Looking for beta testers of new Polar Alignment utility.

[Gonzo]

It's better, I can see now that PPA is trying to use the solve-field however I now have the following issue:

/bin/bash --login -c "solve-field /usr/local/astrometry/etc/astrometry.cfg -u arcminwidth -L 1.00 -H 120.00 -z 2 -z 2 -B none -P none -M none -S none -R none -U none -T -p -O -O  \"/Users/stephanelucas/Astro/PA_session/_1180154.JPG\""Reading input file 1 of 2: "/usr/local/astrometry/etc/astrometry.cfg"...ERROR: Image type not recognized: Unknown image type "['ASCII English text']"augment-xylist.c:592:backtick Failed to run command: /usr/local/astrometry/bin/image2pnm.py --sanitized-fits-outfile /tmp/tmp.sanitized.cgqyoK --fix-sdss --infile /usr/local/astrometry/etc/astrometry.cfg --uncompressed-outfile /tmp/tmp.uncompressed.C5laGc --outfile none ioutils.c:579:run_command_get_outputs Command failed: return value 255

.

my PPA.ini file :

[nova]apikey = r.............. [file]imgdir = /Users/stephanelucas/Astro/PA_session [appearance]geometry = 440x470+92+347 [operations]restrict scale = 0 [local]shell = /bin/bash --login -c "%s"downscale = 2configfile = /usr/local/astrometry/etc/astrometry.cfgscale_units = arcminwidthscale_low = 1.0scale_hi = 120.0xtra = -z 2 -B none -P none -M none -S none -R none -U none -T -p -O

EDIT ~ error above is now FIXED.

Had to remove the configfile line (empty it).

Edited January 18, 2015 by Gonzo

[Gonzo]

Happy to report that the Local solve now works on the Mac.

Thanks Themos.

[themos]

Gonzo,

don't put -z 2 in the extra, that is the option you select via the downscale setting. Also, no need for the -O, I put that in anyway. I always use arcsecperpix as the scale units, there may be subtle bugs if you use anything else. 

[themos]

I had a play last night with a EQ3 Pro mount and Canon 5D3 and 200mm lens. I delibrately offset the PA by a few degrees and then shot a vertical and horizontal image which PPA solved using a local installation of AT in about 5 seconds/image. What threw me was the image that popped up showing the red X etc, because I was about 4 degrees off the magnification of the visual image was too large to see the X and the true NCP. I also struggled to visualise what the move instructions meant i.e. move 4 degrees right and 2 degrees down. Am I right to assume I had the mount 4 degrees left of the NCP and 2 degrees too high??

Also does the alignment of the camera relative to the axis of the mount significantly affect the results?

Yes, to the first. The red x is where you are pointing now. The instructions are to move it (4 right, 2 down) so that it ends up on the NCP.

The alignment of the camera relative to the mount does not matter. What matters is that the image we call "horizontal"  or "improvement" is exactly of the orientation you'd see with your naked eye, that is, horizon to the bottom, zenith to the top.

[Gonzo]

Gonzo,

 

don't put -z 2 in the extra, that is the option you select via the downscale setting. Also, no need for the -O, I put that in anyway. I always use arcsecperpix as the scale units, there may be subtle bugs if you use anything else.

ok, will modify the xtra line.

[Gonzo]

Stephanes-MacBook-Air:PolarAlign stephanelucas$ grep xtra PPA.ini xtra = -B none -P none -M none -S none -R none -U none -T -pStephanes-MacBook-Air:PolarAlign stephanelucas$

results are good

Stephanes-MacBook-Air:PolarAlign stephanelucas$ ./PPA.py ___________________________________________________________/bin/bash --login -c "solve-field  -u arcminwidth -L 0.00 -H 0.00 -z 4 -B none -P none -M none -S none -R none -U none -T -p -O  \"/Users/stephanelucas/Astro/PA_session/_1180164.JPG\""Reading input file 1 of 1: "/Users/stephanelucas/Astro/PA_session/_1180164.JPG"...jpegtopnm: WRITING PPM FILERead file stdin: 2992 x 2992 pixels x 1 color(s); maxval 255Using 8-bit outputExtracting sources...Downsampling by 4...simplexy: found 816 sources.Solving...Reading file "/Users/stephanelucas/Astro/PA_session/_1180164.axy"...  log-odds ratio 134.368 (2.26596e+58), 22 match, 0 conflict, 15 distractors, 26 index.  RA,Dec = (63.7289,89.7015), pixel scale 16.3889 arcsec/pix.  Hit/miss:   Hit/miss: ++++++-+++++++-+-+-+-+-+++--+-------+(best)---------------------------------------------------------------Field 1: solved with index index-4116.fits.Field: /Users/stephanelucas/Astro/PA_session/_1180164.JPGField center: (RA,Dec) = (64.21, 89.7) deg.Field center: (RA H:M:S, Dec D:M:S) = (04:16:49.784, +89:42:12.610).Field size: 13.5411 x 13.5326 degreesField rotation angle: up is 23.4517 degrees E of NCreating new FITS file "/Users/stephanelucas/Astro/PA_session/_1180164.new"...___________________________________________________________local solve time 7.75195407867___________________________________________________________Stephanes-MacBook-Air:

[AndyKeogh]

I had  a failure with a solve tonight

Selecting job id 986652

Sending to URL: http://nova.astrometry.net/api/jobs/986652

Sending to URL: http://nova.astrometry.net/api/jobs/986652

HTTPError HTTP Error 500: Internal Server Error

Wrote error text to err.html

Exception in Tkinter callback

Traceback (most recent call last):

File "C:\Anaconda\lib\lib-tk\Tkinter.py", line 1486, in __call__

return self.func(*args)

File "C:\Users\AKeogh\Documents\PhotoPolarAlign\PPA.py", line 1263, in <lambda

>

nxt = Button(self.wfrop, text='Nova', command=lambda : self.solve('i','nova'

))

File "C:\Users\AKeogh\Documents\PhotoPolarAlign\PPA.py", line 887, in solve

img2wcs(self, self.apikey.get(), aimg, awcs, hint)

File "C:\Users\AKeogh\Documents\PhotoPolarAlign\PPA.py", line 455, in img2wcs

if stat.get('status', '') in ['success']:

AttributeError: 'NoneType' object has no attribute 'get'

with estimated scale 4.14

Andy

[themos]

That's a shame, sometimes the nova.astrometry.net server goes down for a while. This looks like one of those times.

[themos]

I will try to add code to recover from this, probably just keep trying to contact the server, for now, maybe implement a timeout later.

[visionoz]

Used in the field 2 weeks ago on an actual dark sky trip here in Western Australia. Luckily had put some practice time in backyard testing the "local" solve method a few weeks back and all seemed to work well ... except that "Murphy" sturck!! :-/ ... because I wiped out the astrometry.net files inadvertently from my laptop harddrive and had only a single solve file available; unfortunatetly the remote site was in a state nature reserve without internet access and so I just had to go visual instead (didn't bother to drift-align due to the fact it was only a one-niter and time was short) - anyway did enjoy the visual exercise which I missed for a quite long while.

Have to wait for next new moon to try this out again in the field with all my astro-imaging gear decked out ready and just using PPA to polar-align; quite confident it will work as it should 

Cheers

Bill

from a very hot Perth, DounUnder!

[themos]

Thanks for  the report Bill! I will have occasion to try it out at 7 degrees North latitude next week . I can imagine that refraction will cause problems with PPA...

[perdrix]

 I've a few questions:

1) Will this run with Python 3.4.3 instead of 2.7.8?

2) Has the issue of a work directory as distinct from install directory been resolved?

3) What is the recommended plate solver for working offline?

3) Does it matter what the orientation of the camera is for the two shots so long as they are approximately orthogonal?

4) What size FOV is needed?  For example would 49.5' by 27.85' (MS Lifecam Studio in ST80) be large enough or is this way too small?

Many thanks

Dave

[themos]

1) I'm sorry, Dave. I'm afraid I can't do that.

2) I think so, have you tried running the installer?

3) astrometry.net and I've made some attempt to support AstroArt solves.

3)(again) Yes, it does. The orthogonality is not crucial, some large angle will do. What is crucial is for the one marked as horizontal to be horizontal. The reason for that is that we want to give improvement suggestions in terms of the horizon directions, that is, left-right and up-down.

4) I think it will work if you manage to pick up enough round stars for a solve.

Edited March 1, 2015 by themos

[perdrix]

> Sorry Dave

Grin - just wait till you start singing "Daisy, Daisy ..."

OK I'll get Astrotortilla then.  OK understand the bit about horizontal/vertical now.

Once I've got my PE problem resolved I'll try to have a play with this.

Thanks

Dave

[themos]

I am sure at some point we will migrate to Python 3 but not yet. If your skills are up to it, you're welcome to do it yourself, you can fork the code from https://github.com/ThemosTsikas/PhotoPolarAlign.

AstroTortilla is nice to have as it can do some very useful stuff. As a side-effect, it takes care of installing a working astrometry.net on your machine and we can use that.

