PDA

View Full Version : How to pan a camera 360 degrees without rotating it



Evan
04-24-2008, 10:59 AM
I just finished building a mount for the Microsoft LifeCam that I have been experimenting with for use as an all sky camera. I already made a mount that panned it in the conventional manner but found that the requirement for limits and limit switches was too, ummm, limiting. Since the camera is equipped with a very special wide angle lens I put together it has a 100 degree field of view with essentially no distortion. It doesn't need to be panned in the conventional manner to cover the entire visible sky in my location. All it needs is to be tilted to about 30 degrees from the vertical in any direction to see the very high horizon I have. I decided to build a swash plate drive since it will allow infinite rotations in either direction with no cords winding up or other similar problems.

The protective cover is made from Lexan and has a fully multicoated and overcoated optical window in the top to protect the lens. That window is quartz overcoated so it can be wiped with just about anything and is nearly impossible to scratch. I had to piece it together from two pieces so there is a fine line across the center of the field of view, no big deal at all.

http://www3.telus.net/metalshopborealis/skycam1.jpg

The lens is mounted in a proper 65mm lens cell that I salvaged from a photocopier. It is special because it uses the highest quality available negative optics in a configuration that produces a very wide angle view with almost no fisheye distortion. I adapted the general principle of the Stanhope Magnifier lens to a pair of negative lenses. The two lenses are top of the line fully multicoated and overcoated high index eyeglass blanks. I happen to know the optician at a local eyeglass fitter and he sold them to me at, ahem, less than usual price. I can't say how much as I promised not to and there are people in this town that lurk on this board just to see what I am up to. I'll just say that even at a very low price they weren't cheap.

http://www3.telus.net/metalshopborealis/skycam2.jpg

The mount is driven by a small gear motor that powers a worm and wheel. The worm I turned from a piece of steel shaft and the worm wheel was hobbed from some 1/2" Lexan. Nearly all the parts are either Lexan or ABS plumbing components that have been modified to suit the application. This makes the unit very weather resistant and the camera itself is buried deep inside. The circular track is a drain track for any rain that tries to run in because of the tilt.

The lower swash drive is free to rotate under the upper unit which is captured by the pin that rides in the slotted strip on the side. The pin prevents the camera table from turning so the only choice it has is to follow the changing tilt angle of the swash drive.

http://www3.telus.net/metalshopborealis/skycam3.jpg


Here is an animation of it in action. It normally takes about a minute to make a full 360 degrees (although I could speed it up) so I took a short time lapse sequence.

http://www3.telus.net/metalshopborealis/camswash.gif

retusaf99
04-24-2008, 08:33 PM
Evan,
I missed this post before! That's way cool!

I'm trying to imagine what the pictures look like.

Doug

Evan
04-24-2008, 08:38 PM
Give me a few minutes and I will post a short video.

mwechtal
04-24-2008, 08:50 PM
Evan,
From the title I was visualizing a vertical lens with a 45 deg. rotating mirror in front of it. Very nice.
Mike

Evan
04-24-2008, 09:00 PM
Here it is. This video was taken at 1 frame per 15 seconds at sunset. The flashes are when the software adjusts the exposure to compensate for the diminishing light. The overlay is only accurate when the camera points south which is another reason I built the new swash drive. With the new drive the camera is always oriented the same way and I will be able to use it to determine where in the sky something happened. I am hoping to figure out a way to capture meteor trails without having it capturing all the time. It depends on how quickly it can take a shot when the motion detector sees a change.

By the end of this clip is is almost fully dark even though it doesn't look it in this video. Removing the infrared filter increased the night sensitivity a great deal.

Under one meg:

http://www3.telus.net/metalshopborealis/skycam3.wmv

retusaf99
04-25-2008, 01:16 AM
Thank you, Sir.

Very impressive.

Doug

Evan
04-25-2008, 03:14 AM
That particular web cam is by far the best I have seen for low light. I have been messing with web cams for astronomy for about 15 years now. Here is a picture taken with that camera. This isn't the sun, it's the moon and it was completely night when it was taken last winter.

http://www3.telus.net/metalshopborealis/skycam4.jpg

Ken_Shea
04-25-2008, 03:26 AM
Evan,
I find this very interesting, however after reading you post (several times) I still am unable to comprehend the "Panning" as the video appears to be simple time laps :confused:

Evan
04-25-2008, 04:05 AM
The video was taken with the previous mount, not this one. It wasn't smooth enough to make "filming" practical during panning. I have also improved the lens mounting to eliminate the circular vignetting almost completely. The plan is to have the camera slowly scan the sky when aurora are expected. At other times it will just be pointed south away from the sky glow of Williams lake. During the day it will be pointed north to keep the sun from shining on the sensor.

The motor only draws about 40 ma at 6 volts. I have been thinking of placing a sun sensor on the unit to automatically tilt it away from the sun.

Forrest Addy
04-25-2008, 06:02 AM
Isn't that called a "nutating" motion?

Evan
04-25-2008, 08:39 AM
I suppose. The nutation I am familiar with is a movement of a planet (Earth for example) which is a cyclical nodding motion that is added to the usual precession of the poles. In the case of Earth it's caused by the fact that the Moon orbits at an angle to the plane of the Earth's plane of rotation so it tugs more first on one pole and then the other, because they are alternately closer.

Lew Hartswick
04-25-2008, 09:25 AM
Isn't that called a "nutating" motion?
Yep. One of the tracking radars back in the 50s (maybe the SCR 584)
used a nutating feed to the antenna to produce a conical scan.
Been way too many years to remember all the numbers. :-)
...lew...

Dawai
04-25-2008, 09:40 AM
AND THE OTHER METHOD? viewing into the bottom of a spherical mirror, taking the software to split pie sections out.

I have this plastic mirror purchased to film the descent into a local tourist pathway for some small disabled kids.

No, not got it off the ground. anyone want to help some? I'll end up using the laptop in a backpack. THE lil movie camera is not high enough resolution.. AND MULTIPLE CAMERAS with the images stitched.

A child in a wheelchair, who can never climb a rock path could explore with his computer.

oldtiffie
04-25-2008, 09:43 AM
Isn't that called a "nutating" motion?


Yep. One of the tracking radars back in the 50s (maybe the SCR 584)
used a nutating feed to the antenna to produce a conical scan.
Been way too many years to remember all the numbers. :-)
...lew...

That's so Lew.

One that we in OZ had in the 50's and 60's was on a weapons system where the wide slow cone was used for "search/find" and the narrow quick cone was used for track/"Engage".

One of the predictors had a "swash" similar to Evan's where it pointed to/at the future target position as a point in free space but was resolved into weapon bearing and elevation.

Paul Alciatore
04-26-2008, 01:24 AM
Neat.

So are you searching for comets or little green men?

BTW, they do still make fish eye lenses, don't they?

dp
04-26-2008, 02:22 AM
The Astronomy Picture of the Day site has an excellent animation of lunar nutation that is quite surprising. The moon seems unchanging from one observation to the next, but when filmed and presented as an annual series it is quite fascinating. I can't find the link now, but it is worth looking up.

As for your images - for a set of images over a specific time it is certainly possible to stabilize them visually though it would be a lot of work. I think too that lens distortions would quickly reveal themselves. Stabilization could be augmented with a geo-fixed light source so that all the image series components could be aligned on that light point. Said light point could then be removed from each image

In the animation you provided earlier there was a nearly 90 rotation shift - what caused that?

Evan
04-26-2008, 06:08 AM
I was panning the lens around to catch different parts of the sky.

The lens has very little distortion because of the arrangement I used. It's a pair of negative diopter lenses mounted back to back. One lens is a very strong -5.5 diopter and the other is a -2. The -2 is placed in front convex side out and the -5.5 is behind with convex side to the back which is backward. The front lens acts as a corrector plate for the back lens and most of the distortion is canceled.

Here is a shot looking through the lens. You can see both how wide the field of view is and the lack of fisheye. If I had paid full retail for those lenses it would have cost me at least $250. I didn't pay that though.

http://vts.bc.ca/pics3/wideang.jpg

dp
04-26-2008, 01:29 PM
So here is nutation evident during a complete lunation:
http://apod.nasa.gov/apod/ap010218.html

Probably a tongue twister in that.

Evan
04-26-2008, 01:55 PM
That isn't nutation, that's libration. It isn't a motion of the Moon's axis or the Earth's that cause that. It's only an apparent motion, not a real one. It's caused by the varying distance of the Moon from the Earth during an orbit. The Moon has an elliptical orbit. Since the rotation of the Moon about it's axis doesn't change period but the time it takes to travel a particular angular distance in orbit does due to the eccentricity, the leading and trailing edges alternately become more or less exposed as do the north and south poles because the orbit isn't coplanar with the Earths axis of rotation. Whew.