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

PSWAI: A DIY Plate Solved Push To for Visual Astronomy

DIY CMOS equipment astrophotography planetarium software
  • Please log in to reply
17 replies to this topic

#1 hcf

hcf

    Viking 1

  • -----
  • topic starter
  • Posts: 878
  • Joined: 01 Dec 2017

Posted 20 December 2018 - 05:36 PM

Plate solving has been used by AP/EAA ers for some time now to acquire/frame targets. This is an attempt to use a similar technology for visual astronomers.

 

Summary:

Take images with an inexpensive camera mounted on and aligned to a scope, plate solve them for location and hook it up to Sky Safari to display the current location the scope is pointed to in the sky, a Plate Solved Where Am I (PSWAI). I used the camera below, but the same techniques can be used with wide angle lens on other astronomy cameras as well.

 

The Camera:

A lot of people have been trying to use cheap security webcams for EAA with mixed results. Searching for such a cam landed me on the dashcamtalk pages of the Xiaomi Yi Action Cam a 2-3 year old action camera which is very accessable/hackable running a simple linux like kernel inside. It has a IMX206 or Panasonic MN34120PA, 1/2.3", 16MP CMOS image sensor. Functional modification includes adding an autoexec.ash file on the sd card in the camera. There are newer versions (4K) of this camera, but I am not sure they are as hackable.

 

Some related links to Xiaomi Yi Action Camera hacking:

 

https://www.tawbaware.com/xiaomiyi.htm
https://dashcamtalk....ion-camera.153/
https://nutseynuts.b...om-scripts.html

 

The Good:

 

  • Can control shutter speed up to 8sec. ISO upto 25600.
  • Can store images/videos on a micro sd card.
  • Has Wifi connectivity and a not very useful USB connectivity.
  • You can telnet (login) to it over wifi.
  • You can run a ftp server on it for file transfer over wifi.
  • Can enable a rtsp low res live view feed over wifi.
  • Has a hidden RAW+jpeg mode (not needed for this project). The RAW is without headers so you need something like Raw2Dng or Raw2Nef to convert it to a format, that processing apps like Deep Sky Stacker can use.
  • The battery is low capacity, but the camera works when attached to power through USB.
  • Light weight.
  • Hackable, lens can be replaced without too much trouble.
  • Has an app to control settings, take pictures.

 

The Bad:

  • The USB access is useless as connecting a computer to USB puts the camera into storage mode where no pictures can be taken.
  • Wifi is 2.4GHz and a very small range.
  • The battery is low capacity, and you need an external pack. They do sell piggyback powerpacks for it.
  • The included fisheye lens is not very useful.
  • Tiny pixels (1.35 um)

 

The Build:

 

  • A Xiaomi Yi Action Camera ($50-$60)
  • A replacement 50mm F/1.2 Lens with a M12 thread  ($23) The included lens can be removed by taking the front cover off,and removing a little of the glue attaching it to the lens holder (look for youtube videos). This lens gives you about 7x5 degrees FOV which has enough stars for platesolve to work. The Camera has very deep M12 threads so normal webcam adapters don't work. But the M12 lens above works and has enough thread to adjust focus. It can possibly be glued once focused, but I did not need that. I got a lens without an IR filter, there are ones which come with a IR filter too. I found a way to add an IR filter as described later.
  • A light ball head mount ($5) to be able to easily move the camera to align with the scope. Not the best way to do this but works.
  • A linux/unix machine with astrometry.net plate solver downloaded and working. I use a Raspberry Pi 3B. It is better if the machine can do 2 network connections, one for the camera and one for sky safari. This is the fastest, as the camera has a wifi hotspot and connecting to it is the fastest way to download images. I use the wifi on a Raspberry Pi to connect to the Yi, and the ethernet connected to a travel router for connecting to sky safari and my home LAN.
  • A smartphone/tablet used to setup the camera settings and to run Sky Safari. It needs the Yi app (free) and Sky Safari Pro/Plus(?).

