Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Camera FOV Not Centered

  • Please log in to reply
16 replies to this topic

#1 Stargazer27

Stargazer27

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 08 Nov 2016
  • Loc: Midwest, USA

Posted 11 July 2020 - 03:16 PM

I'm using an ASI1600MM on an SCT. I'm noticing that the CMOS FOV center is not the same as the scope's FOV center, so I cannot properly use it to do alignments. How do people get around this kind of problem? If I remove my imaging kit and use eyepieces, the target is a bit in the top of the frame when reattaching. I'd like to use some of the cool software out there for centering alignment stars and annotation, but obviously cannot with this problem. 



#2 scadvice

scadvice

    Vanguard

  • *****
  • Vendors
  • Posts: 2,166
  • Joined: 20 Feb 2018
  • Loc: Lodi, California

Posted 11 July 2020 - 04:20 PM

Here I go again sounding like a salesman. Take a look at N.I.N.A. and the video's about your subject. For me it's been amazing how easy it was to learn this program and use it. I quite frankly have trouble turning a computer on let alone learn a program. Here is a post I did on NINA. If you don't want to read it all read post #7.

 

https://www.cloudyni...eing-set-upwow/

 

Below is a video about your subject using NINA.

 

https://www.youtube....v8apj_eXQ-tkuXG


Edited by scadvice, 11 July 2020 - 05:02 PM.

  • OldManSky likes this

#3 nimitz69

nimitz69

    Vanguard

  • *****
  • Posts: 2,431
  • Joined: 21 Apr 2017
  • Loc: A barrier island 18 miles south of Cocoa Beach

Posted 11 July 2020 - 04:41 PM

Learning to use plate solving is a basic skill in AP image capture. All the popular image capture s/w have plate solving functions which will center your target within pixels in your FOV.

And yes, NINA is a very easy to understand image capture s/w!
  • scadvice likes this

#4 rgsalinger

rgsalinger

    Fly Me to the Moon

  • *****
  • Moderators
  • Posts: 6,579
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 11 July 2020 - 04:42 PM

From reading your post, I think that you are concerned about centering targets since your chip is not perfectly centered in a manner identical to when you use an eyepiece. The answer, if I'm reading this correctly, is that it doesn't matter for photography at all. Whatever software you choose will have plate solving. The plate solving software will identify the center of the chip. Then the next slew you do will center the target every  time.  So, your shot will have the correct RA/DEC coordinates as its center. There's lots of software out there to choose from which does this. I recommend Sequence Generator Pro, myself as I found it easy to learn. You may not as I'd been imaging for a while when I sampled it.. 

 

There are a few caveats, though. First of all, if this is a new camera and the chip is not centered, return the camera. Period. Second of all, it's possible that the chip is so far off center that your filter wheel if going to be a problem because the filters will not cover the entire chip. This would seem amazing. 

 

I have the feeling though that what's actually happening is that you are slewing the scope and expecting, because you've done an alignment that things will be centered. They will not when using a long focal length telescope. The chip is pretty small in the ASI1600 so, even a slew that's even a couple of arc minutes off is going to look distinctly "un-centered" when you take the shot. 

 

Rgrds-Ross



#5 Stelios

Stelios

    Cosmos

  • *****
  • Moderators
  • Posts: 9,542
  • Joined: 04 Oct 2003
  • Loc: West Hills, CA

Posted 11 July 2020 - 06:38 PM

I'm using an ASI1600MM on an SCT. I'm noticing that the CMOS FOV center is not the same as the scope's FOV center, so I cannot properly use it to do alignments. How do people get around this kind of problem? If I remove my imaging kit and use eyepieces, the target is a bit in the top of the frame when reattaching. I'd like to use some of the cool software out there for centering alignment stars and annotation, but obviously cannot with this problem. 

Although they may not have the exact same point as center, if you star align using the camera and crosshairs you will be exactly as accurate for GoTo's as if you star align using an eyepiece. You simply have to do it for 2+2 (2+4 is better) stars at least. It's the *relative positioning* that matters. 

 

I agree with others that platesolving is the ultimate solution (and just so you don't feel forced to N.I.N.A. practically all capture software from BYEOS to APT, SGP, Voyager, etc. also support it). However, if you would rather take things slowly, just star align with the camera and the on-screen crosshairs. Later in the evening you may have to add a star close to your target to the alignment in order to have it in the FOV. 



