283 replies to this topic

[thoughtwave]

I wanted to share an experiment I did back in April during galaxy season. I found that you can purchase very fast 8mm F/0.9 CCTV camera lenses, and with a little 3d printing these can be mounted to a sensitive guide camera such as the ASI290mm. Here's what the camera with mounted lens looks like:

 

My idea was to mount this entire device to a 40mm eyepiece to use with my 10" F/5 Dobsonian. Together, the telescope, eyepiece, and CCTV lens would effectively create a 10" F/1.0 system, which is four times as fast as a RASA or Hyperstar. In theory, this should allow for very short subs, and my hope was that I could get useable images even without tracking by keeping the exposure time below 1s. Here's what the eyepiece-lens-camera assembly looks like:

 

 

To test my idea, I took my telescope and "digital eyepiece" assembly to a nearby Bortle 3 observing site and tried observing a few galaxies. By stacking 0.5s exposures, I was able to get a surprising amount of detail in as little as 10s total exposure. Here are some of the results:

 

M51, 20x0.5s (top) and 74x0.5s (bottom)

 

 

 

 

M81 and M82, 50 x 0.5s (top) and 89 x 0.5s (bottom)

 

 

 

 

Leo Triplet (41 x 0.5s)

 

M101 (61 x 0.5s)

 

Overall, I was pretty happy with the results as well as the experience of using this setup. With the extremely fast image formation and manual telescope movement, I found it to be a lot like visual observing, except that the level of detail I could observe was greatly enhanced. Of course, you could get better images with a tracked setup and prime-focus camera, but for this experiment I was trying to create something like digital night vision. This could be useful for primarily visual observers who want to experiment with EAA, but don't want to invest in a whole new rig with tracking, etc. I can also see something like being turned into an integrated product to replace the Revolution Imager. Anyway, perhaps a fairly niche application of EAA, but feel free to discuss.

[MunichAtNight]

Hello!
 

... My idea was to mount this entire device to a 40mm eyepiece to use with my 10" F/5 Dobsonian. Together, the telescope, eyepiece, and CCTV lens would effectively create a 10" F/1.0 system, which is four times as fast as a RASA or Hyperstar. In theory, this should allow for very short subs, and my hope was that I could get useable images even without tracking by keeping the exposure time below 1s. Here's what the eyepiece-lens-camera assembly looks like:
 
To test my idea, I took my telescope and "digital eyepiece" assembly to a nearby Bortle 3 observing site and tried observing a few galaxies. By stacking 0.5s exposures, I was able to get a surprising amount of detail in as little as 10s total exposure. Here are some of the results: ...

Well done! I tried something similar, but couldn't arrive to put it correct together to achieve focus! Can you please explain more in detail. How did you arrive to get the right and necessary distances? For example the 3d-printed adapter between the camera and the lens? The length of this adapter did you find by calaculation? How? Or simply by trying? How did you fix it to the Televue Plösl? It is a TV Plösl, isn't it?

 

Very interesting! Congratulation for your expertise!

 

Servus - MunichAtNight - Ewald

[alphatripleplus]

Interesting. I platesolved the 2nd image for M51, and your FOV is 75.1' x 42.2' with a pixel scale of 2.33 arcsec/pixel. That corresponds to an effective focal length of about 257mm, in line with your description of a f/1 system on a 10 inch scope. 

Edited by alphatripleplus, 24 September 2023 - 08:01 AM.

[BrentKnight]

Nice.  Something like an NV device but costing less than $500.  It would be nice if an external screen could be eliminated.

[dob45]

Very interesting!  Bravo!

[dcweaver]

Looks like a fun way to build a DIY 0.2x reducer. If you had a coma corrector laying around you could put that in the focuser first (before the eyepiece/camera) and see of it cleans up the field any,

[bigbangbaby]

Very cool. Some thinking along similar lines. https://www.cloudyni...hdmi-out/page-2

[thoughtwave]

Hello!
 

Well done! I tried something similar, but couldn't arrive to put it correct together to achieve focus! Can you please explain more in detail. How did you arrive to get the right and necessary distances? For example the 3d-printed adapter between the camera and the lens? The length of this adapter did you find by calaculation? How? Or simply by trying? How did you fix it to the Televue Plösl? It is a TV Plösl, isn't it?

