Hey, I would like to ask if it is possible somehow to make an RC car with a camera mounted on top. And with some extra stuff, that's the part where I get stuck;
The tricky part is that I need this project to do advanced hyperlapses (walklapses), so the car needs the possibilty to adjust the distance it travels in intervals (5 or 10 cm each time fx) and also I need to be able to adjust how long it stands still between each interval of movement.
I hope this makes sense, but basically i need to take pictures from different points on a trajectory, and the pictures can take from a second up to 30 sec, therefore the need for the adjustment of the "stand still time". 1 to 30 sec or more would be great.
So far it's not obvious that you need anything more than a conventional RC vehicle controlled which you control manually via an RC transmitter handset. If that isn't sufficient then perhaps you could clarify your requirements to explain what capability you need to add to that starting point.
What does it mean "longer distance"? Longer than what?
You say 1 to 30 s or more, which leads me to believe you want to shoot at night? Taking photos is easy. The problem I see is the light metering. Under 30 seconds is no problem, just put camera in aperture priority and you're set. Let the camera worry about it. That's what they are good at. I don't have my camera with me, but I believe that over 30 seconds you need to be in bulb mode which means Arduino would need to determine the exposure. That could get complicated.
The other and much more complicated problem is making the car autonomous. It would not only need to be capable of navigation and staying on the road, but also obstacle avoidance and decision making.
I would like to make a device that makes my camera travel in intervals from a to b in a straight line, stopping for x time at each interval.
Fx: I want to take 200 pictures along a straight line towards a building, the camera has to make 200 stops on a distance of 20m and if its by day the camera can stay for just a second on each stop, but at night, as you mention, I need more time on the stops
I want to make my camera move forward in a straight line, make stops. And I need to adjust the distance it travels each time and to adjust the time it stops on each interval. It's simple in my head I just need some guidance to guides or similar projects someone here may know of.
That seems to call for a trolley, a track for the trolley to run along, a motor to make it move (a stepper motor would make it easy to control the distance moved), and an LDR to measure the light level and decide how long to pause at each position. It doesn't seem to require any remote control, and doesn't actually need a steerable model car as such - just a chassis with four wheels.
How are you going to synchronise the picture taking with the movement and control the shutter speed?
I would not use light meter to measure the ambient light. I would attach a light sensor (LDR or something) to the LCD of the camera (my SLR is set to show the picture taken as soon as it is taken. Otherwise the screen is turned off. Can you get your camera to work that way? What camera do you plan on using anyway?) That's the only way I can think of to reliably determine if the exposure is completed so you can move on to taking the next picture.
Next is the problem of moving in a straight line. Any bumps in the road and even the road texture could affect the direction. Wheels would need to be soft and large, perhaps 10 cm in diameter.
As Peter said, a stepper would give you control over the distance traveled.
Syncing movement and picture taking is no problem if your camera has an external triggering capability (shutter release cable or IR). (What camera are we talking about?) Just use the same Arduino that drives the chassis to activate the camera as well. As for shutter speed I would leave that for the camera to work out and wait until I have the confirmation the picture is done being taken, as described above.
PeterH: You caught the idea pretty good, but I don't need any connection between the "vehicle" and the camera other than the screw that keeps them together, I will measure the light on my camera and then adjust the right timing on the device. If I can get the device done of course
Shpaget: I have a Canon 60D and I use the magic lantern capability of doing time lapse, so I set my camera to take a picture every X second, and then synchronize it with the movement of the small vehicle. A propos the bumps etc, I will of course check the track before using it and making sure there is no disturbance.
The device needs 4 wheels and 2 "dial buttons", one to adjust the distance traveled and the other to adjust the stand still time at each stop. I hope that I can be able to make some pretty good hyper lapses with this engine if I can find enough information on how to build it.
If you use magic lantern to trigger the shutter you lose the control over when the photos are taken relative to the movement of the car.
Why not make arduino both drive the car and take shots? That way you can be sure the shots are taken while the car is standing still.
Standstill time is dependent on exposure length + a bit more to allow for the vibrations to die down before the actual shot, so why would you set that standstill time before taking the shots? You would need to predict the light conditions for several hours into the future and any cloud could interfere, not to mention that a nightfall could mess it up completely.
The way I would do it is have the car move a certain distance and tell the camera to take a shot. Camera in the Av more will adjust the exposure automatically. Then you check if the process of exposure is done and make another step forward with the car only after the picture is taken.
I would not assume the exposure is done, especially if your timelapse encompasses both day and night (in daylight only shooting you could hardcode the delay of 5 seconds and be sure that a 1/200 or similar exposure will be done by the time the 5 second wait is done, but if the night comes and you get into 1 s or longer exposures, the error will add up. There will be accumulation of error even with daylight shooting, but that can be somewhat compensated for.
I would insist on feedback and information that exposure really is done before moving again. Earlier I mentioned the LDR on the LCD, but now I remembered that my 40D has a red LED that is lit during the exposure. That would be easier to monitor (more consistent than LCD).