#6 Stargazer27

Stargazer27

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 08 Nov 2016
  • Loc: Midwest, USA

Posted 11 July 2020 - 07:41 PM

Here I go again sounding like a salesman. Take a look at N.I.N.A. and the video's about your subject. For me it's been amazing how easy it was to learn this program and use it. I quite frankly have trouble turning a computer on let alone learn a program. Here is a post I did on NINA. If you don't want to read it all read post #7.

 

https://www.cloudyni...eing-set-upwow/

 

Below is a video about your subject using NINA.

 

https://www.youtube....v8apj_eXQ-tkuXG

Thank you so much for the suggestion. I’m actually checking it out now and I really like it. I’ve been looking for a software to stick with, and this may be it. I’m assuming you’re referencing the tutorials on Youtube to example the plate solving corrections? 

 

Learning to use plate solving is a basic skill in AP image capture. All the popular image capture s/w have plate solving functions which will center your target within pixels in your FOV.

And yes, NINA is a very easy to understand image capture s/w!

I understand what plate solving is. I just don’t fully understand how plate solving will help me to know correct center when I’m manually aligning the scope through the CMOS. I guess I’m missing something. I get that plate solving can correct your alignment, but how do I use visual tools for polar alignment when the star I’m polar aligning to isn’t centered on the sensor? If I were to use those tools and center them via the CMOS, they would be off center by the difference between sensor center vs scope center FOV. Do you see what I’m saying? It’s probably a misunderstanding on my part, so if you can see where I’m confused, please say so. 

 

From reading your post, I think that you are concerned about centering targets since your chip is not perfectly centered in a manner identical to when you use an eyepiece. The answer, if I'm reading this correctly, is that it doesn't matter for photography at all. Whatever software you choose will have plate solving. The plate solving software will identify the center of the chip. Then the next slew you do will center the target every  time.  So, your shot will have the correct RA/DEC coordinates as its center. There's lots of software out there to choose from which does this. I recommend Sequence Generator Pro, myself as I found it easy to learn. You may not as I'd been imaging for a while when I sampled it.. 

 

There are a few caveats, though. First of all, if this is a new camera and the chip is not centered, return the camera. Period. Second of all, it's possible that the chip is so far off center that your filter wheel if going to be a problem because the filters will not cover the entire chip. This would seem amazing. 

 

I have the feeling though that what's actually happening is that you are slewing the scope and expecting, because you've done an alignment that things will be centered. They will not when using a long focal length telescope. The chip is pretty small in the ASI1600 so, even a slew that's even a couple of arc minutes off is going to look distinctly "un-centered" when you take the shot. 

 

Rgrds-Ross

Can you explain how plate solving replaces manual alignment during a session? I’m missing something. For instance, if the scope is perfectly polar aligned, and the alignment star is directly center In the SCT’s FOV, but not center in the CMOS sensor, wouldn’t the plate alignment images be misaligned for correction processing?

 

Also, related to what you said, “The plate solving software will identify the center of the chip.” If the plate solving software is identifying the center of the chip, and making corrections to the model with that info, wouldn’t that cause the mount to actually track slightly off center? I think you can see where I’m missing details. I can, but I can’t figure out what it is. 

 

Although they may not have the exact same point as center, if you star align using the camera and crosshairs you will be exactly as accurate for GoTo's as if you star align using an eyepiece. You simply have to do it for 2+2 (2+4 is better) stars at least. It's the *relative positioning* that matters. 

 

I agree with others that platesolving is the ultimate solution (and just so you don't feel forced to N.I.N.A. practically all capture software from BYEOS to APT, SGP, Voyager, etc. also support it). However, if you would rather take things slowly, just star align with the camera and the on-screen crosshairs. Later in the evening you may have to add a star close to your target to the alignment in order to have it in the FOV. 

I had always assumed by way of how accurate polar eyepieces are that you needed to have the star perfectly centered for best alignment. Even with that accuracy and manually aligning via an eyepiece I’m still having trouble with tracking accuracy. I align, polar align, align, polar align, and there are always more corrections to make when polar aligning each time. Ever since I upgraded to a dedicated astrocam from my DSLR I’ve had a heck of a time getting any imaging done because of this alignment problem . I’ve taken to starting the night with a 35mm and a illuminated reticle eyepiece to do my alignment. I let StarSense do its 4 plate alignment, and then center in on the polar star with the illuminated reticle and make my corrections. After that you have to align again, and I always see some error. 


Edited by Stargazer27, 11 July 2020 - 07:45 PM.