 

Very interesting! Congratulation for your expertise!

 

Servus - MunichAtNight - Ewald

Thanks for the question! There was some trial and error involved to design the 3d-printed adapters. Here's what the camera-to-lens adapter looks like in cross-section. The adapter screws into the CS-mount adapter provided with the ASI290MM. The lens itself uses its existing M16 focusing thread to screw into the adapter. I added some PTFE tape to make this smoother, since there are no actual internal threads on the adapter. This allows me to adjust the focus of the lens simply by rotating it. Fortunately, this particular lens has enough back focus that it comes to focus without hitting the AR-coated window on the camera. 

 

Here are the parts disassembled. There are also three set screw holes at the top of the lens adapter to adjust the lens tilt, but I didn't end up using them. However, based on the asymmetrical aberrations in the images I took, it might have helped to dial in the lens tilt after all. 

 

 

The eyepiece adapter is fairly simple. Here's a cross section which I think is self-explanatory. The spacing between eyepiece and lens can be adjusted by moving the camera assembly up and down to get the eyepiece exit pupil to align with the camera entrance pupil. 

 

 

 

Here is a link to the STL files for the adapters if anyone is interested: https://drive.google...NPX?usp=sharing

 

Unfortunately, I'm not sure the exact lens I'm using is available anymore. At least, I couldn't find it again when searching. I did find a similar 8mm F0.9 lens on eBay, but it seems to have a different design and housing, so the adapter parts would probably need to be modified to accommodate it. 

[alphatripleplus]

Thanks for the details on the adapters. Based on your results, I may end up playing with an afocal set-up using a 26mm Plossl and one of those very short focal length CCTV lenses. 

Edited by alphatripleplus, 24 September 2023 - 07:20 PM.

[descott12]

Very good book idea and i am glad it worked so well. I am going to have to try it myself.

[UB_Astro]

I wanted to share an experiment I did back in April during galaxy season. I found that you can purchase very fast 8mm F/0.9 CCTV camera lenses, and with a little 3d printing these can be mounted to a sensitive guide camera such as the ASI290mm. Here's what the camera with mounted lens looks like:

 

My idea was to mount this entire device to a 40mm eyepiece to use with my 10" F/5 Dobsonian. Together, the telescope, eyepiece, and CCTV lens would effectively create a 10" F/1.0 system, which is four times as fast as a RASA or Hyperstar. In theory, this should allow for very short subs, and my hope was that I could get useable images even without tracking by keeping the exposure time below 1s. Here's what the eyepiece-lens-camera assembly looks like:

 

 

To test my idea, I took my telescope and "digital eyepiece" assembly to a nearby Bortle 3 observing site and tried observing a few galaxies. By stacking 0.5s exposures, I was able to get a surprising amount of detail in as little as 10s total exposure. Here are some of the results:

 

M51, 20x0.5s (top) and 74x0.5s (bottom)

 

 

 

 

M81 and M82, 50 x 0.5s (top) and 89 x 0.5s (bottom)

 

 

 

 

Leo Triplet (41 x 0.5s)

 

M101 (61 x 0.5s)

 

Overall, I was pretty happy with the results as well as the experience of using this setup. With the extremely fast image formation and manual telescope movement, I found it to be a lot like visual observing, except that the level of detail I could observe was greatly enhanced. Of course, you could get better images with a tracked setup and prime-focus camera, but for this experiment I was trying to create something like digital night vision. This could be useful for primarily visual observers who want to experiment with EAA, but don't want to invest in a whole new rig with tracking, etc. I can also see something like being turned into an integrated product to replace the Revolution Imager. Anyway, perhaps a fairly niche application of EAA, but feel free to discuss.

Wow!  That can open a completely new path for me to study.  I assume you do live stacking like normal EAA?
 

[tbhausen]

Great work!

[alphatripleplus]

One question I have concerns the aperture of the CCTV lens, i.e. what  was the actual diameter of the CCTV lens used? In order to utilize the full aperture of the scope optics, the CCTV lens aperture should be at least as large as the exit pupil generated by the telescope/eyepiece optics. So, a 10 inch f/5 scope and 40mm focal length eyepiece yield a magnification of 31.75 ( = 254mm x 5 / 40mm), and an exit pupil of 8mm (= 254mm / 31.75). If the CCTV lens is smaller than 8mm in aperture, some of the light from the exit pupil won't be captured, and the system will be operating at a smaller effective aperture, and a larger f/ratio than assumed. ( It's the same calculation visual astronomers do when matching the human pupil size with the exit pupil of a telescope/eyepiece or binoculars.)

