Alternative Polar Alignment using PHD2

[Ken Self]

I'm developing up an application that lets one use PHD2 in a similar way to a Polemaster to polar align ones mount.

The application uses the PHD2 interface. The current incarnation has a browser client using Javascript communicating with a Python webserver.

The general idea is to find where the RA axis is pointing by rotating the mount in RA and recording the position of a guide star at three points on the arc. The three points allow the centre of rotation to be calculated which corresponds to the alignment of the RA Axis.

The mount is then adjusted using the bolts in the usual way so that the RA Axis points at the Pole.

Using PHD2 has the advantage of precise measurement of the star position which in turn allows relatively small movements of the mount to caculate the centre of rotation.

My testing so far is encouraging. I've been able to largely automate the process of rotating the mount and recording star positions.

It works as follows:

Point the scope roughly at the pole by whatever means (platesolving, star hopping, star alignment, sheer luck etc...). An overlay showing the circumpolar stars could be provided to assist.

If completely lost, move the mount to its home position. Platesolve and calculate Alt/az.

Adjust mount accordingly.

Pick a guide star (optional - the app can do this with a call to find_star)

The app records the current star position.

The app assumes a mis-alignment of 5 degrees and calculates how far it needs to rotate to detect that the star is following an arc.

To keep the guide star within the search radius, the app calculates how far the mount can rotate and keep the star within the search radius. It then splits the total movement up into steps of that size.

After each step it calls get_lock_position and get_star_image to determine the star position then sets the lock position to the new star position.

In order to move the mount I connect to EQMOD directly. To do this I have to massage the ASCOM driver name returned by PHD2. I issue an ASCOM SlewToCoordinates to move the mount.

After the initial movement it recalculates the mis-alignment and calculates how much further it needs to move the mount. The assumption is that the mis-alignment is less than 5 degrees. Smaller mis-alignments of the mount require more rotation to detect movement.

After that movement it records the second position.

For the third position is moves the mount the same amount it moved (in total) for the second position.

With three star positions it is relatively simple to calculate the centre of rotation.

The vector from the centre of the image to the centre of rotation can be resolved into two components: a declination error component and a cone error component. The cone error component cannot be adjusted out. The declination component can in turn be resolved into altitude and azimuth adjustments; given the RA of the mount and local sidereal time.

Once the centre of rotation is well within the field of view it is possible to display the positions of the circumpolar stars around the centre of rotation to aid alignment.

A couple of questions for the group:

Is it worth me trying to incorporate the application into PHD2 itself or just leave it as a standalone app? If so I might need a little help. I'm reasonably proficient in C++ but a bit rusty.

If standalone, would it be possible to issue guide commands to the mount via the PHD2 interface rather than direct to ASCOM. 

Also, would it be possible to get the calibration data (specifically the angles) via the interface. I will need this to calculate the required alt/az adjustments neeeded. Currently I am just entering the values manually.

[Andy Galasso]

Ken,

That sounds like a really cool project.

We already have an open request to add methods to the server api to issue guide pulses, but I'd be glad to bump up the priority of that so you can use it in your app.  We can also easily add methods to access the calibration information.  Let me take care of those and I'll let you know when they are available. 

Andy

[CyScape]

I think that incorporating it into PHD is a great idea. I personally don't like PHD'S drift align (no offense). It never, ever really seems to settle for me, making it take way longer than necessary. Buying another camera (polemaster) is way to expensive for a one use camera IMO. Being able to use my QHY5L-II to polar align would be a great benefit as well as making PHD even more robust with useful features.

Thank you for your efforts.

[mahaffm]

Hi Ken,

I like this idea as well. I can see it working with equipment utilizing guide scopes but will it work for OAG's, I'm thinking the FOV would be too small when using OAG's? 

Thanks,
Mark

[CyScape]

I was going to mention that as well, I also use an OAG?

[Brian Valente]

I love the idea as well, not sure how it will work for longer scopes like my 12" RC and I'm using an ONAG for guiding. I don't know if it has the fov to do that well. the automated idea is great

For me, polemaster is one of the best investments I ever made. I have a mobile setup, and I get out and image far more often because I'm not dreading having to polar align each night

Thanks

Brian


Brian Valente
Brianvalentephotography.com

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Ken Self]

I've been using a prototype with my RC8 at 1600mm FL with my imaging cam. The FOV is 38'x29' and it works well. A couple of caveats : a) the imaging cam is an ASI1600 so it has a good sized sensor and I run it at 4x binning in PHD2 so it doesn't get bogged down with downloading the images.(b) I'm in the southern hemisphere and I have three fairly bright stars inside the FOV near the pole: BQOct (mag 7), HIP48752 (mag 8) and TYC9518-405-1 (mag 9) whereas the stars close to the NCP are a bit dimmer HIP3128 (mag 8) and a few others around mag 9 or mag 10. I have also used it with my guidecam and a 400mm ST80 where the FOV is actually smaller. With an OAG its possible a DSLR imaging cam may not work well and the northern stars may be a bit dim for a guidecam through an OAG. Maybe its something you could check out for me by pointing at the NCP and see if you can make out guideable stars. One is sufficient but three is better. With one the algorithm needs to work out where the star needs to be to display on the screen. With three one can place the three stars on concentric circles marked around the pole.

[Brian Valente]

Ken is the 1600 your main camera or your guide camera?

 

One thing I just realized is we’re not limited to the guide camera, we could switch to main camera for PA. that would solve the OAG

 

Thanks

 

Brian

 

 

Brian Valente

Brianvalentephotography.com

 

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of Ken Self

Sent: Tuesday, April 11, 2017 1:58 PM
To: Open PHD Guiding <open-phd...@googlegroups.com>

--

[mahaffm]

Hi Ken,

Yes, that is a good point to be able to use the imaging camera for the PA would certainly help improve the FOV over using the OAG. Not sure if my scope/imaging cameras would work. HD11 with the ASI071 FOV is 29'x19.3'? The FOV of the C8 with the ATIK383+ is 30.7'x22.9'? the ATIK383+ can bin 4x4 however the ASI071 can only go to 2x2 binning. But the ASI071 is CMOS so it has adjustable gain and offset that might help.

Cool project Ken, I think you might be on to something.

Thanks,

Mark

[Andy Galasso]

On Mon, Apr 10, 2017 at 8:59 PM, Andy Galasso <andy.g...@gmail.com> wrote:

We already have an open request to add methods to the server api to issue guide pulses, but I'd be glad to bump up the priority of that so you can use it in your app.  We can also easily add methods to access the calibration information.  Let me take care of those and I'll let you know when they are available. 