#7 OldManSky

OldManSky

    Skylab

  • *****
  • Posts: 4,308
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 11 July 2020 - 10:27 PM

"Can you explain how plate solving replaces manual alignment during a session? I’m missing something. For instance, if the scope is perfectly polar aligned, and the alignment star is directly center In the SCT’s FOV, but not center in the CMOS sensor, wouldn’t the plate alignment images be misaligned for correction processing?"

 

With plate solving, you don't need to do a "star alignment."  Manual or otherwise.

You tell the scope to slew to a target (which has a particular set of RA/DEC coordinates).  The scope will likely NOT be pointing in the right place after that first slew.

That you initiate a plate solve.  The software takes an image, and plate solves it -- which figures out exactly where the scope is pointing, instead of where it thought it was pointing.

Then the software tells the mount, essentially:  "Hey, you're not THERE, you're HERE.  Synch yourself to the HERE coordinates."

Now the mount knows exactly where it is, and that is accurate to a few pixels on your imaging chip.

 

That's it, you're done.

 

Want to slew somewhere else?  Do the same thing.  It will be better because the mount already knows where it is, but even if your polar alignment is off, or you have slop in your gears, or just a mediocre mount that doesn't point well, the same thing will repeat.  The mount will slew to where it thinks you wanted to go.  A plate solve will figure out exactly where it is.  Then if you have "recenter" on in your software, the software will move the mount after the first plate solve to exactly where you wanted it to go, plate solve again, and repeat until it's exactly where you told it to go.  All of that usually takes just a few seconds.  

 

Seriously, it's a game changer.  Get a decent polar alignment, then plate solve -- and don't worry about pointing the rest of the night.  Every single slew to a target, with plate solve, synch, and recenter on, will wind up exactly where you wanted it to.  Without you having to spend all kinds of time star aligning.


  • mtc and scadvice like this

#8 Phil Sherman

Phil Sherman

    Gemini

  • *****
  • Posts: 3,140
  • Joined: 07 Dec 2010
  • Loc: Cleveland, Ohio

Posted 11 July 2020 - 10:30 PM

If you're having problems with polar alignment, have you considered trying a completely different technique? I believe that PHD2 will use any ASCOM camera which should allow it to use your ASI camera as a "guide" camera with PHD2's polar alignment routine. Once polar aligned, you can shut down PHD2 and switch to your imaging program to control the camera.



#9 Stargazer27

Stargazer27

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 08 Nov 2016
  • Loc: Midwest, USA

Posted 11 July 2020 - 10:51 PM

"Can you explain how plate solving replaces manual alignment during a session? I’m missing something. For instance, if the scope is perfectly polar aligned, and the alignment star is directly center In the SCT’s FOV, but not center in the CMOS sensor, wouldn’t the plate alignment images be misaligned for correction processing?"

 

With plate solving, you don't need to do a "star alignment."  Manual or otherwise.

You tell the scope to slew to a target (which has a particular set of RA/DEC coordinates).  The scope will likely NOT be pointing in the right place after that first slew.

That you initiate a plate solve.  The software takes an image, and plate solves it -- which figures out exactly where the scope is pointing, instead of where it thought it was pointing.

Then the software tells the mount, essentially:  "Hey, you're not THERE, you're HERE.  Synch yourself to the HERE coordinates."

Now the mount knows exactly where it is, and that is accurate to a few pixels on your imaging chip.

 

That's it, you're done.

 

Want to slew somewhere else?  Do the same thing.  It will be better because the mount already knows where it is, but even if your polar alignment is off, or you have slop in your gears, or just a mediocre mount that doesn't point well, the same thing will repeat.  The mount will slew to where it thinks you wanted to go.  A plate solve will figure out exactly where it is.  Then if you have "recenter" on in your software, the software will move the mount after the first plate solve to exactly where you wanted it to go, plate solve again, and repeat until it's exactly where you told it to go.  All of that usually takes just a few seconds.  

 

Seriously, it's a game changer.  Get a decent polar alignment, then plate solve -- and don't worry about pointing the rest of the night.  Every single slew to a target, with plate solve, synch, and recenter on, will wind up exactly where you wanted it to.  Without you having to spend all kinds of time star aligning.

There's what I was missing! I have used solving to some degree for fun/annotation and post-processing, but didn't realize how it was used during tracking. Thanks so much. I've been frustrated enough the last couple of weeks I could have chucked the scope over the neighbors fence. I can't wait to try it out. But now that I know, a storm just rolled in, as is customary. I'm very truly grateful for the advice.

