Blind vs Astrometric plate solve

[johnson19938]

I'm new to Observatory, and maybe I'm not setting something correctly. I'm having a problem plate solving a when I use astrometric version vs blind version from nova.astrometry.net. The internet solving works every time, but I can't get astrometric which uses the files on my computer to work at all no matter what settings I change in the drop down menu in the solver. I've read the notification that comes up and do exactly what it says and still no joy. Again I'm new to the app and maybe I'm just not setting something correctly. What confuses me is it will work each and every time from the internet, but the very same image will not work from my computer. Any help will be appreciated. I've downloaded all the database files from the internet to use with the app. Not sure what else to do. I do a lot of my imaging from a location where cell service is spotty at best and can't always share my internet connection from my iPhone to my MacBook pro to be able to use nova.astrometry.net and was hoping to have a plate solver that worked without an internet connection. Again hopefully I've expressed my problem properly.

[Sander Berents]

Please share one of your images with for example Dropbox so I can check what might be causing the issue.

[johnson19938]

Not exactly understanding how you want me to supply an image. One note though that I forgot to mention on my post was I'm shooting raw images with an older Nikon D200 which gives you a NEF raw image format. On further reading, another question that I might have is, does observatory need a FITS file to be able to plate solve on the computer instead of using blind plate solving from the internet. I will supply an image for you to evaluate, but I will make one mention, the blind plate solving is working, and giving me the correct coordinates on my image which I can then use with my planetarium software, namely Sky Safari. I should have specified the camera I was using at the outset. Again if you will be a little more specific as to how to get an image to you I will be more than happy to provide that. Thanks

[Sander Berents]

The camera type doesn't matter. Observatory can plate solve RAW images as well. There's no need to convert them to FITS or XISF format.  The most common cause of plate solving failures are image distortions (wide fields), noise and out of focus stars. But, since blind plate solving works for you, assuming you have selected the option to use Observatory's star detector, the latter two do not seem to be the problem here.

This forum doesn't allow large attachments, so please use Dropbox, OneDrive, Google Drive or iCloud Drive to share one of your images with support@codeobsession.com. Your Mac has iCloud Drive by default, so if you are not familiar with any of these, iCloud Drive may be most convenient for you. Copy one image in iCloud Drive, Control-click it, and in the contextual menu click Share > Share File. In the dialog box that appears, choose "View Only" as permission and then have it email the link to support@codeobsession.com. This will allow me to download the image, and only that image, nothing else.

[johnson19938]

I sent the link to my image in my dropbox account It should come through on the email. ke...es@...net you can reply there or here. thanks again.

[edited by moderator: removed email address]

[Sander Berents]

I used the blind plate solver first to determine the image scale. This should remain more or less constant with the same setup for other images. It's a wide field shot with many bright Tycho-2 stars, so there's no need for the UCAC4 and USNO-A2.0 catalogs. For the highly accurate Tycho-2 catalog Observatory automatically expects the solution to be very accurate as well, which in this case it cannot satisfy because of the image scale.

What I did to plate solve this image was:

• Because it is a wide field shot, I reduced "Frame size" to its minimum. This makes the plate solver look only at the inner 25% of the image, where the distortions are minimal.
• I set the pixel scale to 6.2, which is approximately what the blind plate solver returned.
• I explicitly selected the Tycho-2 catalog. This is not strictly necessary, but knowing that the image has many Tycho-2 stars, this makes experimentation with the parameters faster.
• I set the Search radius and Search spiral to their minimum values. These options are rarely needed.
• I unchecked "Maximum residual", and set it to 5. This increases the tolerance. Observatory should really increase the tolerance automatically when a larger pixel scale is used, but it looks like it doesn't do that enough for this image.

With these settings, it solved in a second.

Attached my settings.

[johnson19938]

Worked like a charm. Question. I shoot through my 12" Meade lx200 set with a focal reducer on it that reduces the focal length to around 1600-1728mm. The range is because it's an older giant easy guider from a company called lumicon. Was purchase back in the late nineties. It is adijustable so I think you get the idea. Would I use those same setting for a shot through it. I guess I could try it but I was just wondering. I'm going to have three situation for shooting. All with the same camera, but with the 200mm lens the 12" and an 80mm short tube refractor. Thanks for your help, this will help a lot.

[Sander Berents]

With the LX200 at that focal length the image scale is around 1"/px, so you need to adjust that in the plate solver. With this camera, you may see severe vignetting with the LX200, so although reducing the "Frame size" setting is typically only necessary for wide fields, in this case you may need that so the star detector ignores the distortions in the corners. There's the added chance that the vignetting is so severe that it affects the background so much that adjusting its sensitivity may be needed, or the Flatten Background adjustment should be applied, or the image should be cropped.

The short tube refractor probably has an image scale close to 3"/px with your camera. For both instruments, you may want to do a blind plate solve to get the correct value if you want to avoid having to calculate it yourself.

With the LX200 there may not be enough Tycho-2 stars in the field of view, so for that the UCAC4 or USNO-A2.0 catalogs probably should be used.