Ken,

Ok, two new methods, guide_pulse and get_calibration_data, have been added to the server API.

Andy

[Ken Self]

Wow! That was quick. Thanks Andy, I'll incorporate those into the code and see how they go.

[Ken Self]

The 1600 is my imaging cam. The guide cam is an ASI120MM. I like using the imaging cam for polar alignment. When I get the alignment right I save the home position so the scope then points where the mount is pointing..

[Ken Self]

Is there a dev build with these changes? The latest I can find is dev3 which pre-dates the server changes

[Ken Self]

Some screen shots showing how it works.

One is the actual page showing the calculated arc and the points used to show how little movement is needed. The Blue and red lines show the offset of the centre of rotation from the centre of the sensor. The blue line is a declination offset and red is a cone error offset. I'm currently working on resolving the dec offset into alt and az adjustments.

The second shot is an overlay of the above on top of a PHD2 screen shot where I rotated the mount over a longish expsoure to show that the calculated centre of rotation is close to the actual

[Andy Galasso]

On Fri, Apr 14, 2017 at 7:28 PM, Ken Self <ken.k...@gmail.com> wrote:

Is there a dev build with these changes?

No, not yet. You can either build PHD2, or, if you are not setup for that, send me an email and I will send you an exe.

Andy

[Ken Self]

I'll have a go at building it first

[Ken Self]

An update on my progress so far:

I was able to buld PHD2 with the new methods. I've only been able to test get_calibration due to the weather as I can use a simulator for that.

I've worked on the algorithm based on the comments so far. I realised that if one selects a known star (e.g. Polaris) then the pole does not need to be in the FOV. We can rotate the mount and record three positions for the guidestar and from that calculate the position of the RA axis in pxels even if it is off screen. Using the mount's RA and LST we can then work out where the guidestar ought to be if the RA axis were aligned with the pole, and show the required adjustments in Alt and Az to move it there.

Given that the mounts RA may not be accurate, e.g. if not polar aligned; the circular paths of several near polar stars are displayed on screen. Adjusting the mount to place three stars on their respective paths aligns the mount to the pole.

So far testing has been with simulation while I wait for a reasonably clear night

[Jay Tee]

Hi Ken,

Will your method account for a misalignment between the RA axis and the optical axis? One of my Imaging setups is a side by side where the scopes may or may not be perfectly aligned with the RA axis.

Cheers,
JT

[Ken Self]

At astronomical distances, any lateral offset is immaterial. Just think about the distance from your OTA to where the polar scope goes. Cone error is another story. I found that the offset of the RA axis from the cenre of the sensor can be broken down into a declination component and a cone error component. The declination component can be eliminated by simply slewing in declination. The cone error can only be eliminated by adjusting the OTA e.g with shims. No amount of adjusting RA, dec, alt or az will make a difference. So one thing I need to do is to account for the cone error when showing what alt and az adjustments to make. But this is only needed for substantial cone error.

[Ken Self]

Another quick update. Endless clouds and a mount rebuild have kept me quiet. I've now written the procedure into PHD2 - my local copy only right now. I think I've done it reasonably non-intrusively.

Last night I had a brief chance to test it out and found a couple of issues. One is that I lost the guide star and the other is that I need to find a way to refresh the display for visual feedback during the procedure.

The first problem  should be easy to fix by trial and error (had the same thing on previous standalone versions) when I have clear enough skies to test it.

I did have one other "bug". During development I was using ASCOM simulators so I had to hard-wire the star coordinates to trace an arc. During the first test all went well but the graphics all ended up centred on the fake position.

[Ken Self]

I've finally made some progress after having to rewrite chunks of code to integrate properly with PHD2. My earlier code did not allow the alignment star to be tracked whilst polar aligning so it failed miserably. The additional module is still fairly non-intrusive, I believe. It works in simulation but we have not had clear skies for months now for field testing. So third clear night I'll start testing for real - first two are reserved to get over imaging withdrawal. 

Here's a sneak preview using the simulators:

I told the tool via the dropdown list that I'm aligning using BQ Oct which I selected manually on the dispay. By rotating in RA it reads the position of my star at three points (blue circles - for the simulation I had to manually select the star at each point as the simulated image does not move as the mount rotates).

It calculates the circle that passes through those three points (dotted magenta).

Knowing the coords of BQ Oct it draws a green circle where it should have tracked had the centre of rotation (red cross) been aligned perfectly with the pole. A smaller green circle is a calculation of where it should be on that circle based on LST. The orange and green (bad choice) sides of the triangle are the altitude and azimuth corrections needed to place BQ Oct where it ought to be for polar alignment. The yellow circles are the arcs of the alternative alignment stars with small circles where they ought to be. Placing the three stars on the respective large circles should also result in proper polar alignment even if the LST calculation is incorrect. The red and blue triangle shows the cone and dec adjustments needed to put the centre of rotation in the centre of the sensor (grey cross) for those with OCD :)

The field test is needed to check that the alignment star can be tracked as the mount slews in RA. Slewing is needed as it would take an eternity to rotate enough with guide pulses. I break down the slew into steps that should keep the star within the search radius. At present I leave no settling time so that may be needed.

I'm confident it will work in the soutehrn hemisphere as there are several useable alignment stars near the pole that would be within the FOV for most guiding setups. The northern hemisphere is not so good. After Polaris theres not much else and Polaris requires a FOV of at least 45 arcmin. I've identified a couple of stars from charts but have no idea how visible they are.

[Ken Self]

Had to abandon my imaging session due to light cloud so I stared field testing the tool. It went very well and the star got tracked through its mvoement - except near the end due to a minor code error. Thats now fixed but heavy cloud came over before I could retest. I think my graphics need fixing as it looked wrong when I binned the camera. Plus I think the orientation of the star positions is wrong also. One outcome of the test is that I think I can reduce the amount of rotation eeded to calculate the centre of rotation.

[Jim Hunt]

Hi Ken,

I'm really interested in you your project and have some thoughts and info which may (or may not!) be useful.

The other night I just happened to be playing with taking images of Polaris through my guide scope. My results may help you with your thought "After Polaris theres not much else and Polaris requires a FOV of at least 45 arcmin. I've identified a couple of stars from charts but have no idea how visible they are"

My guiding setup is an Altair GP-CAM IMX224C combined with a 50mm scope. The scope has an FL of 182mmm which provides an FOV of 90.52 x 67.89 arc mins.

