15 replies to this topic

[MaximusStarHunter]

Hello, I have an 8" classic dobsonian. I am wondering how I can do longer exposures to see nebula and galaxies better. I was thinking there might be some way to do this with a non tracking dob. I have a ZWO camera for it, so I was thinking maybe there would be a way to combine different exposures together, but who knows I'm a bit new to this. I would appreciate your advice.

 

Thanks, Maximus.

[StargazerLuigi]

Look into SharpCap, I think it has 'Live Stacking' and I'd wager there's other software to do it too. This is how the Seestar and other smart telescopes work .. short (10s) exposures live stacked.

[bobzeq25]

A very few experienced imagers do this.  Most of them use big dobs and custom software.  They spend big bucks doing it.

 

Dobs are basically a visual scope, and not well suited to long exposure DSO work.  There is no magic solution, as can be seen by how few people do this, and what they spend doing it.

 

If you want to break into DSO imaging, this would be a FAR better way to go, for many reasons.  MANY people have chosen this path, and succeeded.  Trying to do it with a Dob?  Not so much.

 

 

Here's an image I took with that kind of setup, from a light polluted backyard.  Click on the lousy CN thumbnail for a good version, and details.

 

 

 

 

 

[Jethro7]

Hello, I have an 8" classic dobsonian. I am wondering how I can do longer exposures to see nebula and galaxies better. I was thinking there might be some way to do this with a non tracking dob. I have a ZWO camera for it, so I was thinking maybe there would be a way to combine different exposures together, but who knows I'm a bit new to this. I would appreciate your advice.

 

Thanks, Maximus.

Hello Maximus,

I agree with  Bobzeg25. I have seen other members here on CN that have aftermarket or home made tracking platforms for their Dobs but you will still be limited to short subs of 10 to 20 seconds do to spherical rotation and tracking accuracy of the tracking platform. I think these tracking platforms are good for high magnification viewing for the most part and not for imaging.  Many Imagers still use these large reflectors for astrophotography. Personally I think the best way to approach this is, you will need to turn the Dob into a Newt with parallax rings and mount it on a beefy EQ mount that has tracking. However this will get real expensive really quickly. This is where Bobzeg25 thoughts are much more practical. I would recommend keeping the Dob for visual and purchase a set up like Bob mentioned or maybe a ZWO Seestar S50 or the upcoming new ZWO Seestar S30. Dwarf Labs Dwarf II smart telescopes as an inexpensive entry level way into astrophotography. These gadgets are easy to operate and whole lot of fun for a fraction of the price of a conventional Imaging rig. 

 

HAPPY SKIES AND KEEP LOOKING UP Jethro

 

A few images from my Seestar S50 

 

 

 

 

Edited by Jethro7, 04 December 2024 - 10:40 PM.

[Andros246]

Hello, I have an 8" classic dobsonian. I am wondering how I can do longer exposures to see nebula and galaxies better. I was thinking there might be some way to do this with a non tracking dob. I have a ZWO camera for it, so I was thinking maybe there would be a way to combine different exposures together, but who knows I'm a bit new to this. I would appreciate your advice.

Thanks, Maximus.

If you’re going to use a dob you absolutely need a goto dob for this.

Edited by Andros246, 04 December 2024 - 07:42 PM.

[TOMDEY]

Static telescope moving image >>> There is actually a well-established technique to achieve fairly long exposures using a static/grounded undriven telescope. I used an FLI camera comprising the Kodak KAF 16803 CCD chip (9µm pitch 4096x4096 array). The technique is called "Time Delayed Integration" (TDI) or colloquially "Push Broom / Drift-Scan Imaging". The trick is to align the chip readout to the direction of the moving object space field and adjust the readout clock to exactly match the speed that the image flows over the array. Shutter open, photons continuously falling on the chip during this hours-long exposure, with the total integration time for each and every image element equal to the lateral image flow rate multiplied by the array length in the flow direction. Example: My 36-inch aperture F/3.75 36-inch Newtonian telescope with TV Big Paracorr, aimed near the celestial equator, preferably near the meridian --- scope grounded / shutter open all night. The total integration time for each and every image element (in my example) is therefore (almost exactly) 128 seconds. What you collect in a typical clear 10-hour night is therefore an image that subtends 0.53^o in declination and 10h in RA, exposure time (for each image element) 128s. In actual practice there is a ramp-up at the beginning and end. But ---  "you get my drift".