Optional:

  • An optional border frame. This comes with a 37mm thread and filter, you can get a 37-42mm adapter which lets you attach it to a T-adapter. This kind of mounting is much better if you want to connect the camera to a scope but it is not a must for this project. I use it to put a IR filter in front of the lens. If you get one of these frames, use a rubber band  to secure it to the camera for times when you are taking the camera on/off a mount which requires taking off the 1/4-20 screw attaching the frame to the camera.
  • Orion 7033 Precision Slow-Motion Adapter (you may find used ones for around $30). This makes it a lot easier to align the camera with the scope.
  • A wiimote. This is an easy way to add a remote to a pi. You can also take images from a command line on a VNC session, or even run in a loop without a trigger.
 
Yi with a 50mm F/1.2 lens
 
Yi with frame,lens,Tadapter,IRfilter
 

 

 
Software:

You need to get astrometry.net working locally on the machine, their website/github page has instructions. The rest are just scripts to take a picture, platesolve, and update sky safari. I use the Basic Encoder System on SkySafari and a GEM mount in the settings. I found that a smaller jpg for a 2.7s exposure can be solved  in my Bortle 7 skies in about 20s on Raspberry Pi 3B. A 2.7s exposure on a 50mm lens on this camera produces round stars (so that platesolve works) even on non tracking mounts like my 8" Dob. Depending on your skies you might need to tweak the shutter speed/ISO settings.  I have solve-field (from astrometry.net) dump the RA,Dec cooordinates to a file, and a python script faking an EQ encoder  pick it up from there. Although I use astrometry.net, any plate solver which can be run as a command line and dump the center RA,Dec, can be used. The scripts are shell/python scripts which can run on linux/unix like computers. The wiimote, connected via bluetooth to the Pi is just an easy way to take a picture. I use the rumble mode on the wiimote to provide feedback if/when the command completed, one rumble for success, two rumbles for failure.

 

Pictures of the camera mounted on my scopes.

 

camdob
 
cam280
 

Workflow:

   Prep off the field:

  • Start the camera with the appropriate autoexec.ash file in the sd card.
  • Connect to the camera on a phone app and select image size, turn leds off, turn off fisheye correction, setup auto wifi etc etc. All these are sticky across power cycles, so you need to do them the first time only.

   On the field:

  • Align the camera to the scope. The phone app has a low res live view on which you can enable a 3x3 grid. Get a bright star in your scope and then move the ball head to get the star in the middle square of the 3x3 view.  Doing this with a ball head without fine tune controls is the hardest part of the workflow. If you have the Orion Precision  Slow-Motion Adapter it is much easier.
  • Start the software on the Pi/computer you use, connect the wiimote and skysafari.
  • Do a 2 star alignment on sky safari. This is to teach Sky safari the correspondence between the encoder values produced and the ra,dec of the star. One star alignment is enough, two is better. It also helps a little for EQ mounts if your alignment is off a little. You need to get a bright star/planet in the scope EP, click the wiimote to take a pic, wait for the computer to platesolve it, and update sky safari and once updated, align sky safari to that star/planet
  • Point the telescope where you want, click the wiimote, wait for the rumble and look at sky safari for your position.

 

Without getting into the nitty gritty, here is a picture of my VNC session on the Pi.

 

vnc.label

 

Performance & Accuracy:

On the Pi 3B, I can turn around the taking an picture to updating sky safari with the coordinates at 18-30 seconds depending on the night and sky conditions. On a laptop running ubuntu it is can be a little faster (the best I saw was 12 sec) . I even tried a Pi Zero W ($10) where it is about 40-50 seconds.  Accuracy depends on your alignment to the scope and the lag. On my 8" Dob (camera mounted on the RACI with a hose clamp), the ES82 18mm EP (67x) is close with good alignment to the scope. Failing that I try to match it up in the 8x50 RACI which is easier. Have also used it on my Meade Adventure 80 (mounted on the rings) on the EXOS Nano EQ mount.