I have attached some screen shots of the results. The image used was taken using a 2 sec exposure, the usual exposure I use with PHD2.  Unfortunately I didn't capture a PHD2's view, so I have also stretched the APT image to something akin to that which PHD2 might present.

The screen shots I have provided are;

 - "APT - Platesolved - as taken.JPG" > the image as taken by APT,

 - "APT - Platesolved - stretched.JPG" > an intermediate stretch of the image, hopefully providing something akin to what PHD2 would 

display

 - "APT - Platesolved - fully stretched.JPG" > a full stretch just to show as much detail as possible 

 - "CdC - Platesolve result.jpg" > the result of the Platesolve as sent to CdC by APT.

As you say there aren't many stars around, other than Polaris, which are readily identifiable. In my image there are really only two  other than Polaris, HD14369 and HD5914. Not the most "well known" stars (I only know of them because of CdC!). For a novice like me, I think selecting an appropriate star from a dropdown list and then identyfing it on the PHD display could be more than a little tricky - especially if I've do as you suggest i.e. "...point scope roughly at the pole star...". My only hope would to be point the scope pretty precisely at the pole star in the first place, then take an image, platesolve it and then identify in CdC what I am looking at. Not the complexity you intend I think!

I have tried thinking of some solutions to the problems posed, but I confess I haven't really come up with anything yet - except to only ever use Polaris as the alignment star.

Hope this helps!

Cheers, Jim

[JST200]

P.S. It's just occurred to me it would probably be helpful if I also gave you the original image taken by APT! It's too big to upload, so it's here:

https://www.dropbox.com/s/tdfnv65zd5m1lx1/Single__Bin1x1_2017-09-14_21-00-02.fit?dl=0

Cheers, Jim

[Ken Self]

Thank you Jim! Thats an enormous help. 

And you are right, I am trying to avoid the complexity in plate solving. Maybe a popup aid that shows the main stars near Polaris as a guide would help.

Its kind of ironic that in the ays of polarscopes I used to curse how "northerners" have a handy, bright pole star while we antipodeans had to rely on Sigma Octans which is very hard to find. But once you zoom in closer to the pole we have an easily identifiable asterism, several in fact for different FOVs, and a pair of stars within a couple of arcminutes of the pole.

Let me work on this conundrum a bit and see what I can come up with. Worst case is that Polaris is the only choice but that pretty much excludes use by anyone with a FOV less than about 1 degree.

[JST200]

Hi Ken,

Glad to help! Although I'm sorry it hasn't moved you forward that much.

This morning I thought I had an epiphany moment about how to resolve the problem - just use 2 random visible stars, calculate their centres, etc. Then realised it was nonsense! :-)

However, I do keep wondering if a plateSolve might help somehow. Since finding out about platesolving through the Astrophotography Tool - APT, I have become a convert. To do the equivalent to a 3 star alignment by just imaging a random part of the sky, platesolve it and issue a synch, as many times as you like is bliss. And far more accurate than I could achieve by eye.

APT integrates with PlateSolve2. Even on my old laptop a platesolve takes about 6sec. If you could pre-requisite Platesolve2, then couldn't it's results provide you within the info you need?

However, I guess this goes against you complexity criteria.

Anyway, I'll keep thinking.

Cheers, Jim

[JST200]

BTW Apilogies about the odd incorrect word/spelling and punctuation. Sometimes auto correct on a phone is a royal pain in the...

:-)

[Ken Self]

Plate-solving is useful for identifying stars and for an initial alignment, Back when I was prototyping this idea I was using platesolving from a camera attached to the polarscope to do just that. SharpCap has its own in-built platesolving polar alignment but it requires two measurements 90 degrees apart. My aim is to have to slew the mount no more than about 10 degrees using the accuracy of the centroid algorithm in PHD2. Its certainly an option as the ANSVR tool was developed by on A Galasso :) I even looked at incorporating a simplified platesolve algorithm like Sharp Cap does but its code is closed and the open astrometry.net code was phenomenally complex

Having looked closely at your images I think the best approach is a human plate solve.Your image showed all the stars that were options for alignment and many more. They were not at all faint and I could see some useful asterisms that would help someone trying to get their orientation.

I think I can simply provide a template image of the stars around the pole, oriented according to the position of the mount and camera.The stars need only be labelled A, B, C, D, E - except maybe Polaris. The user would match their on screen image to the template and choose an alignment star of their liking from the dropdown. It does require that they be pointing in roughly the right direction but that can be done via a star alignment or platesolving in the normal way before using this tool.

Right now I'm just mulling over whether to have a popup display or use the PHD2 display for the template. I'm leaning towards the popup and maybe even have an option to fine tune its rotation to make matching easier

[JST200]

We'd already agreed you wanted to avoid Platesolving. I don't know what I was thinking! :-)

I think in my arsenal of images I might have some shots of Polaris, other than the one I've already given you. They probably won't be taken through my Guidescope. They are more likely through my main scope with my EOS 1100d, so the FOV will be different. However, they might give you a better idea of the stars available around Polaris that don't show up on CdC, etc.

If the clouds disappeared I'd specifically take them for you through my Guidescope. But that seems unlikely!

I'll be in touch.

Cheers, Jim

[Ken Self]

it was worth looking for a plate solving solution but it would have tobe simple.Anyway I'm well underway on showing a picture of the star positions.

What I'm particularly interested in image-wise is the stars on the other side o the pole from Polaris but close to the pole. 

[JST200]

Hi Ken,

I've just checked and none of the images I have are going to help. Sorry.

I'd forgotten that it just so happens that the FOV of my 1100D & Skywatcher 150mm is almost exactly the same as my IMX224C & 50mm guidescope. So all the images I have are of Polaris in the centre, with some just clipping the pole and some missing it altogether.

If I get a clear night I'll set out to acquire some images circumnavigating the pole, but the weather outlook isn't good for quite some while.

Cheers, Jim

[Bryan]

Would "images" from Cartes du CIel, using an accepted star catalog, e.g. UCAC4, be useful?  There are several other star catalogs available as well.

Bryan

[Ken Self]

No problem Jim. The weather here is pretty bad too - thats one of the reasosn I've gotten cracking on this project. Don't waste valuable imaging time on my account! :)

Hi Bryan - I can get data from Stellarium, CdC and Kstars. Funny thing is they all give slightly different coordinates for the stars. Overall its easier for me to generate the stars positions based on their coordinates. I was looking for an image a around the north pole to see just how visible the stars are in reality and also how easy/difficult it is to identify them visually.

I've got the prototype template image mostly working now with simulation. Just a bit more testing needed to test the orientation of the image.