 

If the CCTV lens diameter is actually 8mm, and the 8mm exit pupil from the eyepiece exactly lines up, there should be no wasted light. 

Edited by alphatripleplus, 25 September 2023 - 10:08 AM.

[Ubuntu]

Very nice setup !!!

F1 is very quick !!

With my 10" F5 manual dob, a camera and an SVbony focal reducer I only managed to go as low as F1.8

[thoughtwave]

One question I have concerns the aperture of the CCTV lens, i.e. what  was the actual diameter of the CCTV lens used? In order to utilize the full aperture of the scope optics, the CCTV lens aperture should be at least as large as the exit pupil generated by the telescope/eyepiece optics. So, a 10 inch f/5 scope and 40mm focal length eyepiece yield a magnification of 31.75 ( = 254mm x 5 / 40mm), and an exit pupil of 8mm (= 254mm / 31.75). If the CCTV lens is smaller than 8mm in aperture, some of the light from the exit pupil won't be captured, and the system will be operating at a smaller effective aperture, and a larger f/ratio than assumed. ( It's the same calculation visual astronomers do when matching the human pupil size with the exit pupil of a telescope/eyepiece or binoculars.)

 

If the CCTV lens diameter is actually 8mm, and the 8mm exit pupil from the eyepiece exactly lines up, there should be no wasted light. 

According to the specs it is 8mm F0.9, so the aperture should be 8.89mm. But it's true specs may not be accurate. Do you know a good way to measure the actual aperture?

[alphatripleplus]

According to the specs it is 8mm F0.9, so the aperture should be 8.89mm. But it's true specs may not be accurate. Do you know a good way to measure the actual aperture?

I think that you could physically take a ruler and measure the diameter of the front of the CCTV lens - assuming it is not recessed. That would at least give you an upper bound on the physical aperture. Maybe post a picture of the top of the CCTV lens with a ruler to give us an idea? 

[mgCatskills]

Wow!

 

Thanks for the question! There was some trial and error involved to design the 3d-printed adapters. Here's what the camera-to-lens adapter looks like in cross-section. The adapter screws into the CS-mount adapter provided with the ASI290MM. The lens itself uses its existing M16 focusing thread to screw into the adapter. I added some PTFE tape to make this smoother, since there are no actual internal threads on the adapter. This allows me to adjust the focus of the lens simply by rotating it. Fortunately, this particular lens has enough back focus that it comes to focus without hitting the AR-coated window on the camera. 

 

Here are the parts disassembled. There are also three set screw holes at the top of the lens adapter to adjust the lens tilt, but I didn't end up using them. However, based on the asymmetrical aberrations in the images I took, it might have helped to dial in the lens tilt after all. 

 

 

The eyepiece adapter is fairly simple. Here's a cross section which I think is self-explanatory. The spacing between eyepiece and lens can be adjusted by moving the camera assembly up and down to get the eyepiece exit pupil to align with the camera entrance pupil. 

 

 

 

Here is a link to the STL files for the adapters if anyone is interested: https://drive.google...NPX?usp=sharing

 

Unfortunately, I'm not sure the exact lens I'm using is available anymore. At least, I couldn't find it again when searching. I did find a similar 8mm F0.9 lens on eBay, but it seems to have a different design and housing, so the adapter parts would probably need to be modified to accommodate it. 

Wow!  I do have a 290mm guiding camera, so would be intrigued to play with this.  

 

Before I read this, I just went looking for an f0.9 8mm lens and came up empty... the closest was a listing on AliExpress, but is no longer available.  Plenty of 8mm F2 lenses, but I guess too slow.  I did find some f1.0 4mm or 6mm lenses on AliExpress for about $10 (+4 shipping).

 

If someone cracks the code on a source of lenses and creates a 3D printing file that works with it, I'd be very interested.  Attempting to do that myself is way outside of my comfort zone. 

 

However, perhaps we could get together a small group of us to place a combined order for lenses and 3D printing services.   We might get a better price than we could do individually, but the bigger payoff is just having one person dealing with the hassle of ordering from AliExpress or eBay and acquiring the adapters... 