 

 

Pros and Cons:

 

   Cons first:

  • The most obvious one is the lag between movement and update.  If you are used to a encoder based pushto it will drive you nuts.
  • Needs a mount side computer to work, although a small one like a raspberry pi will do.
  • A lot of different hardware software components. When things go wrong needs a laptop to debug.

   Pros:

  • Can be attached to most telescopes, Alt Az/EQ, from large Dobs to department store telescopes, (maybe even to Binos?). Usual camera 1/4-20 mount.
  • Is lightweight enough not to cause too many mounting porblems.
  • No wires needed to connect the computer to the camera. The camera will need a power source connect due to the terrible battery life.
  • Is portable across telescopes. Just need one camera. You can move it from one telescope to another very easily.
  • Easy to use. You can move the scope with slow motion, by loosening the clutches, by motors, whatever way you want to, as the camera is mounted on the OTA the alignment will hold.
  • Reasonably accurate.

A sample jpg of a 2.7 second exposure ISO 200 without an IR filter.

 

A single jpg capture (reduced for upload).

 

This is intended more as an aid to starhopping rather than a replacement for a proper encoder based pushto. Once you have one, installing it on a new scope is much easier than an encoder based DSC. You can use it to get close to the DSO in question, and then use the finder/EP. It also helps for faint DSOs to make sure you are at the right spot.

 

And if you want to do some basic AP with it, the other day after finding 46P/Wirtanen on my 80mm F/5 refractor on the EXOS Nano EQ mount with the PSWAI, barely visible in my LP skies, I took some shots with the PSWAI (same 50mm F/1.2 lens). The mount was approximately polar aligned (I have no view of polaris) and tracking. 40x8" lights (ISO 1600 with an IR filter)  with darks, and bias stacked and processed a little in Deep Sky Stacker. The comet is moving fast, so is a small line over the duration of all the shots.

 

46P/Wirtanen

 

 

I put my scripts on github here in case anyone is interested. Some more details on the scripts there.

 

C&C welcomed, and thanks for reading.


Edited by hcf, 21 December 2018 - 03:03 PM.

  • Dave Mitsky, PrestonE, lambermo and 13 others like this

#2 Adun

Adun

    Mercury-Atlas

  • *****
  • Posts: 2,803
  • Joined: 02 Dec 2016

Posted 20 December 2018 - 10:59 PM

Very useful project!

 

I tried to do something like this with a RPi 3, guidescope, and an OV2710 USB camera, but the camera turned out to not be sensitive enough, so I ended up switching to a RisingTech EAA camera, which did the trick, but bound me to windows.


  • hcf likes this

#3 hcf

hcf

    Viking 1

  • -----
  • topic starter
  • Posts: 878
  • Joined: 01 Dec 2017

Posted 21 December 2018 - 12:14 AM

Very useful project!

 

I tried to do something like this with a RPi 3, guidescope, and an OV2710 USB camera, but the camera turned out to not be sensitive enough, so I ended up switching to a RisingTech EAA camera, which did the trick, but bound me to windows.

Thanks.

 

I tried this with my AR0130 based webcam a while back, but the maximum exposure of 0.5s was the big limiting factor. So when I saw shutter speed upto 8s on the Yi, with a larger sensor, I jumped at it, even though it has tiny pixels. Look at the Yi forums for all the possible settings, they are amazing.  And it does not cost much more than the ELP AR0130 based webcam.

 

RisingTech needs to wake up to making drivers/software for ARM chips/linux environments.



#4 robertasumendi

robertasumendi

    Analog Sky Giant Space Binoculars

  • *****
  • Vendors
  • Posts: 56
  • Joined: 27 May 2017
  • Loc: Springfield, Oregon USA

Posted 21 December 2018 - 12:34 AM

Nice project! There definitely needs to be a device like this for visual use.