 

This is extremely efficient for survey work, sucking in around ten hours of data every favorable night. You then aim the scope a half-degree higher or lower, park it there, and are ready for the next good night.

 

Drift-Scan is a well established technique for film or solid state imagery of the ground --- think airplane flying over the ground in a straight-line pass, with the film or electrons dutifully keeping exact pace with the pilot. (These pilots are highly trained specialists.)

 

I ran the camera through its paces in lab simulations under ideally controlled conditions... and easily achieved small sub-pixel smear over the entire 4096 x 4096 array with images feet long (at the array).

 

Technical considerations:

 

> near celestial equator

> large file size

> array clocking tolerance

> camera must have uniformly adjustable readout rate

 

Similarities / Relateds:

> slit-scan

> panoramic

> Google Earth Ground vehicle rotating street cam

> USGS Survey Maps

> air planes, helicopters, drones, satellites

> visiting earth, moon, planets, asteroids

> search and rescue --- land and sea

 

[I worked photogrammetry in my travels. Lockwood Support serviced our region here. "Air-Breathers" were the primary platform for many many decades.]

[bobzeq25]

Static telescope moving image >>> There is actually a well-established technique to achieve fairly long exposures using a static/grounded undriven telescope. I used an FLI camera comprising the Kodak KAF 16803 CCD chip (9µm pitch 4096x4096 array). The technique is called "Time Delayed Integration" (TDI) or colloquially "Push Broom / Drift-Scan Imaging". The trick is to align the chip readout to the direction of the moving object space field and adjust the readout clock to exactly match the speed that the image flows over the array. Shutter open, photons continuously falling on the chip during this hours-long exposure, with the total integration time for each and every image element equal to the lateral image flow rate multiplied by the array length in the flow direction. Example: My 36-inch aperture F/3.75 36-inch Newtonian telescope with TV Big Paracorr, aimed near the celestial equator, preferably near the meridian --- scope grounded / shutter open all night. The total integration time for each and every image element (in my example) is therefore (almost exactly) 128 seconds. What you collect in a typical clear 10-hour night is therefore an image that subtends 0.53^o in declination and 10h in RA, exposure time (for each image element) 128s. In actual practice there is a ramp-up at the beginning and end. But ---  "you get my drift".
 
This is extremely efficient for survey work, sucking in around ten hours of data every favorable night. You then aim the scope a half-degree higher or lower, park it there, and are ready for the next good night.
 
Drift-Scan is a well established technique for film or solid state imagery of the ground --- think airplane flying over the ground in a straight-line pass, with the film or electrons dutifully keeping exact pace with the pilot. (These pilots are highly trained specialists.)
 
I ran the camera through its paces in lab simulations under ideally controlled conditions... and easily achieved small sub-pixel smear over the entire 4096 x 4096 array with images feet long (at the array).
 
Technical considerations:
 
> near celestial equator
> large file size
> array clocking tolerance
> camera must have uniformly adjustable readout rate
 
Similarities / Relateds:
> slit-scan
> panoramic
> Google Earth Ground vehicle rotating street cam
> USGS Survey Maps
> air planes, helicopters, drones, satellites
> visiting earth, moon, planets, asteroids
> search and rescue --- land and sea
 
[I worked photogrammetry in my travels. Lockwood Support serviced our region here. "Air-Breathers" were the primary platform for many many decades.]

Did I mention experienced imagers using custom expensive methods that have NO relevance AT ALL to the original poster? <smile>

[TOMDEY]

Did I mention experienced imagers using custom expensive methods that have NO relevance AT ALL to the original poster? <smile>

Rumor has it that WalMart sells Drift-Scan Christmas Cameras that mate with Christmas Dobsonian Telescopes.      Tom

[PIEJr]

Rumor has it that WalMart sells Drift-Scan Christmas Cameras that mate with Christmas Dobsonian Telescopes.      Tom

Rumor has it...