 

If you're having problems with polar alignment, have you considered trying a completely different technique? I believe that PHD2 will use any ASCOM camera which should allow it to use your ASI camera as a "guide" camera with PHD2's polar alignment routine. Once polar aligned, you can shut down PHD2 and switch to your imaging program to control the camera.

I did not know that PHD2 had a polar alignment routine. Because of the craziness with alignment and a bad CGEM II mount, I have only had the opportunity to check out PHD2 about 2 times so far since I got my guide cam. I'll definitely check that out. I saw today that SharpCap also has a polar alignment technique. I might have to see about that as well. I saw a software that does what appears to be live solving and annotation here: https://youtu.be/diM...2-oBYI3A?t=240 

 

Does anyone know what that is? It's too blurry for me to read.



#10 TelescopeGreg

TelescopeGreg

    Surveyor 1

  • -----
  • Posts: 1,905
  • Joined: 16 Jul 2018
  • Loc: Auburn, California, USA

Posted 11 July 2020 - 11:18 PM

Actually, it can be even easier than that.  I'm using CCDciel + ASTAP for my at-mount control.  I tell it where I want the scope to point to, and it looks up the coordinates, does a slew, takes a picture, plate solves, tells the mount (sync), and assuming it's not exactly where I wanted it to be, it does another slew and repeats the process until it's done (within an error margin I can define).

 

Magical to watch it work... 

 

Should I want the actual target to be slightly offset from the catalog position (for better framing), I can then right-click on the last image and tell it to center the scope there.

 

The next night, I can repeat the final positioning by loading up one of the prior night's subs and tell the software to center the scope where that one was.

 

Game, changed.


  • OldManSky likes this

#11 Stargazer27

Stargazer27

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 08 Nov 2016
  • Loc: Midwest, USA

Posted 12 July 2020 - 05:39 PM

Actually, it can be even easier than that.  I'm using CCDciel + ASTAP for my at-mount control.  I tell it where I want the scope to point to, and it looks up the coordinates, does a slew, takes a picture, plate solves, tells the mount (sync), and assuming it's not exactly where I wanted it to be, it does another slew and repeats the process until it's done (within an error margin I can define).

 

Magical to watch it work... 

 

Should I want the actual target to be slightly offset from the catalog position (for better framing), I can then right-click on the last image and tell it to center the scope there.

 

The next night, I can repeat the final positioning by loading up one of the prior night's subs and tell the software to center the scope where that one was.

 

Game, changed.

Wow! I'll try this tonight too. And most excellently, I can use CCDciel and ASTAP natively on MacOS. I can't do that with N.I.N.A. as far as I can tell unfortunately. I have a decent setup right now where I use an external Thunderbolt 3 NVME drive with my MacBook Pro at ridiculous external transfer speeds of about 2,700MB/s to capture and store images on my remote rig. I can then plug that drive into my iMac for processing in PixInsight or APP and work with the images instantly with no transfer times. Thank you!


Edited by Stargazer27, 12 July 2020 - 05:39 PM.

  • TelescopeGreg likes this

#12 Stargazer27

Stargazer27

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 08 Nov 2016
  • Loc: Midwest, USA

Posted 13 July 2020 - 01:41 PM

Actually, it can be even easier than that.  I'm using CCDciel + ASTAP for my at-mount control.  I tell it where I want the scope to point to, and it looks up the coordinates, does a slew, takes a picture, plate solves, tells the mount (sync), and assuming it's not exactly where I wanted it to be, it does another slew and repeats the process until it's done (within an error margin I can define).

 

Magical to watch it work... 

 

Should I want the actual target to be slightly offset from the catalog position (for better framing), I can then right-click on the last image and tell it to center the scope there.

 

The next night, I can repeat the final positioning by loading up one of the prior night's subs and tell the software to center the scope where that one was.

 

Game, changed.

Well I guess I jumped the gun on getting excited. I can't get plate solving to work for me at all. I'm investigating, but I lost another clear night to failures with that and eventually, DEW. ASTAP just keeps telling me it has reached a star magnitude some number of times. I've tried in APT and SharpCap. :/ If its not one thing its another. Sheesh.



#13 Stargazer27

Stargazer27

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 08 Nov 2016
  • Loc: Midwest, USA

Posted 13 July 2020 - 01:42 PM

If you're having problems with polar alignment, have you considered trying a completely different technique? I believe that PHD2 will use any ASCOM camera which should allow it to use your ASI camera as a "guide" camera with PHD2's polar alignment routine. Once polar aligned, you can shut down PHD2 and switch to your imaging program to control the camera.

I tried this, but it can't solve the plate. No matter what I do its just not working out with plate solving. 



#14 TelescopeGreg

TelescopeGreg

    Surveyor 1

  • -----
  • Posts: 1,905
  • Joined: 16 Jul 2018
  • Loc: Auburn, California, USA

Posted 13 July 2020 - 02:14 PM

Well I guess I jumped the gun on getting excited. I can't get plate solving to work for me at all. I'm investigating, but I lost another clear night to failures with that and eventually, DEW. ASTAP just keeps telling me it has reached a star magnitude some number of times. I've tried in APT and SharpCap. :/ If its not one thing its another. Sheesh.

Deep breath.  Relax.  Very likely just a couple of settings. 

 

ASTAP needs to know what your camera pixel size is, and what the focal length of the scope it's attached to is.  That gives it an idea of the scale of the image it needs to solve.  I also found that I needed reduce the image, in order to keep it withing ASTAP's memory capability (I'm running mine on a Raspberry Pi).  There's a "Downsample" setting under Preferences / Astrometry.  I think the final image going from CCDciel to ASTAP was something like 1,500 x 900 pixels, vs the 6,000 x 4,000 from the camera. 

 

I presume that CCDciel has a connection to the mount as well as the camera.  It needs the mount connection for slewing, but also to know where the mount thinks it's pointing, to give to ASTAP for a starting point for its search.



#15 OldManSky

OldManSky

    Skylab

  • *****
  • Posts: 4,308
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 13 July 2020 - 07:27 PM

Greg gave good advice above.

Also, you don't need a very long exposure for the image you're going to plate solve -- just long enough to show stars.  I typically use a 5-second exposure (at gain 50) with my 183mm-pro camera.  That solves anywhere in the sky with ASTAP, even with only a dozen or so stars visible.



#16 Stargazer27

Stargazer27

    Sputnik

  • -----
  • topic starter
  • Posts: 28
  • Joined: 08 Nov 2016
  • Loc: Midwest, USA

Posted 16 July 2020 - 05:17 PM

Deep breath.  Relax.  Very likely just a couple of settings. 

 

ASTAP needs to know what your camera pixel size is, and what the focal length of the scope it's attached to is.  That gives it an idea of the scale of the image it needs to solve.  I also found that I needed reduce the image, in order to keep it withing ASTAP's memory capability (I'm running mine on a Raspberry Pi).  There's a "Downsample" setting under Preferences / Astrometry.  I think the final image going from CCDciel to ASTAP was something like 1,500 x 900 pixels, vs the 6,000 x 4,000 from the camera. 

 

I presume that CCDciel has a connection to the mount as well as the camera.  It needs the mount connection for slewing, but also to know where the mount thinks it's pointing, to give to ASTAP for a starting point for its search.

I'm ok LOL. I have everything set in ASTAP. I even tried directly through it vs. using an imaging program. No matter what I do in the application, it will not plate solve. It does plate solve in N.I.N.A however (using ASTAP). It just works. The same settings as the config in NINA fail in ASTAP directly. I gave up.

 

I am having a little bit of trouble though. I have mostly gotten the plate solving issues down, and did actually get some imaging done right towards sunrise. The ring nebula. However, if you see my image, there is quite a coma. I am using a 6.3 focal reducer/corrector here, and the sensor needs to be 105mm from the back of the corrector. It is. I'm measuring with the actual mm specs from each item in between, and as a secondary assurance, with a metric caliper. Right on the money. It's quite strange. It is the red channel, stacked and stretched.

 

M57 Red Channel Stacked and Stretched CN

 

 

Greg gave good advice above.

Also, you don't need a very long exposure for the image you're going to plate solve -- just long enough to show stars.  I typically use a 5-second exposure (at gain 50) with my 183mm-pro camera.  That solves anywhere in the sky with ASTAP, even with only a dozen or so stars visible.

The exposure I was using for plate solving is 2 seconds. It's the default in NINA. 



#17 TelescopeGreg

TelescopeGreg

    Surveyor 1

  • -----
  • Posts: 1,905
  • Joined: 16 Jul 2018
  • Loc: Auburn, California, USA

Posted 17 July 2020 - 12:01 AM

So, I'm not using ASTAP directly; it's called from CCDciel which takes the image and (I presume) does the image down-scaling before calling ASTAP with the image and where in the sky the mount claims it was taken from.  That prevents it from being a blind solve, which makes it much more likely to succeed.




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics