Motion Time Lapse. Engineers? - DPChallenge Forums

Canon RF 24-70mm f/2.8 L IS

10/04/2007 09:11:15 PM · #1

OK, I have an idea for a project involving moving a still camera along a track while tethered, taking time lapse images. Iâ��m running into a roadblock with my limited mechanical/electrical knowledge (nearly none). With the wealth of knowledge on DPC, I figured that someone here might be able to help.

Iâ��ve determined that I need the camera platform to move about .2 in./min. or 1 ft./hr. This would give me a ten-foot travel over the course of a ten-hour time lapse. So far the only solution I know I could build would be a pulley/counterweight system, with the counterweight being a water container with a controlled drip (Incredible Machine?). I have two problems with this design:

1. It would be extremely difficult to time and to repeat consistently.

2. Would the camera platform accelerate towards the end of the cycle? This would be undesirable.

I imagine an electrical motor system would be more flexible and more reliable. But I havenâ��t found one â��ready madeâ�� that moves this slowly. I suppose I could build or modify something, but I have no idea where to begin with such a design. Any ideas or suggestions?

kirbic

Canon EOS R5

10/04/2007 09:23:15 PM · #2

How often are you planning on shooting, an how far off the ground does the camera need to be?

Nikon AF Nikkor 50mm f/1.8D

10/04/2007 09:29:54 PM · #3

I would plan on shooting in anywhere from 10 to 60 second intervals. Height is not yet determined, but I imagine that it would optimally be flexible. It would probably be no higher than 5 ft., but it could be as low as ground level. I would like to use this system, once I get it together for multiple shots and (with some modification) multiple perspectives.

_eug

Nikon D50

10/04/2007 09:45:14 PM · #4

Are we talking vertical movement or horizontal?

It seems you want a motor with a known rotation attached to a drum that the "tether" winds around.

Nikon AF Nikkor 50mm f/1.8D

10/04/2007 09:54:18 PM · #5

Thanks, I had considered both for different applications, but horizontal would be the most useful(perhaps even on a curved track). I know nothing about motors, so I would need a lot of help in that area. Would you know of a good primer for the complete lay person online?

Also, one of my main concerns would be very steady and constant movement. I would like the end result to look like a smooth track shot done on film or video (only in time lapse).

ETA: I can also imagine applications for a circular track for TL's of unlimited duration. But I suppose that would be another project entirely.

Message edited by author 2007-10-04 21:58:45.

_eug

Nikon D50

10/04/2007 10:09:30 PM · #6

You might consider the motors used for RC cars.

The only resource that comes to mind is Make Magazine.

ralph

Nikon Z9

10/04/2007 10:10:16 PM · #7

//www.epanorama.net/circuits/diskstepper.html

is a good place for a simple stepper motor controller that is possible to program to do what you wish ..
cheap & simple you just need a track & small enough spindle /string diameter

10/04/2007 10:11:50 PM · #8

Thanks, I'll check out those links!

Canon RF 24-70mm f/2.8 L IS

10/04/2007 10:15:40 PM · #9

Woah! I only just started to look at it, but makezine.com is definately going in my favs!

muz64

Canon EOS-1D X

Canon EF 50mm f1.2L

10/04/2007 10:16:55 PM · #10

Interesting project. I'm an electronics engineer, so I can help you with the motor side of things. You need a really highly geared DC motor - a BBQ rotisserie motor springs to mind, but I dare say that this would move too fast for you. 1'/hr is quite slow. If you'll have the shutter open while its moving, you'll need to make sure your track positively locates the camera, and that there is no backlash in the drivetrain - so you get motion in only one plane.

I'd be looking in the robot supply section at your local electronics/hobbie shop - they have some really neat gear for doing things like this - including DC motors of all sizes and high reduction gearboxes, couplings, etc. How you go from there depends on your budget and your fabrication skills.