 

Are you optimizing the star database to stars in the mag 6-7 range or are you just using the whole dataset?

 

Just to be clear, are you using RPi3B or RPi3B+?

 

Are you processing 8 MP images or binning them down to 1-2 MP and processing those?

 

Thanks for sharing!

 

Robert


Edited by robertasumendi, 21 December 2018 - 12:42 AM.

  • hcf likes this

#5 hcf

hcf

    Viking 1

  • -----
  • topic starter
  • Posts: 878
  • Joined: 01 Dec 2017

Posted 21 December 2018 - 01:54 AM

Nice project! There definitely needs to be a device like this for visual use.
 
Are you optimizing the star database to stars in the mag 6-7 range or are you just using the whole dataset?


Thanks! Thats a good point, I am not optimzing the database at all, I have all the index files. I do use the flags
--scale-units degwidth --scale-low 3 --scale-high 8  to speed up search.
I think I could do with fewer/smaller index files, but I use the same scripts for different setups, like
with my DSLR during AP as well. Something to look at for sure. Have you done something like this successfully?
 

Just to be clear, are you using RPi3B or RPi3B+?


Its a RPI 3B (not a RPI 3B+)
 

Are you processing 8 MP images or binning them down to 1-2 MP and processing those?
 
Thanks for sharing!
 
Robert


This is just from the image size settings on the Yi. The camera has a 16MP sensor. The 8MP image (second smallest) is a jpeg image (not RAW) 3264x2448 pixels about 2.1MB in size. I think it is just jpeg compression/scaling or whatever the camera does. The RAW files are always the max size (about 30MB), so I don't think any binning is going on.

The file transfer over wifi can be slow so I tried to keep the image size small for the plate solving use case.
I do not bin the images externally. The astrometry.net platesolver has a --downsample flag, but I found
that slows things down in this case, possibly because it writes to the slower sd card.


For the last picture with 40x8" lights, I used RAWs which use different settings in the autoexec.ash file, saved them on the uSD card and post processed them later in DSS.

Edited by hcf, 21 December 2018 - 02:05 AM.


#6 Adun

Adun

    Mercury-Atlas

  • *****
  • Posts: 2,803
  • Joined: 02 Dec 2016

Posted 21 December 2018 - 10:50 AM

RisingTech needs to wake up to making drivers/software for ARM chips/linux environments.


Yeah, that would be the dream!

 

Thanks! Thats a good point, I am not optimzing the database at all, I have all the index files. I do use the flags
--scale-units degwidth --scale-low 3 --scale-high 8  to speed up search.
I think I could do with fewer/smaller index files, but I use the same scripts for different setups, like
with my DSLR during AP as well. Something to look at for sure. Have you done something like this successfully?
 
The file transfer over wifi can be slow so I tried to keep the image size small for the plate solving use case.
I do not bin the images externally. The astrometry.net platesolver has a --downsample flag, but I found
that slows things down in this case, possibly because it writes to the slower sd card.

 
The biggest optimization has already been done and it's the use of the 50mm FL lens, since the 7x5 degrees FOV will make astrometry.net use wider field indexes, and those are much smaller files, quicker to read (and to process). I get much faster platesolves from my 25mm FL CS lens than with my 190mm FL guidescope, on the same camera & computer.
 
From my own experimentation, what I've found to be the other useful performance tuning points are:
 
 
#1, biggest improvement came from providing the pixel scale :
 

--scale-units arcsecperpix --scale-low 0.386 --scale-high 0.406

After using this with the actual pixel scale obtained from the first blind platesolve (+/- 5%), things quite sped up, specially at 190mm FL.
 