Cheers

[Brian Valente]

Have you looked at the stars used in something like polemaster? They identify about 6 of them for alignment

 

 

 

Thanks

 

Brian

 

 

Brian Valente

Brianvalentephotography.com

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Ken Self]

The Polemaster has a very wide FOV: 11 degrees by 8 degrees but my tool is intended for use with a guide camera or imaging camera so folks can save their money and buy other stuff instead :) These will typically have a very much smaller FOV - in the order of 1 degree or less in many cases. For instance, an ST80 with ASI120 guide cam has a FOV of 40' x 30'. The Polemaster stars are even further from the pole than Polaris, which is already a problem, so they wont show up at all.

The stars I've identified already show up pretty well in Jims image but it would be nice to find one or two on the other side of the pole

Attached is a crop of Jims image with the NCP marked with a red dot and alignment stars labelled. The ones labelled 9.2 and 9.7 (their magnitudes) don't have memorable names. The L shaped asterism around HIP3128 (or is it a teeny tiny dipper) is a useful landmark if you cant see Polaris in the FOV. But even that star is 30 arcmin from the pole

[JST200]

Hi Ken,

That's looking very promising. I'm now worried that I might have over estimated what PHD will actually show on the screen.

I think PHD does its own stretch which you can't control (though you can of course change the gamma slider - but I don't think that's quite the same as a stretch?). So what PHD displays may be less clear than APT's stretch.

So I now feel more obliged than ever to try and take some specific images through my Guidescope combination with PHD. Lol

Don't worry about me losing imaging time. My interest in this hobby is as much about the computing and the technology as it is about the photographic results. :-)

I just can't see it being anytime soon. :-(

Cheers, Jim

[Michael Phillips]

I'm possibly a bit late to this party but isn't this kind of thing implemented in SharpCap already?

I'm not 100% sure but you can download the previous version of SharpCap for free.

The polar alignment tool does a nifty job of alignment. Point the guidescope/camera at Polaris. Wait for it to auto-plate-solve. Rotate RA about 90degrees. Wait for another plate solve. It then calculates the PA error and tells you which way to move the mount using a handy overlay and real-time screen refreshes to help you home in. Quick and simple for me. I use a qhy5l-iiC and qhy mini guidescope (130mm FL approx f4)

I then close SharpCap and get on with the real business of PHD2 guiding :-)

[Ken Self]

Jim - thanks. That will be a great help. Those mag 9+ stars are fainter than what I deal with down here. The polar stars in the south are all around mag 7 or 8. I was testing with BQ Oct at mag 6.8 and had to turn the gain right down to avoid saturation.

Michael - there are several reasons behind this. One is that not everyone wants to use SharpCap, for instance they might be a Linux user. Another is SharpCap requires a 90 degree rotation which can be a hassle if you have a dome as you need to move the hatch. This tool should need only about 10 degrees rotation. Third reason is because I can :)

[JST200]

I'm right behind you there Ken. I wish I could! Lol. Although I've done a fair bit of coding in my time, i can't say I am particularly good at it.

Anyway, my plan is to try and take a wide'ish view of the pole area using my 300mm camera lens so you have a general view of it's vicinity. Then take as many views of the area using my Guidescope. I'll try and take both screenshots of PHD and also PHD Image saves. I'll also try to platesolve them and get a screen shot of the solve overlaid on CdC.

Cheers, Jim

[mahaffm]

Hi Ken,

I would like to help if I can. If this would be helpful, I can take a few images around Polaris using my rig setup; C8 with OAG. FL=2032, Celestron OAG with a QHY5iii178 as my guide camera with 2 binning via PHD2 or I can use SharpCap to take some images as well.  Let me if this wold be helpful for you and the next clear night I'm out I'll take a few images. I assume if this help you would want the images to be in the fit format, send via DropBox?

Thanks,

Mark

[Ken Self]

Thanks for your offer, Mark. What I'm after is images of the polar region taken through PHD2. JPEG or even a screen shot is ok as the intent is to see how easy it is (or not) for someone to identify and select some of the faint stars there. 

Its not an issue in the southern hemiphere as there are many stars brighter than mag 8 very close to the pole. But in the north there are only a handful of stars and they are fainter than mag 9.

Your setup with the OAG could well be pushing the limits. I calculate the FOV at 12 arcmin and the closest star I can find (mag 9.7) is 10 arcmin from the pole. So it might be the only thing visible which makes it hard to identify.

What I would like to see is a shot of the NCP with a reasonable normal exposure length to see what, if any, stars are visible

Alternatively, if you can connect to your imaging cam with PHD2 that could well produce a larger FOV. If its a DSLR that probably wont work. But a CCD or CMOS camera should work.

Cheers

[Ken Self]

An update: last night I got the template working - I think. The pics match up with what I see on Stellarium. I must say it nearly did my head in getting the orientation right. What with normal trigonomtry where angles increase counter clockwise, versus RA values that increase counterclockwise in the north but clockwise in the south, having to correct the orientation due to the mounts RA readout indicating "up" on the sensor as opposed to "right" and pixel values having the y values increasing down. I now just need to add in a few helper stars (not for tracking) to fill out the asterisms visible on the pics provided so far. I will also experiment with the star sizes to indicate magnitude to further aid identification.

[JST200]

Would you believe the clouds cleared tonight. Not for long mind, but long enough to capture most of what I wanted.

The images are here:  https://www.dropbox.com/sh/jdxen7ojd48jl31/AACAXolW2YH2B3-uguD0bLJga?dl=0

There are 7 sets of 3 images. The 3 images are:

- "phd pole'n'.fit" > the .fit file captured by PHD2

- "phd pole'n'.JPG" > a screen capture of what PHD2 displays

- "cdc pole'n'.JPG" > a screen capture of CdC, from the results of a platesolve of the .fit file by APT and PlateSolve2 sent to CdC

Where 'n' ranges from 1 to 7. The suffix  'a' shows the images where I had to increase(?) the gamma to show the stars better (see 3 and 3a for an example of the difference). I've attached an example of a set's .jpg (the .fit is too big to attach).

I have also added the equivalent 3 files for an image taken with my EOS1100D and a 70-300mm lens (set at 250mm). The image was taken using APT, platesolved and then sent to CdC.

I haven't sequenced the images very well, but they do cover most of the area around the pole. I think I have missed one area though.

All images were taken with PHD2 set for a 2sec exposure. Because I wanted to get going before the clouds came in, the images were taken during twilight and nautical twilight.  So not particularly dark. Also, I'm about 10 miles due south of Manchester, so the light pollution to my north is pretty horrendous.