Themos

[rwg]

Hi Themos,

maybe jumping in here without knowing what I'm talking about, but if you have plate solved the two images *and* you know the location and time of the images, can't you calculate which direction is horizontal/vertical in each image?

cheers,

Robin

[themos]

That's true, Robin, and I had some experimental code that drew a blue line on the image for the computed horizontal for confirmation. The reason I am reluctant to put that feature in is that people might have the wrong time set on their camera, or the wrong time zone or the wrong long/lat coordinates and then they would get bad advice. I also lost those 10 lines of code during a disk crash (but they would be pretty easy to rederive).

[rwg]

Hi Themos,

Ah, yes, if you're relying on jpeg date/time stamps then you are most likely stuffed with regard to accuracy. In my mind I was thinking more like AstroTortilla which can grab images itself (from APT for instance) and talks to EQASCOM. In that case you have the date/time and location available and could even automate the whole measurement procedure.

cheers,

Robin

[themos]

I think I sent the AT people a message about this method, suggesting they might like to incorporate it into AT, with the automation opportunities that might arise. But I am keen to keep this incarnation of the method something that can be used even with a barn-door tracker. Also, AstroTortilla development seems to have stopped.

[malc-c]

Gonzo has just referred me to this thread as I'm searching for a simple way to fine tune my polar alignment.  It seems straight forward but I have a few questions.

• Can this be used with a DSLR conneted to a reflector, or do you need to attach the camera directly to the mount in place of the telescope.
• If it can be used through the scope, what orientation must the scoe be.  I've seen an image which shows a DSLR camera parallel to the tube, but also horizontal, ie standing infront of the OTA the focusre is at the 9 O'clock position.  Mine is at the 8 O'clock position due to getting the balance right - will this have an effect on the results when plate solving images ?
• Is there any documentation to suggest how much to turn the adjustments - seen the 20 down, 6 right type numbers, but what do they mean ?

is there a link to the latest build and other software requirements ?

[themos]

Hello Malcolm,

Yes, it can be used with a reflector, with some extra care. The longer focal length of the typical scope will give you a small field of view and perhaps too fine an image scale. You can compensate for that,  to an extent, by choosing a reduced resolution mode on your DSLR. The online plate solver (through nova.astrometry.net)  should handle almost anything you can throw at it, the local solver might need some tweaking of parameters to work quickly. There's two other issues with the reflector: parity and angle.

Parity, first: the image produced by the camera may have the wrong orientation. Take an image of a terrestial feature you recognise through the telesceope and work out if you need to flip it right-left or upside-down to produce a correct image. I just did that with my newtonian and the image seems to come out with the natural orientation.

Angle, next: However you place the camera in the focuser tube, you will have to figure out  which RA-axis position corresponds to the horizon going along the long side of the sensor. If I place my newtonian horizontal, and have the focuser tube also sticking out horizontally, and then place the camera so that the long side of the sensor is horizontal, then I get the needed angle: the horizon looks horizontal! I haven't come up with a foolproof procedure so some thinking on your part will be necessary to determine how to shoot a horizontal image. If you get it wrong, the feedback provided by the utility will not be much use as it will have a wrong sense of what "up" or "left" means.

The numbers reported are arcminutes, 1/60th of a degree. I aim to go under 5.

The repository is in https://github.com/ThemosTsikas/PhotoPolarAlign

[malc-c]

Thanks for the reply.

Having the scope permenatly mounted in the observatory makes it a tad awkward to het visual on land based targets, but I'll try it out once I've resolved the issue I'm getting with Python, but that will be after I fix the observatory PC which crashed on me this afternoon !

[themos]

The alternative is to recognize the star pattern around Polaris, compare it to a planetarium view and rotate the camera until it looks correct, that is, similar to the horizontal planetarium view. I've been thinking about adding an automatic horizon finder, marking it with a blue line on the image, but I need to know the correct geographical position and time of capture. 

[perdrix]

I tried to use PPA the other night for the first time, but failed to get it to work.

I imported the two png files I had, and it said it had solved them.  But when I clicked on "Find Polar Axis", nothing appeared to happen. I tried to paste in a screenshot here but was told it wasn't supported (how do I do that)?

What did I do wrong?

Thanks

Dave

[themos]

Hello there, I think it's best to send me your image files and I'll have a look.

Themos

I tried to use PPA the other night for the first time, but failed to get it to work.

I imported the two png files I had, and it said it had solved them.  But when I clicked on "Find Polar Axis", nothing appeared to happen. I tried to paste in a screenshot here but was told it wasn't supported (how do I do that)?

What did I do wrong?

Thanks

Dave