 

I'd volunteer to take on the administration (ordering, assembling and shipping out the individual packages, collecting payments) if someone with the technical expertise to make sure this all works can advise me.

Edited by mgCatskills, 25 September 2023 - 12:16 PM.

[thoughtwave]

Hi all,

 

Not sure what the rules are regarding posting links to products, but here's a link to an 8mm F0.9 lens that is very similar to the one I used. I am pretty sure it's the same manufacturer, but they seem to have changed the housing from metal to plastic. I believe this is also sold on Aliexpress, potentially for a lower price.

 

https://www.ebay.com/itm/385476241445

 

Other lenses and focal lengths could potentially work as long as they are fast, F1.0 and below. The formula for effective F# of the system is:

 

(Telescope F#) * (CCTV Lens focal length) / (Eyepiece focal length)

 

which works as long as the system F# is higher (slower) or equal to the CCTV lens F#. This assumes the lens aperture stop and eyepiece exit pupil are perfectly aligned, which is why I included some margin by using an F0.9 lens to form an F1.0 system.

 

So for an F/5 telescope, other working combinations for an F1 system could be a 30mm eyepiece and 6mm lens, or 25mm eyepiece and 5mm lens. I chose the 8mm because it claims to support the largest camera format (1/1.8") and because it's a good match for the relatively inexpensive (but high quality) plossl. 

 

Glad this topic has gained some interest! I'll try to reply to your individual questions when I get home. Definitely motivated to continue experimenting now...

[Varrium]

This topic is super cool and inspiring. I'd love to try and replicate some of that success with the gear I have:

•     Telescope: 8-inch manual Dobsonian with f/5.9, 203 mm diameter, and 1200 mm focal length.
•     Eyepiece #1: 2" 32mm Q70 from Orion. (it being a 2" makes it unusable for this particular case right?)
•     Eyepiece #2: 25mm plossl that came with the dob.
•     Camera: Got an SVbony SV305 Pro on the way.

I mean, I'm not sure how different it'd be using my camera compared to the ASI 290mm, especially since they both have the same sensor. Still trying to wrap my head around all the techy details here.

If I go with an 8mm M16 like thoughtwave did:
(5.9*8)/32 = 1.475

It's not close to the f/1.0 system, but I'm wondering if I can still get some results with what I've got. Anyone got any tips? Can I play around with this without buying a ton of new stuff?

 

As a side note, I have limited experience with 3d printing but I have access to printers so I'm willing to give it a go for whatever is needed like the mentioned adapter.

Edited by Varrium, 01 October 2023 - 11:55 PM.

[thoughtwave]

Hi all,

 

I've been pretty busy lately, so apologies for the lack of replies. I did want to share some updates and experiments I've done over the past few weeks.

 

First of all, I discovered several other lenses that I bought as part of my original project. These are shown below.

Two of them, including the 8mm lens used in the OP, have metal barrels and appear to have been discontinued. The other two, a 6mm F0.95 and 8mm F0.9 lens, have plastic housings and are still available for purchase through eBay or AliExpress. 

 

 

I took alphatripleplus's suggestion and measured the actual apertures of each of these lenses. Unfortunately the aperture stops seem to lie somewhere in the middle of each lens, rather than at the first surface, making accurate measurement more difficult. I ended up putting each lens on an LED tracing panel with a set of calipers for scale and taking a top-down picture from about 1m away to minimize the effect of distance on the image scale. Here's what this looked like:

 

I used paint.net to measure the aperture diameters in pixels, and compared them to the known caliper width (set to 30.00mm) and lens barrel diameters in pixels to get measurements in mm. As expected, the two references gave me slightly different image scales because the top of each lens is slightly closer to the camera than the calipers, but I got the following ranges for aperture diameters and F#s:

 

8mm "F/0.9" lens (metal barrel):  Aperture: 7.29 - 7.42mm (F/1.08 - F/1.10)

8mm "F/0.9" lens (plastic housing):       Aperture: 7.67 - 7.80mm (F/1.04 - F/1.06)

6mm "F/0.9" lens (metal barrel):       Aperture: 6.05 - 6.11mm (F/0.98 - F/0.99)

6mm "F/0.95" lens (plastic housing:  Aperture: 5.91 - 5.94mm (F/1.01 - F/1.02)

 

As you can see, each lens falls short of its advertised aperture by 10-20%, which is a bit disappointing, but important to know. This means I was only using about 9.1" of my 10" telescope aperture in my original post. 

