DSLR's, cheap EQ's and wide-field drift aligning

graphworlok (James) · #1 30-04-2011, 12:44 AM

So, about 15 months or so ago i was given a "toy" plastic 50mm scope as a present... a few days later i was trying to image through it with a webcam.
Just over a year ago i managed a bargain on ebay - 8" newt, EQ3 with motor, and a pile of assorted bits - mainly, a clock motor. After assorted images of planets, etc and the moon, we bought an EOS40D, which is way too heavy to be used with my scope/mount combo, seems to work quite nicely on the EQ mount itself..

I have mainly been imaging sections of the milky way (30 second exposures while pointed towards it, tweak position until the cloudy stripe is "centered")

Now i am after longer exposures and getting more picky about what i try to image, as well as running into other assorted issues (dew, too much balance, bright moons, etc)

My question - Drift aligning with a wide-field type DSLR lense (18-55mm "kit" lense@55mm"

My current technique i have been trying over the past few weeks has been this:
Align tripod with compass, level, etc.
Point camera "straight ahead" (so should be pointing directly at true southern pole,)
(shamelessly modified from many other websites)
Using laptop+gphoto2 to control camera, take a 480 second exposure@ISO100. No tracking. Tweak the east/west and lat. knobs as needed until non-rotating stars are centered. Repeat images until stars appear centered.
Then take repeat with same eposure/ISO settings, but with motor engaged, and controller/timer in hand
Track for 30 seconds(or so)
2x forwards for 220 seconds
Reverse for 220 seconds
If my alignment is off, i will see the stars appear as V's - the 30 seconds of tracking creates a bright spot for the "start" point, so i can tell which way they are moving
Stars above/below the south pole will either drift up or down compared to the bright "start" point
Stars to the east/west will drift left or right
I then log the image name, examine which direction and how much the drift is to the "up" and "right" (whether stars are drifting up/down and left/right)
Tweak the latitude or east/west alignment knobs. Log which way and how much (turns of lat CW or CCW, same for east/west adjustment)
Image again
Compare changes in star drift
Repeat
Eventually i get to the point where the stars are solid lines (as far as i can tell) in both axis rather than V's, and as i progress, the turns-of-the-knobs go from being 2 turns to 1, 1/2 or 1/4 turns until i find myself zooming on on a 10MP image@200x to try and see which way the lines end up

So, i have to ask

Am i completely insane? Does this make sense to anybody else? I have gotten to the point where i can then track for a full 7 minute exposure and the stars still appear as points, but once i point my camrea "up" at the milky way, i still get oval stars. I am sure part of it is slop in my cheap mount , as some is the increased distance of the stars from the pole exaggerating any error, however further alignment becomes problematic due to the camera being at a very odd angle, making it hard to determine how to correct

Any suggestions?

dannat (Daniel) · #2 30-04-2011, 05:57 PM

James how long is this process taking you? If you goto the imaging sub forum there is an eos utility, it has red lines so you can align your mount -then start tracking & see which way it moves..

Or you can drift align with the camera like a telescope- just remember the camera gives a right side up image..pick a star low in E, adjust s necessary then pick a star above & do the same..

Your method should get you there I just wonder how long it takes you

tlgerdes (Trevor) · #3 30-04-2011, 06:42 PM

1) Your method appears resonably sound, convoluted, but sound.

2) Why do you want 7 min exposure of the galaxy for widefield. You would be better off at 800ISO and 2mins.

Oval stars could also be lens aborations.

mswhin63 (Malcolm) · #4 01-05-2011, 11:22 AM

Hi James, I would go as low as 800 ISO and I mostly use 1600 and take many more shots. to reduce the noise levels in the final stack.

I have the same type of mount although believe mine is a cheap EQ2 not a 3.

Last night I captured 2 minutes x 19 exposure at 1600 ISO and night before 90 second at ISO 800 and worked well. Unfortunately I cannot post the link to my site as I upgraded the skin and it is not working. I will fix it today and hope show you the result.

I still experience slight drift even after 90 seconds though. They are not the best of mounts for AP.

graphworlok (James) · #5 01-05-2011, 02:46 PM

Quote:

Originally Posted by tlgerdes

1) Your method appears resonably sound, convoluted, but sound.

2) Why do you want 7 min exposure of the galaxy for widefield. You would be better off at 800ISO and 2mins.

Oval stars could also be lens aborations.

The 7 minute exposure is more just for alignment/tracking purposes. Once i am satisfied with the stars overlapping in forwards and reverse, i will usually image for a much lower period of time (30 to 240 seconds, ISO 1600 to 400)

The long exposures are purely to get the mount aligned as accurately as possible - Any errors are amplified once i point the scope away from the pole - or it may just be slop in the mount throwing mount out of alignment slightly once the weight shifts.

If course, i'm looking at stars in a 10MP image at 100% - if the image is scaled down, the oval stars become much less noticeable.

Theres a selection of images i have taken here (Yes, slow site) http://220.233.71.133/Astro/

graphworlok (James) · #6 01-05-2011, 02:51 PM

Quote:

Originally Posted by mswhin63

Hi James, I would go as low as 800 ISO and I mostly use 1600 and take many more shots. to reduce the noise levels in the final stack.

I have the same type of mount although believe mine is a cheap EQ2 not a 3.

Last night I captured 2 minutes x 19 exposure at 1600 ISO and night before 90 second at ISO 800 and worked well. Unfortunately I cannot post the link to my site as I upgraded the skin and it is not working. I will fix it today and hope show you the result.

I still experience slight drift even after 90 seconds though. They are not the best of mounts for AP.

The long exposures are purely for the alignment - before i leave the camera to image proper, i will usually take a set of images ranging from 30 seconds up to 300 seconds, with varying ISO's, preview them to see whats the longest exposure i am happy with, then leave the camera clicking away while i do other things - i will periodically check to see if a building, etc has drifted into view, (or clouds, or greyed-out image from dew). Also, i will usually recharge my laptop and camera batteries bfore imaging proper. And i'm also starting to bring the whole camera inside to warm up again (damn dew!)

graphworlok (James) · #7 01-05-2011, 03:01 PM

Quote:

Originally Posted by dannat

James how long is this process taking you? If you goto the imaging sub forum there is an eos utility, it has red lines so you can align your mount -then start tracking & see which way it moves..

Or you can drift align with the camera like a telescope- just remember the camera gives a right side up image..pick a star low in E, adjust s necessary then pick a star above & do the same..

Your method should get you there I just wonder how long it takes you

I have actually just started leaving my tripod outside now - and my viewing location does not give me much to see to the east or west. the EOS utility looks handy, but alas my laptop for imaging runs linux - i have simple shell scripts that i preset to adjust my exposure time and ISO, and the laptop saves the images directly to a network share, so i can view the raw .CR2 files on a desktop with a 1920x1200 screen as soon as they are saved.

I'm getting tempted to rig up a small cradle with servos to press the appropriate buttons when required to change the clock motor speed/direction, so i will only need to be outside for the time taken to adjust the mount, instead of waiting for the timer i am using currently to tell me.

Another issue i ran into recently was having my mount and camera balanced too well - without some tention on the gears, instead of getting straight trails, when i switched from forwards to reverse, with no weight to take up the slop the lines ended up becoming wiggles instead of small arcs

I guess the main issue i am seeing is once i point the camera up at something worth imaging, any errors are amplified, or more errors are introduced by the mount, and due to the camera being at an odd fiagonal angle, its tricky to make any corrections

tlgerdes (Trevor) · #8 01-05-2011, 08:43 PM

You might be seeing some sort of periodic error that is see-sawing the stars back and forth due to the worm gear not being smooth.

joecool (Mark) · #9 05-05-2011, 11:59 AM

You are seeing exaggerated errors I believe when you move the camera away from the SCP. As a test I mounted my 450D using the 18-55 lens at 20mm unguided on a tripod with 30sec exposures the other night. At the SCP the stars are pretty round, but on Orion setting in the west the stars are 1cm long lines!

This page I have found to be an excellent description
http://www.astrophotoinsight.com/pub...d-drift-method
but swap south for north.

Mark.

graphworlok (James) · #10 05-05-2011, 09:09 PM

Hmm, i believe this is likely the problem - I guess i will try pointing the camera straight up for some extra alignment.. I just bought an old Pentax SUper Takumar 135mm prime lense, which should help me reduce the time needed for noticable drift too.. Of course, looks cloudy for the foreseeable future!

Quote:

Originally Posted by joecool

You are seeing exaggerated errors I believe when you move the camera away from the SCP. As a test I mounted my 450D using the 18-55 lens at 20mm unguided on a tripod with 30sec exposures the other night. At the SCP the stars are pretty round, but on Orion setting in the west the stars are 1cm long lines!

This page I have found to be an excellent description
http://www.astrophotoinsight.com/pub...d-drift-method
but swap south for north.

Mark.

Moon (James) · #11 06-05-2011, 11:17 AM

James,
You are doing a great job pushing your gear to the max and learning a lot on the way. It is harder to track in some parts of the sky, so you can either stay away from there or reduce the exposure time or try to work on the polar alignment as you suggested.

Also try taking some shorter exposures and see how round you can get the stars in the center of the image by tweaking the alignment of all the components.

James

graphworlok (James) · #12 01-06-2011, 12:31 AM

I think i'm starting to hit assorted mechanical issues - I am moving house in a few weeks, so will strip down the scope and see what i can clean up after that

I did have an idea of a different method of polar aligning my scope, purely using a DSLR which i tried tonight - My lense started fogging up though, so i have left the mount outside o continue tomorrow if it remains clear down here

This method has much shorter steps, but is frustrating in different ways. I'm not sure if its a workable solution either. Basically:

Position mount with counterweight pointed straight down, and camera straight ahead. Take a long ( > 60 second) exposure) - Motor not engaged. Look at the position of the polar stars, compared to image resolution. Adjust until they are in the center of the image - so the Y position should be half of the image height. (stellarium, etc helps for this)

Repeat until stars are centered.

Tilt head 90 degrees so that the counterweight is horizontal (i use a spirit level) - Take another series of images, adjusting the heads east/west alignment until the stars are centered.

To my mind, this should give a pretty good approximation of polar alignment, without the 8+ minute iterations my previous method used. I spent a few hours, and generated over 330 meg of images trying to get this method sorted out - Calling it a night due to lense fog and work in the morning! - i was reachine the point where i was continually overshooting my intended position due to sloppy mechanical issues (cheap old EQ3 mount)