That said, the good news is that I believe there should easily be enough stars visible for me to make this work.

[acb]

Would it also work without a view to the North Pole?

On Monday, April 10, 2017 at 4:14:03 PM UTC-6, Ken Self wrote:

I'm developing up an application that lets one use PHD2 in a similar way to a Polemaster to polar align ones mount.

The application uses the PHD2 interface. The current incarnation has a browser client using Javascript communicating with a Python webserver.

The general idea is to find where the RA axis is pointing by rotating the mount in RA and recording the position of a guide star at three points on the arc. The three points allow the centre of rotation to be calculated which corresponds to the alignment of the RA Axis.

The mount is then adjusted using the bolts in the usual way so that the RA Axis points at the Pole.

Using PHD2 has the advantage of precise measurement of the star position which in turn allows relatively small movements of the mount to caculate the centre of rotation.

My testing so far is encouraging. I've been able to largely automate the process of rotating the mount and recording star positions.

It works as follows:

Point the scope roughly at the pole by whatever means (platesolving, star hopping, star alignment, sheer luck etc...). An overlay showing the circumpolar stars could be provided to assist.

If completely lost, move the mount to its home position. Platesolve and calculate Alt/az.

Adjust mount accordingly.

Pick a guide star (optional - the app can do this with a call to find_star)

The app records the current star position.

The app assumes a mis-alignment of 5 degrees and calculates how far it needs to rotate to detect that the star is following an arc.

To keep the guide star within the search radius, the app calculates how far the mount can rotate and keep the star within the search radius. It then splits the total movement up into steps of that size.

After each step it calls get_lock_position and get_star_image to determine the star position then sets the lock position to the new star position.

In order to move the mount I connect to EQMOD directly. To do this I have to massage the ASCOM driver name returned by PHD2. I issue an ASCOM SlewToCoordinates to move the mount.

After the initial movement it recalculates the mis-alignment and calculates how much further it needs to move the mount. The assumption is that the mis-alignment is less than 5 degrees. Smaller mis-alignments of the mount require more rotation to detect movement.

After that movement it records the second position.

For the third position is moves the mount the same amount it moved (in total) for the second position.

With three star positions it is relatively simple to calculate the centre of rotation.

The vector from the centre of the image to the centre of rotation can be resolved into two components: a declination error component and a cone error component. The cone error component cannot be adjusted out. The declination component can in turn be resolved into altitude and azimuth adjustments; given the RA of the mount and local sidereal time.

Once the centre of rotation is well within the field of view it is possible to display the positions of the circumpolar stars around the centre of rotation to aid alignment.

A couple of questions for the group:

Is it worth me trying to incorporate the application into PHD2 itself or just leave it as a standalone app? If so I might need a little help. I'm reasonably proficient in C++ but a bit rusty.

If standalone, would it be possible to issue guide commands to the mount via the PHD2 interface rather than direct to ASCOM. 

Also, would it be possible to get the calibration data (specifically the angles) via the interface. I will need this to calculate the required alt/az adjustments neeeded. Currently I am just entering the values manually.

[mahaffm]

Hi Ken,

I use the ATIK383L+OSC so I should be able to connect it to PHD2 and take a few images via PHD2 through the main imaging scope of the polar region. The weather is suppose to be clear this weekend and if so I should be able to get a few shots of the polar region for you. Once I have them I post.

Regards,

Mark

[JST200]

Hi Ken,

Looking at the images I managed last night, I have realised stitching them into something coherent will be a nightmare I think. It didn't help that CdC kept changing it's orientation. However, hopefully they serve as a proof of concept?

If I get another clear night I'll try and produce a more coherent set. But it definitely won't be for a couple of weeks now at the earliest. Unless someone else beats me to it. ;-)

Cheers, Jim

[Ken Self]

Not in the northern hemispere. This tool is an alternative to drift alignment for those who don't have the views required for that. It needs to be able to view te region around the poles.

[Ken Self]

Very helpful Jim. The region around the NCP really is quite thinly populated.

[Ken Self]

That would be great. I calculate a FOV of around 30 arcmin for your setup

[Ken Self]

Latest screenshot showing the template for the northern hemisphere. Polaris is just outside the FOV

[Bret McKee]

I'm in the camp that think this seems like a perfect place for plate solving since it covers only a small, known in advance portion of the sky.

Some time ago I was thinking about trying plate solving in PHD2 to calculate the frame by frame offset and direction, and I wrote some plate solving code to play with because I had no idea how it worked. I ran out of time/motivation before I actually tried it in PHD2, but I probably have the code around somewhere. 

Let me know if you are interested in it and I'll see if I can find it. It was a "proof of concept" and might even have been written in c#, but it might be enough to get you started.

Bret

On Wed, Sep 20, 2017 at 4:56 AM Ken Self <ken.k...@gmail.com> wrote:

Latest screenshot showing the template for the northern hemisphere. Polaris is just outside the FOV

[Ken Self]

Irrespective of whether plate soving is used the principle requirement is to establish the centre of rotation and then to indicate the adjustment needed to align to the pole. Where plate solving would be most useful is in establishing an initial position near the pole. But that can equally be done outside PHD2 with any of the available tools. The alignment itself can use plate-solving to record the position of an arbitrary point at two or more orientations of the mount as input to calculating the centre of rotation. This is what SharpCap and Ekos do. The code to do this is completely separate from the core function of PHD2. My algorithm utilises the core functions of PHD2 of tracking a star position and therefore requires very little additional code. 

I did have a good look for any open source plate solving code and found only two. One was the ubiquitous astrometry.net and the other a defunct project APSS. There was one other defunct project Unimap which claimed to be open source but contained no source code! 

I'd be interested to know the accuracy of plate solving. And I'm always curious to look at algorithms. But if I were to use a plate solving approach in PHD2 I'd most likely just interface to ansvr/astrometry.net

One thing I've just realised from thinking anout the SharpCap solution is that I can reduce the amount of rotation needed by recording just two positions. I use three because my prototypes were using a non-goto mount. With a goto mount the direction of rotation is known so only two points are needed.

[JST200]

Hi Ken,

I've been thinking again - dangerous I know!!

Although I'm happy to try and take some more coherent photos of the Pole for you, I was wondering if it's necessary.

As it happens, because of something Ivaylo Stoynov (APT developer) said in one of his posts, I have been taking a look at SkytechX. I was surprised to see that it readily show stars well below a magnitude of 10 - much better than CdC. It seems to be quite controllable in what it displays and what you can add. So with a bit of playing I have been able to produce the attached two screenshots.

As I think we have concluded that using a lower end (?) guidescope configuration and with pretty awful light pollution, I can PHD can still display magnitude stars of less than Mag 10. Therefore I have created a star map showing stars of less than Mag 10 and labelled some of the more prominent ones (see Mag 10.jpg). I have also added a circle to mark the actual pole. The star maps FOV is about 5 minutes.

It's extremely easy to add the label to the stars and I could have labelled many more, but am running out of time. Hopefully the ones I have done give a flavour. The labels are just their designation, but I could have just as easily added their magnitude (or any other text) and I could have left the EQ grid visible if required or change it for the AA grid.

I then created a similar 2nd star map with stars of magnitude of less than 9. I could easily have produced more either above mag 10 or below as required.

Once I worked out how to do it, it probably took me about 30 mins to produce thees maps.

What do you think?

Cheers, Jim

[Ken Self]

Hi Jim,

I agree with what you say. I'm pretty confident that stars down to mag 10 should not be a problem and even then a slightly  longer exposure would help. The best course of action for now is for me to finish the code I've got and do my final testing. The latest changes have been drawing proportionally sized stars and solving the alignment with just two stars instead of three - which halves the amount of rotation and reduced the time taken considerably.

Once I've done that I'll see if Andy and Bruce are happy to accept the code for wider testing. During that time the star plots can be validated and it is very easy to add more. All I need is RA, Dec, Magnitude and a name.

The only thing holding me up now is the weather here

Cheers

[Bryan]

If you install the appropriate catalog, CdC will go to mag 20.

Bryan

[JST200]

Thanks Bryan, I should have guessed as much. :-) Perhaps I should have added the rider "out of the box"? ;-)

I've been using CdC for a number of years and really like it. In fact I now have a dichotomy; I actually prefer the CdC look and feel, but I like the fact that SkytechX can zoom into the moon's detail. Although I haven't tried it yet, I think this means I can manipulate close up moon images without reverting to something like Virtual Moon. Unless I've missed that in the CdC extras too!

Cheers, Jim

[Ken Self]

I've just completed a very successful test of the tool. Unfortunately, I had just corrected the code that displays the correction needed when the clouds rolled in. I had introduced a minor bug in the drawing codewhen I moved from a 3 point alignment to a 2 point alignment. The bug that was causing the mount to slew too far has also been fixed and the alignment star tracked perfectly through an arc of 11 degrees.  I'm now just making some functional changes so that people with On Camera guiding can use the tool by slewing manually. They will need to collect three alignment points whilst people with goto mounts connected with ASCOM can run automated and only two points are needed.

Attached are two screen shots from tonight. You can see how well the template matches the actual stars. The second screen shot shows the first alignment point and the preliminary second point. The tool calculated the centre of rotation roughly in the position I've drawn in. Ive also marked the position of the pole. You may be able to see two faint stars of around mag 11. One of these is a mere 2 arcmin from the pole. So the tool was estimating my polar misalignment at around 15-20 arcmin. 

To further test the accuracy, I slewed the mount through about 60 degrees and the alignment star tracked the drawn circle very well.

As an added bonus, when the clouds first scudded over, PHD2 lost the alignment star which stopped the tool until PHD2 was able to lock on again, and it then resumed.

[mahaffm]

Very nice Ken, you are making great headway on this. Do you still need me to take some images around Polaris? If the weather holds this weekend I can do so if you need me to.

Regards,
Mark

[Ken Self]

I think you can hold off for now. Wait till I can distribute the prototype so you can then check whether the stars offered are usable.

[JST200]

Looking great, Ken. Can't wait for a trial!

In your test the template is the same orientation as the PHD image. Is this chance or do you rotate the template according to the current time?

Cheers, Jim

[Ken Self]

I rotate the template according to LST, the RA of the mount and the camera orientation as determined by PHD2 calibration.

[JST200]

Awesome!! :-)

[mahaffm]

Hi Ken,

Ok, I'll wait for the prototype. BTW - I use the QHY PoleMaster for my Polar Alignment. The PoleMaster works great, but having a "free" way of doing PA would be great.  It will be really interesting to see how the accuracy of two compare.

Thanks again for your work on this! I think if this works and gets incorporated into PHD2 it will be a big plus.

Regards,
Mark

[Ken Self]

Just another quick update. I'm nearly ready to issue a pull request - the code is sitting in my Github repo. The manual alignment option is working and I've cleaned up the code quite a bit. Just want one more field test to confirm some of the changes I made. The forecast for tomorrow night is looking promising.

[Ken Self]

SUCCESS!!! There was just enough of a gap in the clouds to have a clear view of the pole.

This is the first automated run showing the centre of rotation is at the bottom left of the display. The SCP is near the two faint stars at centre left.

I slewed the mount manually with a long exposure to see how accurate the calculation was. It looks pretty good. I think the deviations at the edges of the display are due to the astigmatism of the RC scope.

Next I tried with the manual method which give a similar result.

I changed the code a little to speed up the process. The algorithm uses a constant number of pixels deviation from a straight line to determine when it is safe to calculate an arc. I reduced this from 8 to 5. The next pic shows that this had no adverse effect and it was much quicker. The mount only had to rotate less than 10 degrees.

As indicated on the display I adjusted the azimuth knobs and reran the alignment. Muc closer but not quite there

After another az adjustment its much closer - almost impossible to see the correction needed. When correct the three visible stars should rest on their respective circles. That suggests I overcorrected slightly in az and need to correct a little in altitude. The astigmatism may be affecting the position of BQ Oct.

If I want to bring the centre of rotation closer to the centre of the display I need to adjust for cone error. This moves the centre left/right. A simple dec correction moves it up/down.

Time to raise a pull request I think.

[Andy Galasso]

If anyone would like to try the Static Polar Align tool, it is now available in PHD2 version 2.6.4dev1.

If you are already running the dev series builds, the easiest way to upgrade is to select Help => Check for updates.

If you are running 2.6.4, you can switch to the dev series builds by clearing the option "Only check for major releases", then select Help => Check for updates.

If you are running an older version of phd2, you can get version 2.6.4dev1 from here: PHD2 Development Snapshot Builds

The instructions for the Static polar align tool have not yet been added to the user manual, so you'll need to read this discussion thread for instructions.  Please post your feedback here in the forum.

Many thanks to Ken Self for the work on the new tool and the contribution to PHD2!

Andy

[Brian Valente]

So cool! thanks a million guys

 

 

 

Thanks

 

Brian

 

 

Brian Valente

Brianvalentephotography.com

--

[Ken Self]

And my thanks to Andy for making it available in this release. I'm expecting bugs and enhancements and I'm working on the documentation. So any questions just fire away.