#2, ensuring that astrometry.cfg has the "inparallel" option active. Since finder use involve wider fields, the small indexes do fit in memory without problem. Maybe for the widest fields this doesn't help as much, but for 190mm it sure does. It also depends on the computer, so an original RPi (1) won't benefit as much as a RPi 3
 
#3, for some reason, "--downsample 2" helped my 190mm setup cut a couple seconds. Not so with the 25mm FL CS lens. And "--downsample 4" actually takes longer. Go figure.
 
#4, skipping the generation of the plots with --no-plots  (might not be necessary depending on the astrometry build).
 
I was surprised that when I tried using --ra, --dec, --radius to speed things up (using the coordinates of a previous platesolve and a radius of -say- 10º) it did not produce a noticeable speed improvement. I was expecting this to be the #1 tuning factor, but it hasn't. I'm still trying to cut a few more seconds from the platesolving time though.


Edited by Adun, 21 December 2018 - 10:54 AM.

  • hcf likes this

#7 hcf

hcf

    Viking 1

  • -----
  • topic starter
  • Posts: 878
  • Joined: 01 Dec 2017

Posted 21 December 2018 - 01:30 PM

#3, for some reason, "--downsample 2" helped my 190mm setup cut a couple seconds. Not so with the 25mm FL CS lens. And "--downsample 4" actually takes longer. Go figure.
 
#4, skipping the generation of the plots with --no-plots  (might not be necessary depending on the astrometry build).
 
I was surprised that when I tried using --ra, --dec, --radius to speed things up (using the coordinates of a previous platesolve and a radius of -say- 10º) it did not produce a noticeable speed improvement. I was expecting this to be the #1 tuning factor, but it hasn't. I'm still trying to cut a few more seconds from the platesolving time though.


Yes --no-plots is a must. I have used --downsample in the past when I used these scripts with my DSLR to frame DSOs for AP/EAA. --downsample reduces the number of stars which seems to have an effect on improving the speed. For this project I found 30-100 stars are the best. Then I found the --sigma flag which sets the noise level. I found that this also reduces the number of stars, and is faster than --downsample on the Pi.



#8 hcf

hcf

    Viking 1

  • -----
  • topic starter
  • Posts: 878
  • Joined: 01 Dec 2017

Posted 29 December 2018 - 11:42 AM

Added INDI support using the indi_dsc_telescope driver. Now, in addition to Sky Safari,  Kstars can be used to view the location. Once the alignment in Kstars/Ekos has been done, the INDI driver can be used by Stellarium as well using the telescope plugin (Could not find a way to align the indi_dsc_telescope driver from Stellarium directly).

 

This allows you do run the platesolving scripts and Kstars/Stellarium on one linux laptop connected to the Camera Wifi hotspot.

 

Updated github.

 

 



#9 hcf

hcf

    Viking 1

  • -----
  • topic starter
  • Posts: 878
  • Joined: 01 Dec 2017

Posted 24 January 2019 - 03:16 AM

Added an option to distinguish between Tracking and Non Tracking mounts.  This makes the position of the FOV on SkySafari/Kstars/Stellarium more accurate in between the shots.

 

Running Indi/Kstars/Stellarium on an ubuntu laptop, the lag is about 10s now.  Recently had a session where I used the 11mm EP on the Dob, and had it accurate enough to be usable.

 

Updated github.


  • Bill Kocken likes this

#10 hcf

hcf

    Viking 1

  • -----
  • topic starter
  • Posts: 878
  • Joined: 01 Dec 2017

Posted 14 May 2020 - 11:59 AM

Recently added a new LX200 based interface to the PSWAI based on my work on the PS-G2.

 

This lets you start the platesolving from SkySafari itself and not use an external remote control like the wiimote.

 

Some more details on the github page.

 

 


  • cmooney91 likes this

#11 synfinatic

synfinatic

    Viking 1

  • *****
  • Posts: 794
  • Joined: 22 Dec 2013
  • Loc: San Jose, CA

Posted 14 November 2020 - 05:48 PM

wow, this is really cool.  thanks for sharing!