 

Well, you opened the door Tom.... 

 

I did crawl around Walmart, but no luck.

So I went web crawling and it brought me back here.

 

https://www.cloudyni...ed-eq-platform/

Edited by PIEJr, 04 December 2024 - 10:24 PM.

[Andros246]

Well that’s something

[MaximusStarHunter]

Static telescope moving image >>> There is actually a well-established technique to achieve fairly long exposures using a static/grounded undriven telescope. I used an FLI camera comprising the Kodak KAF 16803 CCD chip (9µm pitch 4096x4096 array). The technique is called "Time Delayed Integration" (TDI) or colloquially "Push Broom / Drift-Scan Imaging". The trick is to align the chip readout to the direction of the moving object space field and adjust the readout clock to exactly match the speed that the image flows over the array. Shutter open, photons continuously falling on the chip during this hours-long exposure, with the total integration time for each and every image element equal to the lateral image flow rate multiplied by the array length in the flow direction. Example: My 36-inch aperture F/3.75 36-inch Newtonian telescope with TV Big Paracorr, aimed near the celestial equator, preferably near the meridian --- scope grounded / shutter open all night. The total integration time for each and every image element (in my example) is therefore (almost exactly) 128 seconds. What you collect in a typical clear 10-hour night is therefore an image that subtends 0.53^o in declination and 10h in RA, exposure time (for each image element) 128s. In actual practice there is a ramp-up at the beginning and end. But ---  "you get my drift".

 

This is extremely efficient for survey work, sucking in around ten hours of data every favorable night. You then aim the scope a half-degree higher or lower, park it there, and are ready for the next good night.

 

Drift-Scan is a well established technique for film or solid state imagery of the ground --- think airplane flying over the ground in a straight-line pass, with the film or electrons dutifully keeping exact pace with the pilot. (These pilots are highly trained specialists.)

 

I ran the camera through its paces in lab simulations under ideally controlled conditions... and easily achieved small sub-pixel smear over the entire 4096 x 4096 array with images feet long (at the array).

 

Technical considerations:

 

> near celestial equator

> large file size

> array clocking tolerance

> camera must have uniformly adjustable readout rate

 

Similarities / Relateds:

> slit-scan

> panoramic

> Google Earth Ground vehicle rotating street cam

> USGS Survey Maps

> air planes, helicopters, drones, satellites

> visiting earth, moon, planets, asteroids

> search and rescue --- land and sea

 

[I worked photogrammetry in my travels. Lockwood Support serviced our region here. "Air-Breathers" were the primary platform for many many decades.]

How would one do this in Sharpcap or something?

[TOMDEY]

How would one do this in Sharpcap or something?

My FLI camera came with the software drop down to control the data readout clocking rate. Simple and straight forward. It was up to me to figure out the camera θ_z rotational clocking angle and derive the nominally-needed temporal clocking rate for my focal length, system distortion, and target declination. You set up that approximately, run a few minutes exposure on a star field (mechanically grounded static telescope of course!) evaluate the image to determine any cross and/or in-track smear (should be small) and then tweak θ_z and clocking rate to achieve perfection. That one iteration is enough in the hands of an experienced technician. So, within a few minutes, you're ready for the night's run. Go sleep and the data is ready in the morning. NOTE: Active ~guide correction~ is entirely eliminated, not needed... because the earth's axis direction and rotational rate are ideally constant. If you change only declination for the following night's run, just reset the clocking rate by the (obvious) Δcos(Dec) factor and are ready to roll. So much of the baloney associated with guiding entirely circumvented. 

