right to see which astrometric index file was being looked at, at the time.

Eventually it will successfully plate solve and it will tell you so, showing you which astrometric index file was the successful one in matching your exposure. Take note of the astrometric index file name. Please do remember that from one target to the next, it may end up plate solving successfully with another astrometric index file, though usually this is an adjacent one (e.g. it may normally succeed using index-4206-11.fit but another night on another region of the night sky it will succeed using index-4206-10.fit). Normally it appears two or three astrometric index files does the job for a particular telescope-camera pair. Do note that, as in my case, if you use different telescope-camera pairs, you will no doubt have a couple of useful astrometric index files per telescope-camera pair.

Once you have determined which astrometric index files do it for you, you can physically move the unused ones out of the usual folder so that they are missing during plate solving and thus AstroTortilla does not waste time in looking through them (as they are useless to you anyway). To access the astrometric index files you have, assuming you installed everything in its default directories, go to the following folder:
Here is my own such folder:

There are clearly a good number of astrometric index files there, as per my installation of AstroTortilla. Now, let us say we determined that actually only the following were useful for me:
It is then a simple process of selecting all the other astrometric index files (just the files with .fit as file extensions) and moving them to a custom sub-folder I made, named Not Useful.

Above I demonstrate the created Not Useful folder, selection of all astrometric index files except the useful ones, moving the selected ones to the Not Useful folder and then how the only available astrometric index files are now the ones I deemed useful. By experience, this should make a plate solve that usually takes around 30 seconds (as reported by AstroTortilla) succeed in just under 5 seconds. That is certainly a big deal. The reason the astrometric index files are not just simply deleted is because we may actually need them in the future, which brings us to the caveat of this optimisation technique.

The caveat is that if you do not experiment with plate solving enough to know which astrometric index files are really useful, you may end up moving a useful one inadvertedly and thus causing plate solving to fail on that particular target. This is a simple thing to fix however. Simply access the Not Useful folder and move one astrometric index file back over. Choose one that is above or below the ones you currently have available. For example, if I have the above three but plate solving failed, I would choose to move back either index-4206-08.fit or index-4207-00.fit. I would move only one and retry plate solving until it succeeds as there is no point in deviating too far from the usually useful astrometric index files or introducing too many new ones into our availability pile. 

Kayron Mercieca


  1. Thanks for such a detailed write-up, only one minor update: the -L and -H switches are not necessary, AT will use those switches internally based on the configured scale limits, and you can further let AT automatically adjust those by setting the "Scale refinement" to a some fraction for tolerance, e.g 0.2 for 20%.

    For speeding up a bit, the camera settings dialog supports binning for almost all image sources and with MaxImDL and Nebulosity you can also force a specific filter for AstroTortilla use. Personally, I use Luminance filter, 2x2 binning, no downscaling, and a 2s exposure for all solving purposes on 550mm focal length (1.6x1.2 degrees) and 5s on 1200mm. With those I get a fast download from my camera, a 6-7 second solve time on the imaging laptop for a very quick goto correction.

    Antti from AstroTortilla

    1. Thanks very much for your comment and clarification of the -H and -L switches on "Custom options". "Scale refinement" does sound like a good option for those wishing for extra tolerance on the FOV limits. 2x2 binning does sound like a good idea for plate solving given its still-high resolution, stronger signal and significantly shorter image download time!

  2. Great write up. A few additional things that may help:

    1. If using EQMOD, make sure the 'Append on Sync' option is selected (it should be by default). This ensures that the EQMOD pointing model is updated each time AT successfully solves an image and thus improve the accuracy of subsequent gotos through the night.

    2. Two additional parameters can be used in the "Custom options" box to help speed up solving:


    This re-sorts all the identified stars, or 'sources' in astrometry.net speak, and starts solving with the brightest.

    -objs 50

    This limits the solving attempt to only use the first 50 stars or whatever number you specify. I find that if you can't get a solution from 30-50 stars then the attempt will usually fail so you might as well fail quickly and try again rather than waiting several minutes for the solver to give up. You should use -r and -objs 50 together as far as I can tell.

    3. You can use -c 0.02 (or more) to relax the star shape detection routine. The default value is 0.01 so if you are getting no stars detected try adding it and increasing the value slightly if your stars are bloated or non-round.

    4. You really shouldn't need to use a value as high as 45 degrees for the search radius. Depending on your camera field of view and the initial accuracy of your mount's gotos (before you've done any solving) you can almost certainly use a much smaller value like 5 or 10 degrees. Once you have started building a pointing model (i.e. have a few alignment points) you can usually reduce the value even more, perhaps a couple of degrees depending on your FOV. This will speed up the solving process even more.

    5. When trying to work out the image scale for selecting the index files to download, if you can't find your scope/camera in the Sky at Night (12 Dimensional String) calculator give my imaging toolbox a whirl: http://www.blackwaterskies.co.uk/p/imagingtoolbox.html It has 588 telescopes and 383 CCD and DSLR cameras plus support for focal reducers/barlows and can calculate the height and width of the field of view (in degrees) of any combination.

    6. Alternatively, if you already have a (single unstacked) image from your camera/scope you can upload it to http://nova.astrometry.net/ and have it solved online using the same back-end software as AT. Look in the 'Calibration' section once your image is solved and it will state the height and width in degrees which you can then use for the index files download.

    1. I would like to thank you for your contributions. You reminded me of the -r command and taught me about the -objs command. I have updated the tutorial and screenshots accordingly and have credited you for it! :)

  3. Thanks for taking the time to assemble this useful reference!

    Can you tell me, if you use AstroTortilla to set parameters in Nebulosity (like switching to 2x2 binning or specifying a filter) will AstroTortilla restore the imaging camera to the previous settings upon completion? Or is it up to the user/different software to re-establish the Nebulosity settings?