  • hcf likes this

#12 hcf

hcf

    Viking 1

  • -----
  • topic starter
  • Posts: 878
  • Joined: 01 Dec 2017

Posted 14 November 2020 - 08:44 PM

wow, this is really cool.  thanks for sharing!

Thanks !

 

That reminds me, some other related news I should link in here. Initially I ran the platesolver on raspberrypi 3 and then on an Android TV Box (4GB RAM) running Armbian.  I have also been able to run the platesolver on two environments that can run both the platesolver and SkySafari on one box.

 


Edited by hcf, 14 November 2020 - 08:46 PM.


#13 metalpoet

metalpoet

    Mariner 2

  • -----
  • Posts: 259
  • Joined: 20 Jul 2018
  • Loc: Cape Town, South Africa

Posted 21 February 2021 - 03:47 PM

I have a question: could this be adapted for Alt- az mounts like Dobsonians ?

That could turn it into a replacement for the Nextstar Explorer app without the need to buy a whole scope just to get the app.

Edited by metalpoet, 21 February 2021 - 03:48 PM.


#14 hcf

hcf

    Viking 1

  • -----
  • topic starter
  • Posts: 878
  • Joined: 01 Dec 2017

Posted 21 February 2021 - 04:45 PM

Yes, I use it on a Dob as well (see picture in the first post). Because the camera is aligned to the scope, it shows you in Sky Safari where the scope (center of image taken by the camera ) is pointed to, irrespective of the mount.

 

The Celestron StarSense explorer, also uses the phone's sensors to show you real time changes in position, between platesolves. The PSWAI on the other hand is not real time. It takes 10-20 seconds (depending on the platesolving computer) to show you the location, after you hit the button on Sky Safari. It does adjust the position based on whether the mount tracks or not. So while you can use it as a pushto, it is much more convenient to get close to the target using the PSWAI and star hop from there. This is especially good for light polluted skies where there are few naked eye visible stars. The camera can pick up a lot more.

 

The other advantages are because you do not have to mount the phone, you can use Sky Safari on a tablet you hold in your hand. And it is not limited to the brightest DSOs, I have never failed to get to any DSO, NGC or Messier that I have tried from Sky Safari. You do have to adjust the exposure time and ISO on some nights. With good alignment of the Camera and my 8" Dob, I can match stars in the FOV between the location Sky Safari shows with the PSWAI and my ES82 18mm EP in Bortle 7 skies.


Edited by hcf, 21 February 2021 - 05:33 PM.


#15 metalpoet

metalpoet

    Mariner 2

  • -----
  • Posts: 259
  • Joined: 20 Jul 2018
  • Loc: Cape Town, South Africa

Posted 21 February 2021 - 10:11 PM

Yes, I use it on a Dob as well (see picture in the first post). Because the camera is aligned to the scope, it shows you in Sky Safari where the scope (center of image taken by the camera ) is pointed to, irrespective of the mount.

 

The Celestron StarSense explorer, also uses the phone's sensors to show you real time changes in position, between platesolves. The PSWAI on the other hand is not real time. It takes 10-20 seconds (depending on the platesolving computer) to show you the location, after you hit the button on Sky Safari. It does adjust the position based on whether the mount tracks or not. So while you can use it as a pushto, it is much more convenient to get close to the target using the PSWAI and star hop from there. This is especially good for light polluted skies where there are few naked eye visible stars. The camera can pick up a lot more.

 

The other advantages are because you do not have to mount the phone, you can use Sky Safari on a tablet you hold in your hand. And it is not limited to the brightest DSOs, I have never failed to get to any DSO, NGC or Messier that I have tried from Sky Safari. You do have to adjust the exposure time and ISO on some nights. With good alignment of the Camera and my 8" Dob, I can match stars in the FOV between the location Sky Safari shows with the PSWAI and my ES82 18mm EP in Bortle 7 skies.

