Re:Home made Astro Sky Cam

Here is my home made Astro Sky Cam I built today.

I wonder why I posted it on a Jimny page?

Sent from my moto g42 using Tapatalk

Ingenious use of Jimny parts if ever there was!

Cool.
It's going to leak when the bearings go 

Hi Martin,

I have need running an All Sky Camera for several years now, it uploads every 5 minutes to www.drrobin.co.uk .

Note no Jimnys were harmed during the making of this camera.

Robin

DrRobin

Website looks good.

My camera is here: -  www.lines-associates.com/allsky/

It is off/on as I experiment with it and get the different functions to work and also will move around (and show the inside of my study if it is switched on while i play with it)

Your camera seems to be working fine just now, which software are you using?.

I am on my second all sky camera, the first was a mono but not good in low light. I moved to a colour imaging source camera I had spare and rebuilt the housing. Software is AstroVideo running a custom script to go to 65 second exposure and will capture the main stars on a clear night.

 

The box in front is a cloud/rain sensor and the dome white dome behind is my smallish observatory complete with computer controlled 10” RC imaging scope.

Robin

Fascinating.
Can you explain in very simple terms how you obtain those photos?
I see problems with rotation of earth / long exposure etc with a fixed camera (lack of tracking) unless the software is the key.
Genuinely interested but, absolutely no idea of how you do this.
Electronically ok ish, even know what a rasberry pi is, but after that welcome any explanation you wish to offer to enlighten me further.



 

I am using a raspberry pi running the thomasjacquin allsky software.

However testing shows it leaks ( left in the rain this evening) so I will have to look at using some sealant. The Suzuki seal leaks (as normal!).

Camera moved inside and pointing North as the Aurora forecast has gone Red.

[Scmikie] As this is an allsky cam it is not for "detail" shots so tracking etc is not necessary for this. My main rig is below which does track the rotation of the earth for detailed long exposures. The raspberry pi is a small computer to drive the system

Sent from my moto g42 using Tapatalk

Scimike, as Martin says the 'All Sky' camera is fixed, they use a wide angle lens (a fish eye in my case to get [nearly] 180 degree coverage) so even over a minute or so you can't see the 'star trails'.

My main telescope is much longer focal length and that does need to track for anything over a couple of seconds.

The tracking is performed by the mount, mine is an 'Equatorial' mount, the same basic design as Martin's, these have one axis aligned to the axis of the earth and the other as 'elevation' over the equator.

In theory an equatorial mount only needs to move in one axis to track and it has stepper motors that move the mount at the same speed as the rotation of the earth. In practice the stepper motor is never precise, there gears are never perfect, the balance of the scope is never perfect, the wind and vibration in the ground will move the scope slightly and for it to work the mount itself has to be perfectly aligned to the earth's axis, almost impossible to do with a tripod where you move it in and out of the house every session.

A tracking mount, if somewhere close will allow exposures up to a minute or two, depending on the focal length of the scope.

To over come the limitations of the tracking, you then add a second scope and camera that you focus on a bright star. This acts as a guide scope and the software constantly corrects the motors on the mount to keep the guide star in the right place.

Once you run an equatorial mount, that is fairly close to good polar alignment and is guided then it can track and compensate for long exposures. My system is fairly reliable out to 10 minutes, but it should be capable of 30 - 60 minute exposures, although I find that cloud or a plane or a neighbour switching a light out runs the exposure if I try for longer. Anyway even at 10 minutes, I only get 6 exposures an hour and not enough in one evening for a picture, most of the ones on my website were taken over several nights.

Once you have captured your long exposures then you use a piece of software that analyses the images, aligns them together then adds them, this increases the intensity of the stars and reduces noise, it's called 'stacking'. Then it's a case of stretching the brightness and contrast of the image to bring out the detail, usually in Photoshop.

I have an added complication, I use a mono camera (a purpose built 35mm CCD that is electrically cooled to about -30C to reduce noise), so to build a colour image I take a number of images using a clear filter (this gives a luminance plane) and then take a number of images using a red, then green, then blue filters to give a colour plane. All of these are stacked and processed separately then combined in Photoshop to give the final image.

The luminance plane contains all of the detail, think of it as a child's colouring book, there is a black and white image printed on the page, the child then adds the colour on top of the detail, it's really just the same, except the luminance and colour planes in these photos are grey scale rather than black/white or colour/no colour.

Finally the pictures of the [colourful] gas clouds are done with narrow band glass filters, the filter is set to the wavelength of the light that the gas produces, this is where you need the mono camera and some of the narrow band filters let so little light through you need the long exposures.

The colours in these are false, but represent the nearest colour in the visible spectrum of the light. If you looked with the eye, you might see some of these gas clouds but you don't see the detail or the colour as our eyes are just not sensitive enough and the image we see is dominated by light coming from the stars and other wavelengths.

Robin

My mount is the same basic design as Martin's, except rather than it been a classic equatorial mount where the scope hangs off one side with a large counterweight, it is a signle fork design.  This allows the scope to sit close to the main axis and reduces the counterweight size to almost nothing.  Not only does this provide for higher performance (less mass to move), it also makes it very compact and that was what I needed to fit the 10" scope in a small fibreglass dome.  Unfortunately there is only one manufacture of this type of mount and I think they machine it out of a solid block of gold, they cost a fortune.

 

 

The mount (in red at the bottom) is supported by a home made pier, about 160Kg of concrete, SS reinforcement rods and a home made aluminium pier top to hold the mount.  It's very stable and isolated from the rest of the dome by earth and rubber gaskets.

The scope is a 10" carbon fibre truss tube design, they cool down very quickly and are also very stable.  The scope is optimised for photography, it gives a large flat field with very little distortion even across a 35mm camera.

The guide scope and guide camera are the white tube and blue box under the main scope at the top right.

Robin

Martin, did your camera capture the aurora last night?

Mine has a problem, if it tries to ftp up to the website and it fails, it hangs until I manually clear the error. Unfortunately, it did it last night at 21:10, so was off all last night.

I am going to write a bit of code to automatically clear the error, it's happening too frequently.

Robin

I believe I did for just a few frames between the clouds and lights. I do not have a good view north but I left the camera in a north facing window overnight just to see.

Here are some looped frames and I feel you can just see the sky flash pink at one point which would represent a few minutes as it is a time lapse of course. So this is in the middle of light polluted Thames Valley