    Thanks again! Great reference!

    Craig Smith

    1. In my experience with AstroTortilla and Nebulosity, if before using AstroTortilla I have captured an image in Nebulosity, it won't respond to AstroTortilla's commands. What I do here is close Nebulosity and re-open it quickly (to avoid the CCD sensor temperature rising too much as Nebulosity controls it).

      Once Nebulosity is "freshly opened", it responds perfectly to AstroTortilla. However, once AstroTortilla is done with it, the settings in Nebulosity are kept as what AstroTortilla set them to be. This is because AstroTortilla generates a script from your chosen settings and interfaces with Nebulosity, which opens this script and executes its commands. Once finished, Nebulosity remains configured with this script.

      So, if Nebulosity doesn't seem to be taking the exposure AstroTortilla wants, close it and re-open it. Once AstroTortilla has done its job, it's your choice whether to close Nebulosity and re-open it again or just alter the settings. Check everything - binning, exposure time, number of exposures, series name, folder where series will be stored when captured, etc. The last point (folder for storage) is important because AstroTortilla sets a really odd temporary folder to store its plate solving image so you might later be unable to find your actual images (without a search)! :)

      You are very welcome. I am glad this tutorial is helpful to you.

    2. Thanks Kayron,

      I find the sane thing with Nebulosity, that is, if it is used for some captures, then it seems that AstroTortilla cannot access it without restarting Nebulosity,

      As a result of this work-around, I have at least superficially started learning to use scripting in Nebulosity to restore settings after restarts. Makes things easier, even if I wasn't using AstroTortilla.


    3. Nebulosity does have a few quirks from AstroTortilla point-of-view. It has no way of telling us where images are saved currently, and trying to save an image to an absolute path doesn't work either. Thus AT has to set the image saving directory in Nebulosity script to a known location (temp dir) and it the restores to another known dir you can define in AstroTortilla from the Camera Settings -dialog. The default is something like c:\Astro\\ but you can set it to your liking.

      If there's a problem accessing Nebulosity after acquiring images, this clearly needs to be investigated. Can you reproduce the problem with the (ASCOM) simulator camera? If yes, write a bug report with the steps to reproduce the error at astrotortilla.sf.net


    4. I love your tutorials very clear and easy to understand thank you for putting this together. I am now using ASTRO TORTILLA and its so much easier to get my targets aligned correctly. More time imaging and less time searching around brilliant.

  4. Without question this is the best and clearest guidance on Astrotortilla that I have found. Even though I already have gotten it to work for me this has still helped me to refine my use of the program. I only wish I'd had this in the beginning when I was trying to figure everything out for myself.


    1. My thanks for your comment. I am glad it was helpful! :)

  5. Thanks Kayron

    I'm looking to automate an alignment process and this is a great tutorial for me to do this!

  6. Superb tutorial. Can't wait to get out there now - just waiting for the moon, clouds and fog to clear! :)

    Thanks Kayron :)

    1. You're very welcome! I'm glad this is helpful and wish you success. Many seem to have found this tutorial useful and I can tell you that to date, I still use the exact AstroTortilla settings described here, with 100% success rate! :)

  7. Used this for the first time the other night, during one of our increasingly rare clear nights. Gave me something to play with while the moon was fat. Absolutely brilliant. Slewed to the first target, took a snap with AT (all seamless), solved in seconds and then centred the target - spot on. Did this seven times. I could have happily played with it all night, but it was getting a bit nippy!

    One thing I did notice: I'd slewed to M1 and noticed after it took the first snap that it was more or less centred anyway. AT seemed to get a little confused here and kept shifting slightly backwards and forwards in RA. I had to close it down with task manager, but following that it was fine, I suspect it was trying to make a tiny adjustment that my EQ5 isn't capable of doing. Bless :)

    All in all, success beyond my wildest imaginings - thanks for the tute Kayron :)


    1. Doug, my apologies for the late reply. I am really happy to hear that using AstroTortilla after the configuration was very successful! I'm sure that you have since used it further (weather allowing! :)). AstroTortilla is by far one of the most useful programs I use in my procedures. Certainly makes mosaic work spanning several nights a hell of a joy! :)

  8. Have you used AstroTortilla for polar alignment and if so have you any comment?

    Ian Wardlaw

    1. Hi Ian, I have not personally used AstroTortilla for polar alignment. There was a bit of a discussion about this going on the thread I created to advertise this tutorial:


      People's general consensus seemed to be that AstroTortilla does not yet have a mature-enough polar alignment feature to really be helpful. Personally I use EQMod for polar alignment and that's it, though I have in the past used Alignmaster to refine this (Alignmaster is excellent).

  9. Hi,

    Looks like a useful piece of software I'd like to have a go with .

    One thing though, the Index files that are loaded on install, are they available fo download separately ? I'll be installing it onto a laptop that doesn't have internet access (it's in the middle of nowhere in Spain), sop would like to have all the install files and index files ready on a memory stick for it.

    Can I do that, and will the installer take the index files from local disc instead of the internet ?



    1. Hi Stuart,

      Thanks for your comment. During AstroTortilla's installation on your computer here, install everything including the astrometric index files you need. Make a copy of both the "AstroTortilla" folder in your Program Files folder and the "cygwin" folder in your drive C: on a USB stick.

      For AstroTortilla's installation on your remote laptop, only select to install "AstroTortilla" (as "Cygwin and astrometry.net" and "Astrometric index files" will require an Internet connection). Once that's installed, copy-paste the "AstroTortilla" folder from your USB stick to replace your "AstroTortilla" folder on your remote laptop. Do the same with the "cygwin" folder.