I'm keen to see how it performs in the northern hemisphere and its accuracy over a bigger sampe than one. There is a threshold constant I've set in the code that is a balance between accuracy and speed.

[mahaffm]

Hi Ken,

If the weather holds this weekend, i'll give the Alternative Polar Alignment a try. However, I'm not exactly clear on the workflow/process. I'm using a OAG so I'll be using the main imaging camera for the PA. Would you be able to list out the workflow/process steps that should be followed? I think this would help in the respect that if everyone can follow the same process maybe to results would be more consistent and if there are bugs easier to diagnose and track down.

Thanks,

Mark

[Ken Self]

Star with the scope pointing at/near the celestial pole and PHD2 running and connected and calibrated.

Select Menu > Tools > Static Polar Alignment

I'll assume you have a fully functional mount which can slew under computer control and report its position:

The graphic shows stars that should be in the FOV with a letter next to each one - and it should correspond to the main PHD2 display.

Use the letter to pick one of the stars in the drop down list and select the same star on the main PHD2 display as though you were going to guide on it

Click Rotate and watch as the mount records the stars position and slews around. 

When it finishes an overlay appears on the main display.

Adjust the alt knobs to move your selected star along the red line on the overlay

Adjust the az knobs to move your selected star along the blue line on the overlay

Those adjustments will move the celestial pole to the centre of the circle which corresponds to your mounts polar axis.

If you had to make big adjustments, manually move in declination to bring everything back into the FOV and repeat the whole process to check if any more adjustment is needed. You may also want to restore the mount go its starting position in RA.

When done, close the tool and start guiding/imaging

[tym...@gmail.com]

I'm not sure what my FoV is, but here's my attempt at a calculation.
Scope has Focal Length of 952mm
imaging camera has a sensor width of 22.3mm
FoV=2 * arctan (22.3/2/952) = 1.34degrees.
The height is ~1degree.
is that the correct way tofind FoV for the camera?

On Sunday, 17 September 2017 23:45:48 UTC+1, Ken Self wrote:

Thank you Jim! Thats an enormous help. 

And you are right, I am trying to avoid the complexity in plate solving. Maybe a popup aid that shows the main stars near Polaris as a guide would help.

Its kind of ironic that in the ays of polarscopes I used to curse how "northerners" have a handy, bright pole star while we antipodeans had to rely on Sigma Octans which is very hard to find. But once you zoom in closer to the pole we have an easily identifiable asterism, several in fact for different FOVs, and a pair of stars within a couple of arcminutes of the pole.

Let me work on this conundrum a bit and see what I can come up with. Worst case is that Polaris is the only choice but that pretty much excludes use by anyone with a FOV less than about 1 degree.

[bulrichl]

Yes, both your formula and result are correct.

Bernd

[tym...@gmail.com]

Thanks.

[mahaffm]

Hi Ken,

I give the Alternative Polar Alignment a try this weekend but was unable to complete. Here's what I did.

1. Polar aligned use QHY PoleMaster - I thought I would give everything a test first with being accurately Polar Aligned.

2. Connected PHD2 to the ATIK 383 camera, Set the camera binning to 4, approx FOV 30" x 22"

3. Started PHD2 Alternative Polar ALignment and started looping.

From there I got lost.

I could not identify what star I was suppose to select. I save the image of my FOV via PHD2, Please see the image below

Below is the PHD2 Alternative Polar Alignment Screen.

For me I don't know the stars well enough to know which start I was suppose to select in my FOV so I kind of got stuck.

Let me know if you have any suggestions on what you want me to try. We are suppose to have another clear night, with a bright Moon, so I can do some testing since I will not be doing any imaging.

Regards,
Mark 

[mj.w...@gmail.com]

Hi all

I'm hoping to try this out tonight.

Attached is my guidecam view of the NCP (ASI120MM and 300mm guidescope, FOV 55x41 arcmins)

Has three of the stars on your A-F list of Sept 20th.

Michael

Wiltshire UK

[Ken Self]

Thanks for the input Mark. Do you have a plate solve of your image, or can you send it through?

A couple of things you can try:

Select one of the stars you can see on the display. Don't worry about identifying it.  The run the automated routine. This will at least check if tha part fo the code works and show where the centre of rotation is. 

Another thing to try is to slew to a known star on the list e.g. Polaris. The pole will not be in the field of view but the process should* still show how much polaris needs to move to get alignment. 

*Note: I'm tracking down a reported bug which shows the corrections incorrectly for stars some distance from the pole

[Ken Self]

Thanks Mike. I'll add those identifiers in place of "unnamed" on the list. Given that the NCP need not be at the centre of the FOV (an advantage of the small slew angle needed), you could well get Polaris into the FOV. For the north in particular, I'm wondering if I should apply the declination to the helper image, or centre on the selected star rather than the CP. 

[mahaffm]

Here's the plate solve. However, as per the plate solve, it is no where near to where I was pointing. Not sure what's up with that.

I can send you the fit file if you need it. via drop box. 

In the meantime, I'll play around some tonight with what you suggest and will report back what I find.

Thanks Ken,

Regards,

Mark

[Ken Self]

That is an odd plate solve - could not be further from the pole.

Is it perhaps an old image?

For now I'd suggest you align on Polaris as it is readily identifiable. Some people have had some issues with the tool aligning on Polaris so I've just submitted a bug fix. You may want to wait for the next dev release with the bug fix. 

[Andy Galasso]

On Tue, Oct 3, 2017 at 8:56 AM, Ken Self <ken.k...@gmail.com> wrote:

You may want to wait for the next dev release with the bug fix. 

I just uploaded a new dev release, 2.6.4dev2, with the latest fixes included.

Help > Check for updates to upgrade

Andy

[mj.w...@gmail.com]

Tried the new PA routine last night.

I made the scope and the guidescope concentric on the moon, just to make pointing easier.

Slewed to Polaris which was big, bright and overexposed on the guidecam, and selected this for the first run (after reducing exposure). 

Result is shown in PA01, I overexposed Polaris again to make it obvious on the screen.

Moving the wedge the correct way was easy (unlike the Equatorial PA routine where it's a guessing game)

I ran the routine again with Polaris, the result is PA02 - almost there.

I did two more runs, the result is shown in PA04.

Limiting factor was adjusting the Meade wedge !

I didn't occur to me to slew the scope towards the centre of the circles after the first run, and then choose one of the other Reference Stars on the list - would that have made a difference?

PS - how do you paste an image into the text?