 

You can derive the (profoundly simple) set-up and tweaking math yourself. The entire concept is nice and easy, not rocket science. For our simulations I of course also (got a friend to do) the math to drive the scene generator to run the camera through its paces. I did the scene generator optics... he (they) did the generator multi-axis mechanicals and drives. His task was tougher than mine! Thanks Jerry, Mike, and Bill! Far as I'm concerned, this technique is a goldmine of recovered simplicity and reliable image quality. Sure, a bit arcane... but fantastic for survey work (supernovae, variable stars, things that move relative to the stars, etc. etc.)    Tom

[MaximusStarHunter]

My FLI camera came with the software drop down to control the data readout clocking rate. Simple and straight forward. It was up to me to figure out the camera θ_z rotational clocking angle and derive the nominally-needed temporal clocking rate for my focal length, system distortion, and target declination. You set up that approximately, run a few minutes exposure on a star field (mechanically grounded static telescope of course!) evaluate the image to determine any cross and/or in-track smear (should be small) and then tweak θ_z and clocking rate to achieve perfection. That one iteration is enough in the hands of an experienced technician. So, within a few minutes, you're ready for the night's run. Go sleep and the data is ready in the morning. NOTE: Active ~guide correction~ is entirely eliminated, not needed... because the earth's axis direction and rotational rate are ideally constant. If you change only declination for the following night's run, just reset the clocking rate by the (obvious) Δcos(Dec) factor and are ready to roll. So much of the baloney associated with guiding entirely circumvented. 

 

You can derive the (profoundly simple) set-up and tweaking math yourself. The entire concept is nice and easy, not rocket science. For our simulations I of course also (got a friend to do) the math to drive the scene generator to run the camera through its paces. I did the scene generator optics... he (they) did the generator multi-axis mechanicals and drives. His task was tougher than mine! Thanks Jerry, Mike, and Bill! Far as I'm concerned, this technique is a goldmine of recovered simplicity and reliable image quality. Sure, a bit arcane... but fantastic for survey work (supernovae, variable stars, things that move relative to the stars, etc. etc.)    Tom

I still don’t really get the full concept, how does a non tracking scope track all night?

[Andros246]

I still don’t really get the full concept, how does a non tracking scope track all night?

if it makes you feel better i've been doing this for a few years and I have not a clue what hes talking about

 

 

The "software" or whatever hes talking about came from a $10,000-$128,000 camera

Edited by Andros246, 06 December 2024 - 05:21 PM.

[TOMDEY]

I still don’t really get the full concept, how does a non tracking scope track all night?

Think of it as a type of relativity - In a nutshell --- the photoelectrons follow/pace the imaging photons point to point >moving object→•←moving image< in perfectly matched pace across the face of the camera array and therefore are stored in the downloaded image unsmeared. The telescope is static, the array is static (relative to the earth) --- only correlated photons and photoelectrons are moving. That's the ~Executive Summary~ We are relying on the rotating earth to do the moving for us, and that is extraordinarily/ideally uniform, simple, and well understood motion. Pretty close to magic and so simple that the concept cruises right over most otherwise smart people's heads. Simplifying Gestalt... while 99.9% of other imaging fans are neck deep in the weeds. Feynman, Noether, Albert, Kingslake, Emch --- it's all about coordinate transforms simplifying rather than complifying our theories and executions.

 

But I must admit --- when I got my simulations (sims) to run in the lab... even though I completely understand the theory... I was still amazed that my static camera indeed captured the giant high-res image of the moving target perfectly. It's the sort of thing --- once you experience it in action... you want it!

 

[I was also immersed in photogrammetry and flight simulators on and off for years/decades, so the challenges of imaging ~on the fly~ was familiar turf. My daughter ran sims to certify/qualify our pilots and crews. A family thing.]     Tom

Edited by TOMDEY, 06 December 2024 - 09:15 PM.

[TOMDEY]

Another way to think of it: Spend a lot of $$$ on an ideally capable tracking mount that makes your telescope follow the stars ~or~ Spend a lot of $$$ on an ideally capable camera that makes your photoelectrons follow the stars. It comes out about the same. And on gargantuan scopes... the scanning alternative is actually a lot cheaper. Some similarity to William Herschel's and Ross's survey scans along the meridian.    Tom

Attached Thumbnails

• 
• 

Edited by TOMDEY, 06 December 2024 - 09:38 PM.