      Once done, in theory the laptop will have AstroTortilla installed properly and since you copy-pasted the data from your computer, you will have all the data required including the astrometric index files. This may or may not work 100% - I have not tested it, but it seems sensible to assume it will work. Worth a try within your own computer (or another laptop) I guess.

      Best of luck!

  10. aha, will give that a try, thanks !

  11. This comment has been removed by the author.

  12. I am wondering why for me nothing works

    1. Hi Luca,

      What are the errors/problems you are encountering with AstroTortilla?


  13. Hi Kayron
    It does not work under windows 8.1 unless everything (including installer) is run using windows7 compatibility.
    Once installed it must be run using administrative privileges otherwise it won't connect to an ascom telescope.
    Now it works great!
    Thanks for the tutorial

  14. This comment has been removed by the author.

  15. First of all, great tutorial, thanks for taking the effort to publish.

    I have come across an issue I hope someone can help with. When I use the file dialog box to solve a plate, AT works great. But as soon as I connect AT to my telescope with ASCOM, it becomes very unresponsive. I can't edit any text boxes without waiting an exceedingly long time, and even trying to move the AT window on my screen takes upwards of ten seconds to respond. I can press the 'solve' button butAT to snap the picture, then it takes upwards of a minute just to get to where AT starts solving. After that it starts to crawl through the index files so slow I have to give up on it.

    Anyone have any ideas? This program has such great potential! I am running a Windows 7 laptop with an Athlon A4 processor.


  16. Hi Craig,

    Thanks for your comments, I appreciate them. As for your question, I think this is a case of a problem with the laptop as opposed to the program in its own right. I don't mean to suggest a component in your laptop is faulty, but there's something going on. It could be up to very little free memory and/or a very fragmented harddisk. What I would suggest trying is first and foremost, run the program "msconfig" (just type this in Run and click OK) and under "Startup", uncheck any and all programs you don't need running. Anti-virus programs tend to take a lot of computer resources, for example. The less you have checked on the "Startup" list, the less programs will start up with Windows and take up memory. Please note doing this does not uninstall anything - it just stops the programs from starting with Windows when you boot up.

    Once that's sorted out, reboot and give it a go. If you removed a lot of programs from starting up, you will immediately notice much more responsive performance from your laptop. You can also defragment the harddisk (if your laptop does not use an SSD as for SSDs, this is not needed). When installing/uninstalling and moving stuff around, files get fragmented into pieces and shoved all over the place. This means that the harddisk head has to move around extensively to access a single file at times. Windows has a "Defragmenting Tool" that will tidy everything up and speed up accessing of files on your harddisk. To run this tool, open "Computer", right click on your harddisk drive (C:), go to "Properties" and under "Tools", click "Defragment now". This process can take minutes to an hour or so and is best you do not touch your laptop in the mean time until it is done. Reboot afterwards just in case.

    If all that does not alleviate the problem, I would investigate whether other programs also suffer from sluggish performance. It could be down to a corrupted drive. Windows can check for areas of the harddisk that are corrupted and unpartition them (remove them from being accessible, improving performance as it avoids the problematic areas). To do this, again go to "Computer", right click on your drive (C:), go to "Properties" and under "Tools", click "Check now". Check both "Automatically fix file system errors" and "Scan for and attempt recovery of bad sectors". This can also take a while (an hour and more at times) and is best you leave your laptop unattended for it. A reboot afterwards is also recommended.

    Further to that, you could try adding more RAM to your laptop if possible. Replacing the harddisk with an SSD would massively improve performance as well, whether or not your harddisk has problematic areas since SSDs are flash memory based and incredibly faster than harddisks.

    I hope this gives you ideas to look into!

  17. Hi Kayron --

    Thanks for your response. I'll do as you suggest and see if it picks up AT's performance. Again, great tutorial.

    Best regards


  18. Because I do not have an st-4 port I can only connect telescope to one program at a time. I can either connect to starry night or AT. Because of this I do not think I can use the program. Am I missing something or am I out of luck?

    1. Hi Richard,

      The ST-4 port on your mount is purely for autoguiding purposes and should connect to your autoguiding camera. How is your mount connected to your laptop? I use an EQDirect USB cable. These work on ASCOM, which are a set of free drivers on your laptop that connect to every astronomical device (mount, autoguider, camera, etc).

      In reality, many programs can access one single device at the same time through ASCOM. I normally have EQMod (for slewing), Stellarium (for GOTO) and AstroTortilla connected to the mount at the same time. They all interact beautifully through ASCOM, as if they're one program.

      I hope this gives you a better idea of how it all works together in software. You can read more about the software side of things here:


      If you wish to ask me anything else, please feel free to post here on to e-mail me at: kayron.mercieca@gibastrosoc.org

      Best Regards,

    2. Just to embelish on the above, the EQDirect USB cable connects to a USB port on the laptop but on the mount, it replaces the standard Hand Controller the mount comes with. It goes into the same port the Hand Controller normally does. The ST-4 port is an extra, on which I have my autoguider camera connected.

  19. Very well written tutorial on Astrotortilla. Much Appreciated!

    1. Thank you for your comments. You're very welcome!

  20. Great tutorial thanks, it's threads like this that improve our skills in the steep learning curve of AP, without which I would never improve!!

    I have just been advised about AT for plate solving. From reading the above comments, I see this is a great tool for matching the night sky to your mount accurately with alignment for gotos.

    However, I have a portable mount which I have to set up each night that I want to image the night sky. Unfortunately, in a single night session, I do not have enough time to get enough data to achieve a great final result. I have never managed to match a target from different nights to add data. Would AT be able to plate solve an image taken from a previous session to ensure that when I set up again on another night that I am exactly lined up?

    1. Thank you for your comments, I appreciate your thoughts on the usefulness of this tutorial to your personal experience. Indeed AstroTortilla is one of those "once you use it, you won't go back" programs.