One idea could be just buying a rolling chassis with rubber wheels (from an RC car for example) that you could easily mount your camera on. Connect up a small battery driven DC motor through a high reduction gearbox. Setup a flat surface to run on, and you can get your curved path by setting the steering off centre. This'll only give you horizontal motion. There's probably 100 different ways to do it. Let me know how you go.

kirbic

Canon EOS R5

10/04/2007 10:29:46 PM · #11

On a straight track there are lots of simple options. A curved track makes for a much harder challenge. Let's consider a straight track for a moment:
- A track can be nothing more complicated than two lengths of pipe
- The camera carrier can be just a board with rollers that runs on the pipes
- To provide motion, find a small motor with a gear reduction (low-speed output shaft). Look at industrial supply houses like Grainger or McMaster-Carr. There should be small units that still run on 120V AC
- To provide motion, either mount a rubber "drive wheel" to the output shaft and let it engage the pipe(s) or let the shaft wind up a string.
- Another option is not to provide motion at all, but just to move the camera platform along the pipes by a certain amount for each exposure.
Let's take a look at a little simple math... you want a ten-hour total timeframe, and exposures from 10s to 60s apart. That means:
- There are 36000 seconds in 10 hours, so at 10s/photo, you'd be taking 3600 photos! Better have a plan for camera power and card space! One photo per 60s means 600 photos, a much more manageable, though still daunting project.
- At 1 foot (12 inches) per hour of movement, and a 10s interval, your camera moves 0.033 inches between shots. At a 60s interval, make that 0.200".
- The above figures give you an idea of what kind of motion you need to supply, and whether manually moving the camera is feasible (it's probably not).
Let's talk about the motion. Assuming 0.2 in/min, and assuming that the "winding the string" method is adopted, you still need a *very* slow speed shaft. if your shaft is 1/4 inch in diameter, then it has a circumference of 1/4*pi, or about 0.785 inches. If you want a movement rate of 0.2 in/min, your shaft will need to rotate only once per 3.92 minutes (about 0.254 RPM). You could round to one rotation per 4 minutes. It's probably possible to find a unit that rotates as slowly as once per minute; slower than that may be a problem. Unless you're OK with building your own gear reduction (not that difficult, but requires decent mechanical skills and perhaps access to some special tools) then you may be challenged to achieve this slow speed.

Nikon AF Nikkor 50mm f/1.8D

10/04/2007 10:45:54 PM · #12

Wow! Lots of great ideas, thanks! This will probably keep my head spinning for a while.

Kirbic, one question about the string idea ... as the string winds, the effective diameter of the output shaft would increase, increasing the rate of movement. Is this correct? Is there a solution to this?

Also, I have no problem with long TLs. Ten hours is "short" for me! I regularly shoot 48 -72 hour TLs(many thousands of frames)! Ever seen CS3 with 3GB RAM/AMD X2 give up on a batch process?

Muz, I'll visit the local hobby shop tomorrow. It's been a long time since I've been there (used to build model rockets/airplanes)!

_eug

Nikon D50

10/04/2007 11:31:08 PM · #13

Originally posted by nickp37:

... as the string winds, the effective diameter of the output shaft would increase, increasing the rate of movement. Is this correct? Is there a solution to this?

Correct.

Wider spool width for the winding.
Thinner string. You could go with 10 lb test fishing line.

Konica Minolta MF 85mm f/1.7

10/04/2007 11:42:41 PM · #14

Of course! That just makes too much sense.

Still, for flexability, I think it would be nice to have the motor/drive assembly and the camera platform in one unit. That way, curved and circular tracks of any length would simply be a matter of designing a new track.

briantammy

Olympus PEN E-P3

10/05/2007 12:04:13 AM · #15

I have a large clock in my office. I'm not sure how much strength
the hour hand has and if it would be able to pull something around
but it is pretty reliable and repeatable. Using a clock that run
from house current might not drag or slow down. The hour hand would
make the full circle in 12 hours and on my clock of about 1 ft radius
the distance traveled would be about 3-4 inches/hour. Increase
the distance from center until you get 1 ft/hour traveled. Not sure
if this gives you any ideas.