Plate solving at any focal length?

[Rich1980]

Is it possible to plate solve at any focal length? 

Thanks in advance, 

Rich. 

[vlaiv]

Not sure what exactly are you asking, but yes, you can plate solve without knowing any parameters of the image - it is called blind solving.

Takes a bit longer, but in principle - any scale, angle, and parity should be matched against catalog and result produced.

[Rich1980]

54 minutes ago, vlaiv said:

Not sure what exactly are you asking, but yes, you can plate solve without knowing any parameters of the image - it is called blind solving.

Takes a bit longer, but in principle - any scale, angle, and parity should be matched against catalog and result produced.

Sorry, I should've been clearer. 

I didn't know if there was a focal length maximum that would successfully plate solve. I plate solve at the moment at 550mm. Plate solving at 2800mm for instance seems quite a jump although yes it is doing the same process just in a smaller field of view. 

[vlaiv]

3 minutes ago, Rich1980 said:

Sorry, I should've been clearer. 

I didn't know if there was a focal length maximum that would successfully plate solve. I plate solve at the moment at 550mm. Plate solving at 2800mm for instance seems quite a jump although yes it is doing the same process just in a smaller field of view. 

Ok, get it now.

Plate solving works regardless of the scale, however, there are a few things to consider when doing very long focal length plate solving.

First is that you need a number of stars to plate solve on. Second is to have decent SNR on those stars so they can be identified as stars.

FOV gets into this equation because it restricts number of visible stars to plate solve on. Too constrained and you won't have any stars to plate solve on.

Other is SNR, and here its not so much focal length as sampling resolution. If you oversample by a lot - stars are no longer these points of light - they are "blobs" of light and plate solve software might not identify them as stars. Another issue is that you spread light over more pixels and each pixel gets less light as a result - SNR goes down and stars become fainter - another potential issue for star detection.

Luckily, there is way to solve over sampling issue - just use binning for exposures that you'll plate solve on (but in reality, you should also use binning for everything else if you are over sampling) and that will handle it.

I plate solve at 1600mm FL with 0.5"/px sampling resolution with no problem, so at least you know that down to 0.5"/px it should work ok.

[Rich1980]

40 minutes ago, vlaiv said:

Ok, get it now.

Plate solving works regardless of the scale, however, there are a few things to consider when doing very long focal length plate solving.

First is that you need a number of stars to plate solve on. Second is to have decent SNR on those stars so they can be identified as stars.

FOV gets into this equation because it restricts number of visible stars to plate solve on. Too constrained and you won't have any stars to plate solve on.

Other is SNR, and here its not so much focal length as sampling resolution. If you oversample by a lot - stars are no longer these points of light - they are "blobs" of light and plate solve software might not identify them as stars. Another issue is that you spread light over more pixels and each pixel gets less light as a result - SNR goes down and stars become fainter - another potential issue for star detection.

Luckily, there is way to solve over sampling issue - just use binning for exposures that you'll plate solve on (but in reality, you should also use binning for everything else if you are over sampling) and that will handle it.

I plate solve at 1600mm FL with 0.5"/px sampling resolution with no problem, so at least you know that down to 0.5"/px it should work ok.

Ah right, I see. 

Thanks for the detailed response.