      I too have a portable setup as in Gibraltar, we live in flats compacted into a very busy and small city. I also don't generally have enough imaging time in one night to complete an image, particularly since I image in monochrome. AstroTortilla is perfect for what you want to do. Here's what I do:

      1. Go to my location and set up. Polar align and focus the optical system.
      2. Connect AstroTortilla to my mount and send my mount to the target using Stellarium or Cartes du Ciel.
      3. Let AstroTortilla plate-solve to centre the intended target.
      4. Image!

      The next night, I do exactly the same as above! I do nothing to different to ensure the target is centred because AstroTortilla always seems to get the target perfectly, perfectly centred. You can however use the menu at the top of AstroTortilla to open an image file and have it send your mount to that target (issue a GOTO command based on what is in your image). What it will do is plate-solve your image, calculate the co-ordinates and send the mount there with a GOTO command. You can then plate-solve after the mount stops slewing to ensure the target is centred well.

      In any case, I just do the same thing every single night and it always works perfectly.

    2. Hi Kayron, I have now started to use AT however, although it does plate solve, after the solve it syncs but it does not re-centre my image although I have checked these boxes. Have you heard of this, is this a bug and would you know how I can correct this?.
      Kind regards

    3. Hi Kayron, I have recently upgraded to V7.0 from the previous version of AT. All worked fine with this version and I have not changed any equipment.
      My equipment is an Explore Scientific 102 mm triplet APO, Skywatcher HEQ5 Pro mount and a Canon 650D DSLR. I control my Camera using APT, V2.81. I have also upgraded Cartes Du Ciel to their latest version V3.10.
      I use EQMod to polar align my mount and also align using two to three stars and sync them on CDC. I then focus on a nearby bright star and use CDC to slew to my target. This targets very well usually and is within arc minutes. However, as I image on different nights on the same target, AT is a god send in order to exactly align on the target on each night. This is when I use AT to precisely centre my target before imaging. Now when I go to solve and centre an image (with both sync & re slew check boxes ticked), AT solves in 70-90 secs but then does not re slew to target. Previously, AT used to do this just fine and a small re-centre would happen but this was prior to updating each bit of software to the latest version. I do not use the repeat until within check box but the value is set at 1 arcmin.
      Can you help me please as to why AT does not re slew to my centred target??

    4. Hi Julian,

      I just want to say that I too have had this issue once or twice. What fixed it for me was quite simply to park the mount, turn it off, close AstroTortilla completely, turn the mount back on, run AstroTortilla and connect to my mount, slew to my target and then trying to plate solve again with AstroTortilla. It may very well be a bug within either AstroTortilla or the ASCOM connection itself. Since I have had this issue only once or twice, I have been unable to ascertain as of yet what could be causing it. I hope you're no longer having this problem.

      Best Regards,

    5. Thanks Kayron.
      No unfortunately it seems to be happening all of the time.
      However, with the kind help of folks on the stargazers lounge forum, I have been getting advice as how to maximise my settings to get a quicker solve than I am currently getting.
      I am to try the results of these changes to see what happens and I am hoping that it begins to re centre after solve.
      I will let you know after I have had a chance to test AT the next clear night that I get.
      One thing though, is there a strict order for when to open AT? As I have said above, I usually use EQMod to polar align my mount and also align the mount using two to three stars and sync them on CDC. I then focus on a nearby bright star and use CDC to slew to my target. Then I open AT to plate solve. Should I be opening AT earlier in this process?
      Thank You
      Kindest Regards

    6. Hi Julian,

      I see, ok. Indeed the people at Stargazer's Lounge answering questions on my thread for this tutorial are being helpful as always! :) I actually recommend you open AstroTortilla BEFORE you issue the slew command. It should make no difference in theory, but I've found it best to open it before slewing in order to avoid problems such as this one you've had.

      When I used AstroTortilla, I did it all as follows:

      1. Switch on mount.
      2. Start EQMod and connect it to the mount via ASCOM.
      3. Polar align with EQMod.
      4. Park mount.
      5. Start AstroTortilla and connect it to the mount via ASCOM.
      6. Unpark the mount and issue a slew commnad to my intended target in Stellarium or Cartes du Ciel.
      7. Tell AstroTortilla to plate solve to centre my intended target.
      8. Start autoguiding and imaging!

      I actually think I may have found the source of your issue. Let me explain. In the past when I've slewed to my target and plate solved on it, and then gone elsewhere and plate solved again, it seems that for some reason AstroTortilla wasn't actually centering my mount. What would solve it would be to park the mount, switch it off, close all programs, switch the mount back on, start EQMod and AstroTortilla, slew to my target and plate solve on it. By switching the mount off and closing EQMod, I was effectively deleting all previous alignment data, which made plate solving work again.

      So, my recommendation is to NOT plate solve with the two or three stars in Stellarium / Cartes du Ciel BEFORE going to your target. Firstly, this should prevent the issue of not centering (as has for me in the past). Secondly, you do not need to do a star alignment procedure. Star alignment procedures are just there to make sure your slew commands are more accurate. However, they do not prevent polar alignment error and therefore improve autoguiding performance. That is done with more accurate polar alignments such as use of PHD2's Drift Align tool. Going to your target and plate solving ONLY on your target does the job perfectly and acts as a "one-star alignment", in effect, since AstroTortilla sends a Sync command when it corrects your alignment. It seems to me as if AstroTortilla fails to re-centre your mount when the slew error calculated is very, very small. I could be wrong in the specific reason but it's most probably a bug of some kind.

      So moral of the story - don't bother with the star alignment procedure and slew straight to your target and plate solve there only. After all, if at least your target is perfectly centred, what does the rest of the sky matter? :)

      Best Regards,

    7. Thank You Kayron

      I will try your new procedure and skip the alignment.

      I have copied below the log viewer results towards the end of the solve;

      2015-03-22 22:29:25,299 - astrotortilla - INFO - Solved in 68.1s
      2015-03-22 22:29:25,315 - astrotortilla - INFO - Re-centering...
      2015-03-22 22:29:25,315 - astrotortilla.ASCOMTelescope - INFO - Syncing to Coordinate(169.943368, 13.092713)
      2015-03-22 22:29:25,332 - astrotortilla.ASCOMTelescope - INFO - Sync separation is 1.78'
      2015-03-22 22:29:27,430 - astrotortilla.ASCOMTelescope - WARNING - Sync delta detected: 1.56 arcmin
      2015-03-22 22:29:27,446 - astrotortilla.Main - ERROR - Sync failed: Traceback (most recent call last):
      File "astrotortilla\gui\MainFrame.pyo", line 809, in OnBtnGOButton
      File "astrotortilla\engine.pyo", line 627, in gotoCurrentTarget
      Exception: ASCOM Telescope sync error

      Does this fit in with your thinking?

      Kind regards


    8. Hi Kayron

      I have tried your new set up routine that you have suggested in that strict order. Unfortunately, it still did not recentre following the first solve and ended with the sync error as above.

      I then remembered someone else in this thread had a similar issue but with different software and to solve his issue he had to change Eqmod from Append on Sync to Dialogue based, so I did the same. Hey presto, It worked!!!

      I have copied below the end of the action log to see if you can see why this works from the end of the first solve......

      If you want the full action log let me know.

      2015-03-23 20:21:33,299 - astrotortilla - INFO - Solved in 76.8s
      2015-03-23 20:21:33,299 - astrotortilla - INFO - Re-centering...
      2015-03-23 20:21:33,315 - astrotortilla.ASCOMTelescope - INFO - Syncing to Coordinate(170.239959, 12.928753)
      2015-03-23 20:21:33,315 - astrotortilla.ASCOMTelescope - INFO - Sync separation is 1.46'
      2015-03-23 20:21:34,959 - astrotortilla - INFO - Connecting to camera...
      2015-03-23 20:21:34,973 - astrotortilla - INFO - Exposing: 6.00 seconds
      2015-03-23 20:21:35,270 - astrotortilla - INFO - Exposing: 5.9 seconds
      2015-03-23 20:21:35,723 - astrotortilla - INFO - Exposing: 5.5 seconds
      2015-03-23 20:21:36,223 - astrotortilla - INFO - Exposing: 5.0 seconds
      2015-03-23 20:21:36,740 - astrotortilla - INFO - Exposing: 4.5 seconds
      2015-03-23 20:21:37,224 - astrotortilla - INFO - Exposing: 4.0 seconds
      2015-03-23 20:21:37,740 - astrotortilla - INFO - Exposing: 3.5 seconds
      2015-03-23 20:21:38,194 - astrotortilla - INFO - Exposing: 3.0 seconds
      2015-03-23 20:21:38,709 - astrotortilla - INFO - Exposing: 2.5 seconds
      2015-03-23 20:21:39,224 - astrotortilla - INFO - Exposing: 2.0 seconds
      2015-03-23 20:21:39,694 - astrotortilla - INFO - Exposing: 1.5 seconds
      2015-03-23 20:21:40,256 - astrotortilla - INFO - Exposing: 1.0 seconds
      2015-03-23 20:21:40,726 - astrotortilla - INFO - Exposing: 0.5 seconds

      Hopefully this is the answer but don't know why.

      Thank you for all of your time and patience, it is very much appreciated.

      Kindest Regards


    9. Hi Julian,

      Thank you for your updates on this. It's great to hear you fixed it and it's nice of you to share the fix (others experience this as well, I'm sure). When EQMod is in "Append on Sync" mode, it adds a star alignment point when a Sync command is issued. This essentially acts as a star alignment procedure and so using AstroTortilla itself effectively works as star alignment, even if you just do it with your target. Switching to "Dialog Based" means that Sync commands no longer work to add star alignment points to EQMod, so you've disabled that mode of star alignment.

      As I mentioned above in another comment, this poses no disadvantage whatsoever. In one night, you'll slew to your target, miss and use AstroTortilla to centre your camera on your target. That's it - you then guide and image away. No need for star alignment at any point along the process. Even if you do multiple targets in one night, star alignment is not needed if you're plate solving on each target you choose.

      I bet you're happy this will keep working well now! :) AstroTortilla is brilliant.

      Best Regards,

    10. Kayron, YES a fantastic piece of software, thanks for your help. I just need to sort out guiding now, that's another dark art!!

      Kind Regards


  21. Hi Kayron,

    I am an amateur who took up astronomy just over an year ago and I haven't looked back. In this time I have looked up many tutorials trying to understand various concepts, tools, and processes. I came across your tutorial while looking up tools for polar alignment.

    Your tutorial is very well written and thanks ever so much for taking time to do this. Sharing of knowledge is meritorious. Thanks also for taking time to respond all the questions. I have no questions at the minute other than to ask What is plate solving and can it help with accurate polar alignment ? and in the era of tools like Astrotortilla and auto guiding is just a rough polar alignment enough ?

    Thanks and best wishes

    1. Hi Nalin,

      My thanks for your comments on my tutorials - I am happy to hear they have helped you. I will answer your questions.

      Unfortunately there is no substitute for a good polar alignment. The reason for this is that when you polar align, to ensure the mount is physically rotating around the axis of rotation of the Earth. In an equatorial mount, this effectively cancels out field rotation - the apparent rotation of objects in the night sky due to the Earth's rotation. When you're not properly polar aligned, you are rotating off-axis and then field rotation becomes apparent over several minutes of exposure time, though of course, the longer the imaging focal length, the faster the onset of field rotation in your images. Ultimately it depends how far off-axis you are when polar aligning.

      Autoguiding cannot correct for this because autoguiding will not change the alignment to the axis of rotation of the Earth. This is only changed by the azimuth and altitude bolts on your mount (used when polar aligning). AstroTortilla will also do nothing for you in this regard because all it does is plate solve to centre objects in your field of view. Even when you're really well polar aligned, when you issue a Goto command to go to an object, you will inevitably be off even if just a bit. AstroTortilla's plate solving captures an exposure, plate solves it to figure out precisely where it is you are aiming, calculates the error offset and then issues a slew command to the mount to correct for the offset. This effectively perfectly centres your target in your camera's field of view. The alignment to the axis of rotation of the Earth is however not altered, as your azimuth and altitude bolts are not tweaked (this can only be done by hand).

      You may enjoy reading up on the need for a proper polar alignment for long exposures, here:


      Drift alignment is the gold standard of polar alignment, and can be done effectively in 10 minutes or less with PHD2 and an autoguiding camera. No matter how accurate you try to be with your polarscope and EQMod, for example, polar alignment will be off, sometimes by a significant margin. Enough to cause field rotation in exposures of about 10 minutes even on very short imaging focal lengths of 360mm, for example. A drift alignment uses quick, direct feedback to allow you to tweak the azimuth and altitude bolts of a mount in order to get the Declination drift to be as close to null as possible, providing a fantastic polar alignment. After a 10 minute procedure with PHD2, I managed to capture a 30 minute exposure at 360mm imaging focal length with perfectly round stars. That's how successful it can be.

      Drift aligning in PHD2 can correct for a decent or bad polar alignment. The difference is the initial error but ultimately you can polar align really, really well without looking through the polarscope or even at Polaris, by drift aligning. I normally polar align decently well with my polarscope and EQMod and then open up PHD2 to perform a good drift alignment. Once done, I proceed with imaging. Stay tuned though, I will be writing a tutorial on drift aligning in PHD2 sooner or later. Do note however that it isn't very hard to work out how to do!

      Best Regards,

    2. Hi Kayron. I also use PHD2 but not for bettering my PA yet which I do thru' EQMod. Have you done this tutorial yet re drift alignment thru' PHD2?? Would love to view it when you do, please.

    3. Hi Julian,

      My apologies - I forgot about this comment and just saw the e-mail notifying me! I have not yet worked on the PHD2 tutorial. My intention is to have a big PHD2 video tutorial (seeing as I'm doing video tutorials for the time being) and have it included the drift alignment procedure within. However, to do this I need to capture some video recordings of me physically carrying out the procedure outdoors with my own equipment. Given the current weather, I am unable to go out to do this. I will set up my astrophotography laptop to capture video of what I do so that I can then use the footage with commentary as the video tutorial.

      Best Regards,

    4. Hi Kayron,

      This was an excellent read, and I thank you for taking the time to be so detailed. I am just getting beginning to learn AstroTilla, as my setup is also portable, to escape the city lights and go to my club’s dark site. My imaging setup is as follows; Celestron CGEM 8” HD SCT w/ Nightscape 8300 CCD (w/ 0.7 FR) and my modded Meade ETX-90 w/ QHY5 (and 0.63 RF) for auto-guiding. I made the jump to one-shot colour CCD last year after having spent 2-3 years using a Canon DSLR (1000D) and BackyardEOS. I have experience using Nebulosity 3 and PHD2, and plate solving seems like a very attractive option to spend much less time performing setup when I could be imaging 

      I have a few general questions, and I would greatly appreciate if you or others could chime in with your feedback – I find myself continually learning, and it’s through exceptional users/blogs like these that I learn the most!

      1) I note that AstroTilla appears to require EQMod – is this correct? From what I understand for (Celestron) CGEM mounts, EQMod will not work. Is there an alternative software or fix to make EQMod control the mount, considering it has ASCOM drivers.

      2) Related to my first question, is the general setup routine. What has frustrated me is trying to perform alignment and polar alignment using the Nightscape imager. I note from your discussion about how you setup, you balance the mount and power-up, and immediately start polar alignment. Do I need to do a two-star alignment first, or can I simply just power up and start a polar alignment procedure? Does it matter if I use the hand controller polar alignment routine, or can I use software (such as EQMod) instead?

      3) Do you have any other tips for setup? Once everything is balanced, and I have done a rough polar alignment (PA) with my PA scope, and I power up – again related to my second question – can I start a polar alignment procedure, and then park the scope, and then begin plate solving procedure?

      Overall I have been frustrated trying to perform alignment/setup using the Nightscape imager, since I have to take 5-10 second exposures, and then make very fine adjustments to the Alt/Az knobs. I used to do this in BackyardEOS quite simply, because I could use LiveView with my DSLR. So how do others do this fine alignment using their CCDs?

      Again please excuse my ignorance, any details others could offer would help!

      Clear skies,

    5. Hi Tim,

      Thank you for your comments on the tutorial. I'm very happy to hear you have found it useful thus far. Your setup sounds pretty heavy, but no doubt your Celestron CGEM mount can take it! :) I will answer your questions as best I can in the order you asked:

      1. AstroTortilla requires that it connects to your mount via ASCOM. EQMod is one way to connect the mount via ASCOM and have a virtual hand controller on screen, of sorts. Some do not use EQMod at all and instead have software such as MaxIm DL connect to their mounts. I would assume AstroTortilla can work with any such connection as long as it is able to tell where the mount is currently pointing and provide it with slew instructions to correct the calculated Goto error.

      I've not worked with Celestron mounts myself. In theory, if you have ASCOM drivers for the mount and hook up the mount to your computer (even if through the physical hand controller, set to PC mode), then you can select the driver from AstroTortilla and connect to it (as long as the mount is switched on). This should work as well as someone using EQMod with a Skywatcher/Orion based mount.

      2. A star alignment procedure is done for one reason only - to make Goto commands more accurate. In other words, so that when you slew to an intended target, that it more precisely falls in the centre of your FOV. If however you are using AstroTortilla (or similar) to centre the target in your FOV with plate solving, star alignment procedures are a pointless waste of your time and effort. At the end of the day, you won't care if your mount hits the target 30% better because your performed a star alignment procedure. 30% better may not be 100% dead-on centre. Plate solving will get it pretty much dead-on centre, no matter your initial error. Therefore, balance everything, power up, polar align, slew to your target, plate solve on it and start imaging (and autoguiding)! :)

      It doesn't matter if you use the polar alignment routines provided by your hand controller, or software's. I prefer using EQMod's because I don't have a hand controller for my Avalon M-Uno (I bought it without it, knowing I wouldn't use it) and even if I did, I wouldn't want to bother connecting it and then disconnecting it. Polar alignment routines provided by hand controllers these days can be pretty good indeed. Nothing will get you closer to a perfect polar alignment than a drift alignment though, and for that PHD2's Drift Align tool is golden.

    6. 3. Indeed, I would set up my telescope with ALL the equipment I intend to have on it, roughly on the focal point (you can estimate it if you don't already know more or less where). This means your telescope is now producing a realistic moment on your mount, and so balancing it in RA and DEC now will be key to the best tracking when you're imaging. Get it balanced well. Once you're satisfied, lock it back up in the home position and power up. Perform your favourite polar alignment routine and park it again at the home position. Now you're ready to start. Select a target, send your mount there with a slew command and initiate a plate solve. This will centre your target precisely on your camera's FOV. Once it's done, start your autoguiding and then start your imaging run.

      When you refer to being frustrated with performing alignment with the Nightscape Imager, do you mean polar alignment or star alignment? I assume you mean polar alignment since you're touching the mount's Alt/Az knobs. Does your hand controller not have a nicer polar alignment routine that uses a polarscope in the mount? Usually this suffices if the polarscope is well aligned to the mount's axis and your GPS location and date/time are entered correctly. Like I said though, it sounds like maybe a PHD2 drift align might be what you want to go for. It's actually quite simple to perform but seems difficult at first. It's done purely with PHD2 and your autoguiding camera, using two stars. It gets your polar alignment down more precisely than you'll ever manage with your polarscope and related routines, plus you don't even need to see Polaris to do it!

      Once the weather clears, I want to capture some footage of me performing it on my astrophotography laptop so I can compose a video tutorial on PHD2 drift alignment. However, there is already some information online and nothing beats running it and playing with it. Do feel free to ask me for help if you get stuck though. You can always e-mail me at: kayron.mercieca@gibastrosoc.org

      Please don't worry about feeling like you stuck with something - astrophotography has a very steep learning curve but once you fill the small gaps with experience, you'll be right up there with the top imagers! :)

      Best Regards,

  22. I have been trying to use astrotortilla and follow your tutorial but I keep getting the following message 0 [main] bash 3632 find_fast_cwd: WARNING: Couldn't compute FAST_CWD pointer. Please report this problem to

    2015-06-05 00:27:58,490 - astrotortilla - INFO - the public mailing list cygwin@cygwin.com

    I use it with backyardEOS to control my Nikon Camera and PHD2 with an Ioptron IEq45 mount for guiding with a magnificent mini from orion

    1. Hi Lunam,

      My apologies for the late reply. I don't seem to have received an e-mail for your comment so I didn't realise until now! Regarding the error you are getting, I hope you have been able to fix this. It sounds to me as if AstroTortilla cannot properly communicate with the cygwin installation that handles the plate solving in the background. I'm not 100% sure how to diagnose the problem myself as I have not written the software. I would suggest contacting "Antti Kuntsi", the software developer for AstroTortilla. He commented above (top comment) or you can probably contact him through the AstroTortilla website.

      Best Regards,

  23. Your presentation is very clear. I am considering installing it. I would like to know if it is possible to use AT simply to solve a plate taken a night before just for coordinates and orientation and without any connection to another program or to a telescope. Thanks. Jean-Paul Richard.

    1. Hi Jean-Paul,

      Thank you for your comments. I'm pretty certain the Log Viewer (found under the Tools menu) will show you the coordinates of the plate solved image. What I would do if you want to check these out without having a physical mount connected is use the ASCOM Simulator. Run EQMod with the ASCOM Simulator, which acts like a simulated mount for all intents and purposes. Unpark it in EQMod and then go to the menu Tools -> Goto Image in AstroTortilla. It will ask you for an image file. Point it to your desired image and it will plate solve it. Once it does so, it will send your simulated mount to that location, but in the process, you should see the plate solved coordinates in the Log Viewer. if they do not appear in the Log Viewer, they will appear in the Telescope section under "Current" once your simulated mount stops slewing to the image location. That is your best bet with AstroTortilla.

      If however you have an Internet connection when you're trying to plate solve an image, simply go here:


      Upload the image and it will plate solve in seconds. Once it says "Success" on the top, go to the details page linked and on the right side is a lot of information including the image coordinates. I hope this helps! :)

      Best Regards,

    2. Kayron: thank you very much for a very precise answer on AstroTortilla and for the suggestion on nova.astrometry.net.
      I will give these a try. Jean-Paul Richard.

  24. I followed your guide to the end. Very nicely written.

    One quick question, because I haven't tried it out on the field yet, I'm doing it exact as you say using the Simulation tool (Not the EQMod one, but another simulation which should work the same).

    The step afterwards where you launch Stellarium and issue a GoTo command, how were you able to do that and get the data viewed on AT under Current and Target (using the Simulator, or possibly the mount itself)..

    When I launch Stellarium and played with the different settings of the Telescope Plugin, I couldn't find a way to issue a GoTo command and see it resemble on the AT screen.

    I did try EQMod Simulator as well but 2 separate simulators opens up when I used Stellarium in conjunction with StellariumScope..

    Just getting some insight on this before I take my stuff outside. :)

    Thank you!

    1. Hi Rafa,

      Indeed for this tutorial, I used the EQMod Simulator to capture the screenshots and indeed once I issued a GoTo command in Stellarium (with StellariumScope), AstroTortilla would show me the Current and Target co-ordinates of the EQMod Simulator. The way to use the EQMod Simulator is to open a session with it first. This involves running the "Toolbox" shortcut for EQMod, selecting the Simulator from the list and clicking "ASCOM Connect". This will open up the session. Selecting the EQMod Simulator from the list of ASCOM devices in StellariumScope and AstroTortilla will then automatically connect to the currently active session. This is identical to using a real mount on EQMod.

      I hope that if this doesn't help, that you have solved the issue by now! :) Perhaps you've already been out with your mount and answered your own question. Sorry for the delay in replying!

      Best Regards,

  25. I get to plug the combo: Eqmod, Ciel, BackyardEOS and AstroTortilla
    But I'm confused by the default value in Eqmod "Epoch Unknown"

    So i've managed to set all software to J2000 > but i just don't know what i'm doing :(

    Anyway, plate solving on my recent image is working fine and the AT goto image drive my scope on target with the Ascom Simulator and confirmed that in Ciel.

    But i cannot get the re-slew command working as expected, AT said it move, show briefly current value sliding before step back to initial wrong target/current position, Ciel is simply not moving the scope in the re-slew process.
    Same behavior with both Dialog and Append method inside Eqmod.

    Any hints for J2000/JNOW ?
    Did you manage to get the re-slew function working with the simulator ?

    Best Regards,

    1. For re-slew, as i understand, please correct me if i'm wrong.

      re-slew is working only after choosing initial target (pointing with errors)
      then AT plate solve, and compute the delta between the expected target and the solved coordinates
      after that AT make a sync to the current coordinates (plate solved) then issue a re-slew using the preview computed delta.

      When i'm playing with the simulator, i'm in initial wrong position and provide an image that solve to the real target coordinates
      so the delta is the same, but target and current coordinates are flipped and produce something not expected by design...

      So re-slew will work on the field as far as my AT goto image is working ?

    2. Hi Hugues,

      You are correct. The re-slew option applies to imaging in the field with real data, because in the field, you simply send a slew command to your mount via your favourite software and let it go to your target as best it can, then AstroTortilla will:

      1. Capture an exposure of what you're pointing at.
      2. Work out precisely what coordinates you're pointing at.
      3. Calculate the difference between your actual coordinates and your mount's "reported coordinates" (the ones related to your initial slew command).
      4. Sync this difference with the mount so it acts as a "star alignment point" and makes future slew commands more accurate.
      5. Re-slew your mount by the difference between coordinates so that it falls directly on target as intended.

      With the simulator, the only real way to do this (to fool the design), I suspect, is to slew the simulator to a target and then provide AstroTortilla with a pre-captured exposure that's actually slightly out from your simulator's slew target. This will make AstroTortilla carry out the same job as intended, but with the simulator.

      My strong recommendation is that you set up AstroTortilla as per this tutorial and then go out in the field to test it out. You should be happy with results if you followed the tutorial closely, as it has been tested by thousands and thousands (if my tutorial's read count is anything to go by! :)). I myself have used these settings before and several friends of mine do to this very day.

      As for the epoch settings, I would opt for J2000 for EQMod's setup, and any other software that asks. The reason I have everything set up as J2000 is because of an issue I once had with Sequence Generator Pro, which on my setup uses the same Astrometry.net offline service as AstroTortilla to plate solve. It was plate solving fine but I was always slightly off-target. It ended up being that one thing was in J2000 and the other in JNow. I switch both to J2000 and voila - perfect plate solving. I suspect switching both to JNow would have also solved the issue as it was a coordinates reference mismatch, no doubt.

      I hope this helps you out in some way! :)

      P.S. Thanks for the tip on upgrading CygWin. I will look this up for when this tutorial receives an update, providing AstroTortilla hasn't been updated in of itself at that point.

      Best Regards,

  26. FYI
    On windows 8.1, you must upgrade the CygWin version packaged with AT to avoid the annoying "find_fast_cwd" warning that plague your logger.
    At least upgrade the Base/findutils package and dependencies.

  27. Hello Kayron,

    Thank you for your valuable answers. I will leave my setup in J2000.

    As i'm an OSX user It was the main reason i switch from Eqmac to Eqmod in virtual machine.
    (Eqmac is in J2000 and the Stellarium interface sync is done in JNOW wich lead to invalid precession of slew coordinates)
    The same things that you have experienced with SGP.


    1. Hi HoxCa,

      I'm not 100% sure if I should call you Hugues or HoxCa, so please excuse me if I got your name wrong! :)

      I tend to favour having everything in J2000 and that's it. I'm certain that given the option, if everything is set to JNow, it'll work just as well though, since the coordinate references would agree between software. Nevertheless, most of my software appears to use J2000 by default so I make sure everything I use is set to J2000.

      Indeed that is what I experienced with SGP. Everything was in J2000 but I accidentally left SGP's Plate Solver at JNow, giving me a slight mismatch in coordinate reference. The plate solves worked well but the target was always half-way across the images to the top-left, until I realised what was happening! :)

      Best Regards,