Michael

Wiltshire UK

[Ken Self]

That looks like a great result Michael. In automated mode I don't think using another star would make much difference. In manual mode its more accurate to triangulate the stars around the pole.

Where slewing ind ec to centre the pole could help is in familiarising yourself with the star patterns around the pole. I find that makes it easy to check PA visually first each session and if necessary, run the tool to confirm. This works well with a GEM as my park position points at the pole. Maybe not quite so handy if you need to slew there first.

When doing that, note that you can not fully the centre if you have cone error. So the offset of the CoR from the centre of the sensor orthogonal to declination is an indication of your cone error. As you camera is well aligned, the cone error is the offset of the CoR along the sensor x-axis.

As you can see from your three images, that offset varies by about 15 pixels which is a rough indication that the accuracy of the tool is in the order of an arc minute or so.

[Steven Bellavia]

This is awesome!  I can't wait to try it!

Thanks!

Steve

[Brian Valente]

Yeah I’m going to give this a try as well

 

Polemaster seems to work great, but I’m still getting reports of 4 arcmin error, which according to polemaster should be more like 30 arcsec

 

 

 

Thanks

 

Brian

 

 

Brian Valente

Brianvalentephotography.com

 

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of Steven Bellavia
Sent: Tuesday, October 3, 2017 1:48 PM
To: Open PHD Guiding <open-phd...@googlegroups.com>
Subject: Re: [open-phd-guiding] Re: Alternative Polar Alignment using PHD2

 

This is awesome!  I can't wait to try it!

--

[mj.w...@gmail.com]

G'day Ken

You said  " In automated mode I don't think using another star ( than Polaris ) would make much difference"

To clarify then, given that the NCP stars are harder to identify than the SCP ones, is there any point not using big bright obvious Polaris all the time?

Thanks

Michael

Wiltshire UK

[mahaffm]

Hi Ken,

Ok I downloaded and installed the beta 2.6.4Dev2.

Here's what I did

1. Completed a Polar Alignment using PoleMaster

2. Started PHD2 2.6.4Dev2 and connected my mount and imaging Camera, (ATIK 383L+)

3. Started Static Polar Align

4. Selected a star and selected "Rotate" and let the Static Polar Align do its thing.

The image below is what I ended up with but from there I couldn't tell how I needed to move the mount.

I was getting a centre of -1 for the Xpx and a +1 for the Ypx. Also from the image circles I could not see what it was telling me to move and by how much.

Anyway, I'm don't seem to be much help with this testing. Maybe I'll leave it up to others to work on the testing. I think this is a little beyond my understanding.

Regards,

Mark

[Ken Self]

Hi Mark - could you please send me your debug log from that session. The readout on the status line looks suspicious and I'd like to get to the bottom of it.

Thanks

Ken

[Bryan]

Ken

I planned on trying out your PA approach last night.  When I opened Tools, Static PA, the three reference stars shown in the display are

G TYC 4629-37-1 (  28'  from Polaris)

H TYC 4627-6-1 ( 28'   from Polaris)

I  TYC 4661-2-1 (  45'  from Polaris)

The FOV of my guide system (480 mm scope / SBIG ST-i) is 25' x 34.  All of the above stars are outside that FOV.  

Since it seems there is insufficient data in PHD2 to calculate the FOV for a user's system, i.e. no sensor size), I wonder how Static PA determines what reference stars to display.  Should I just use Polaris?

FIT and stretched JPG  images attached, also Debug log.  I never started guiding, so there is no Guide log.

Bryan

[mahaffm]

Hi Ken,

There were actually 2 debug files with the same date and time so I'm not sure which is the correct one so I attached both.

Hope this helps,

Regards,
Mark

[Ken Self]

Hi Bryan

It works out the FOV from the pixel scale and the size of the image in pixels (which it does know). But the display is centred on the pole. This works fine in the SH as the best stars are arranged around the pole. Would it be helpful if I change it so that the display is centred on the selected star or maybe between the pole and the selected star.

Also, based on the feedback I've had so far I'm going to add more bright stars to the list that are further from the pole and remove some of the fainter ones. Aligning on Polaris or another bright star seems to be the best option and others have reported success with that. Just take care that it is not saturated 

[Ken Self]

Thanks Mark - I shall pore over those while the moon is full and the clouds are thick.

[Bryan]

Ken

Thanks for clarifying the FOV.  I forgot that the image size would be embedded in what PHD2 reads.

However,  I still have the issue that the stars shown on the display are outside the FOV of my camera.  Attached are two screenshots from Cartes du Ciel. 

The red rectangle marks the FOV of the ST-i image (also attached).  The rotation angle of the ST-i image is the value found by plate-solving using AstroTortilla.

The two brightest stars in the lower right of the image are HD 5914 and TYC 4627-86-1. 

The two screenshots show the position of G and I stars relative to the camera FOV.  

I totally understand that I may be missing something in the approach!

P.S.  This is in a permanent installation with the mount already polar aligned using PHD2 drift align.

Bryan

[Ken Self]

I've started on a change that will let you pan the helper display around. In the meantime, if you can get Polaris on the main display you can select it and go from there. The small image is just an aid to finding the less obvious stars but has no bearing on the actual alignment.

[Brian Valente]

Ken

 

I’m curious what you think about this – it seems like this alt PA feature would be more accurate than polemaster (or perhaps a good ‘tune up’ from a polemaster align) because of the longer focal length. Do you agree or feel otherwise?

 

Thanks

 

Brian

 

 

Brian Valente

Brianvalentephotography.com

 

From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of Ken Self
Sent: Thursday, October 5, 2017 1:16 PM
To: Open PHD Guiding <open-phd...@googlegroups.com>
Subject: Re: [open-phd-guiding] Re: Alternative Polar Alignment using PHD2

 

I've started on a change that will let you pan the helper display around. In the meantime, if you can get Polaris on the main display you can select it and go from there. The small image is just an aid to finding the less obvious stars but has no bearing on the actual alignment.

--

[Ken Self]

Hi Brian, I'm reluctant to make claims of accuracy till more data is in. The longer focal length and PHD2 centroid calculation (not sure what Polemaster does in this regard) are traded off against a shorter slew distance. Both are affected by seeing and mount precision, the Static PA tool possibly moreso since we are dealing with smaller movements. However, I'm pretty sure with Static PA you can get a "good enough" alignment without having to spend extra.

[Bryan]

Ken

Thanks.

Polaris is easy.  I'll give it a go tonight.  I can reduce the exposure sufficiently to de-saturate Polaris.

Bryan