 

Of the two lenses that are still available for purchase, only the 8mm lens was able to reach focus with my ASI290mm, so I decided to make that the focus of future development. I did have to sand down the barrel by a fraction of a mm to get enough backfocus, but this was easy enough.

 

 

I modified the lens adapter to accommodate the larger housing and printed a new version.

 

 

Initial tests suggest that this lens has pretty decent image quality. There is some coma in the corners, but stars in the center are sharp and round. With an actual F# of around 1.05, it would pair well with a 38mm eyepiece on my F/5 telescope, so I've purchased a 2" 38mm "SWA" eyepiece, as well as a second-hand MPCC to try out coma correction. I hope to test out this new setup in the coming weeks.

 

Finally, here are some links to the parts:

8mm F/1.05 lens (mine was advertised as F/0.9, but this one looks identical and matches the measured F/#): https://www.ebay.com/itm/314154769411

6mm "F/0.95" lens (doesn't reach focus with my 290mm, but could work if you removed the protective window, or with board cameras such as Arducam modules, etc): https://www.ebay.com/itm/254457260320

3D-printed adapter for above 8mm lens to CS mount: https://drive.google...?usp=drive_link

 

[thoughtwave]

This topic is super cool and inspiring. I'd love to try and replicate some of that success with the gear I have:

•     Telescope: 8-inch manual Dobsonian with f/5.9, 203 mm diameter, and 1200 mm focal length.
•     Eyepiece #1: 2" 32mm Q70 from Orion. (it being a 2" makes it unusable for this particular case right?)
•     Eyepiece #2: 25mm plossl that came with the dob.
•     Camera: Got an SVbony SV305 Pro on the way.

I mean, I'm not sure how different it'd be using my camera compared to the ASI 290mm, especially since they both have the same sensor. Still trying to wrap my head around all the techy details here.

If I go with an 8mm M16 like thoughtwave did:
(5.9*8)/32 = 1.475

It's not close to the f/1.0 system, but I'm wondering if I can still get some results with what I've got. Anyone got any tips? Can I play around with this without buying a ton of new stuff?

 

As a side note, I have limited experience with 3d printing but I have access to printers so I'm willing to give it a go for whatever is needed like the mentioned adapter.

Hi Varrium,

 

Nothing wrong with using a 2" eyepiece, you would just need a different 3d-printed adapter to mount the camera to the eyepiece. You could try a 6mm lens with your Q70, which would give you an f/1.1 system, but you would need to find one with enough back focus for your SV305. I couldn't find the specs for this from SVBony, but I actually have an original (non-pro) SV305, which appears to need at least 6.5mm. It might be hard to find a 6mm lens that will work, but I confirmed that I can reach focus with my (slightly modified) 8mm lens. You could try an 8mm lens with something like a 42mm GSO Superview eyepiece, which also gives around f/1.1 and is not too expensive used. I actually have one of these eyepieces that I no longer use and would be willing to part with. DM me if you are interested.

 

Also, for what it's worth, in my experience my SV305 is significantly less sensitive than the ASI290mm or similar color cameras such as my ASI224MC, but perhaps the pro version is better. If you do experiment with this I'd be interested to hear about your results!

[jml79]

Hmm, a very interesting experiment. I wonder how inexpensive it can be done, perhaps with the super budget SV-105c. I have one and a 3d printer so may consider trying it. I also wonder if the eyepiece and lens can be fiddled with to accommodate various focal lengths and smaller, cheaper scopes. My largest is a 102mm refractor, so pretty wide at f/1 but usable for some DSO's at f/2-f/4.

[alphatripleplus]

Thanks for that info on the clear aperture of these CCTV lenses. I'm encouraged by the observation that the 8mm (plastic housing) lens gives decent image quality. 

[alphatripleplus]

I just ordered the 8mm (plastic housing) lens to experiment with.

[MarMax]

It's a bit like delayed night vision that's not that horrible green color and looks to be higher quality. I'll probably not experiment with this but it sounds fun.

 

Really cool thoughtwave!

Edited by MarMax, 10 October 2023 - 08:20 PM.