Those cover a lot of the same reasons I was considering it in the first place, the settings specifying "equatorial" was what made me wonder if it would work at all with Alt-Az mounts - but clearly that's not an issue then. I live under bortle-8 skies where star-hopping is just not very viable, and even finder scopes are not great.

I currently use a manual azimuth setting circle and digital level - which lets me find a lot of things that are otherwise impossible but I could find so much more with a more accurate than "you are now roughly close to the target" solution.

 

I do see the shortcoming though - that the pure platesolving approach is just too slow for real-time tracking - though even so when you are CLOSE to a target and you are just not sure "am I slightly off-target or is it just too faint for this scope ?" the ability to press a button and confirm for sure (and if you are off, exactly which way to go to get back on) would still be awesome.



#16 hcf

hcf

    Viking 1

  • -----
  • topic starter
  • Posts: 878
  • Joined: 01 Dec 2017

Posted 21 February 2021 - 10:39 PM


I do see the shortcoming though - that the pure platesolving approach is just too slow for real-time tracking - though even so when you are CLOSE to a target and you are just not sure "am I slightly off-target or is it just too faint for this scope ?" the ability to press a button and confirm for sure (and if you are off, exactly which way to go to get back on) would still be awesome.

When you are making big moves the platesolving overhead is ok, but as you get closer and move by small amounts, the wait is a bit painful. I usually do platesolving to get the target in my RACI and then use starhopping. Once I think I have it centered, I confirm with a final platesolve. It also comes in handy if you get lost during the starhop part.

 

One of the overheads, is taking a picture on the Wifi cam, and downloading it on the computer over slow Wifi, can take upto 6 secs. If you use different hardware like a Pi4 with one of the new HD cams and a 35-50mm C mounted lens, you might save some of the time. The picture taking script will have to change to accommodate this, but it is not a big change.

 

A key element in accuracy is, aligning the scope and the camera accurately. The Orion 7033 Precision Slow-Motion Adapter is very useful for this.



#17 metalpoet

metalpoet

    Mariner 2

  • -----
  • Posts: 259
  • Joined: 20 Jul 2018
  • Loc: Cape Town, South Africa

Posted 22 February 2021 - 12:07 AM

When you are making big moves the platesolving overhead is ok, but as you get closer and move by small amounts, the wait is a bit painful. I usually do platesolving to get the target in my RACI and then use starhopping. Once I think I have it centered, I confirm with a final platesolve. It also comes in handy if you get lost during the starhop part.

 

One of the overheads, is taking a picture on the Wifi cam, and downloading it on the computer over slow Wifi, can take upto 6 secs. If you use different hardware like a Pi4 with one of the new HD cams and a 35-50mm C mounted lens, you might save some of the time. The picture taking script will have to change to accommodate this, but it is not a big change.

 

A key element in accuracy is, aligning the scope and the camera accurately. The Orion 7033 Precision Slow-Motion Adapter is very useful for this.

My approach would be to put a camera on the scope with USB connection to a laptop running both my planetarium and your platesolving scripts. Though considering I probably won't use the same camera as you I may rewrite the photo-capture script to use libgphoto2 instead.



#18 hcf

hcf

    Viking 1

  • -----
  • topic starter
  • Posts: 878
  • Joined: 01 Dec 2017

Posted 22 February 2021 - 01:29 AM

My approach would be to put a camera on the scope with USB connection to a laptop running both my planetarium and your platesolving scripts. Though considering I probably won't use the same camera as you I may rewrite the photo-capture script to use libgphoto2 instead.

That would work. The FOV of the camera/lens setup is also tied to platesolve time. Mine had 5x7 deg approximately. A laptop with USB is going to be faster. With a similar FOV using a laptop your platesolve+overhead should be below 10s maybe around 6s with the right flags.  A smaller FOV (longer focal length of the lens) would take longer.




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





Also tagged with one or more of these keywords: DIY, CMOS, equipment, astrophotography, planetarium software



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics