remote focus

I'd love to have something like this. Consider that the influx of electronic gimbals is going to create an demand for devices like this.

I'm a veteran programmer and would be happy to help with the software around a system like this. Please PM me for my email.
 
Good point. If gimbals develop into a solid platform, the next step would to get away from just using wide angle lenses. Imagine being able to use a longer lens and start playing with DOF while on the gimbal.

I'll take it another step. Instead of having a second operator running the remote pan and tilt, it would be nice to have some way to track an object in the frame and automatically follow it while the gimbal operator moves.

So get coding! :)
 
Hey. well i'm sure there is a lot of code written already for this, but i also wanted to learn the coding , not necessarily to be able to say "i'm the man", but i'd love to know coding. Oh, and i was thinking about the auto focus, but i don't know if you can make that happen without some heavy processing and coding, most likely the data should be processed in real time via laptop. The google cars use this technology for the driverless cars, but for those they use really powerfull computing. What i want to say, is that probably it will be less likely for people like us to pull it off, with resources like a spare room and internet connection. I just had an idea: let's say you make a brushless gimbal, 3 axis type. And you want to use your camera like you would just hold it in your hand. I was thinking that for panning and tilting you can mount an rc gyro, the small ones, on let's say google glasses. You link the glasses to your camera, for live video, low res is ok, the gyro controls as i said the pan and tilt, and some controls on the gimbal handles are enough for pan and zoom. and the hardest part of this is making the google glasses talk to your camera, But it can work without them, the important part is the head mounted gyro.
 
Oddly enough, I submitted exactly that idea (the concept of google glass and mounting a 3 axis motion sensor) to a companies' internal start-up committee where I worked about 5 years ago. They rejected it. Since they rejected the idea, I asked if I could have it. They said yes, but the lawyers refused to release the concept to me.
 
F#c@ the lawyers man. :)) you cannot have a patent on an idea. At least that's how romanian patent sistem works. You can patent only a technical solution for something. Whatever i make, i want to patent it under some sort of gpl, so that people can enjoy making without threats of lawsuits.
 
grosariustefan said:
Ggrantly, please keep the thread constructive. thanks.
Click to expand...

What? No one thus far has mentioned servocity as a vendor for servo motors and controllers. How is mentioning a good vendor not constructive grosariustefan? They have many parts that are entirely relevant to this thread OK?

Grant
 
They have a lot of product, including some really good soft start programmable controls as well as hardware bits and pieces to make stuff work. Too many products for a single link. They can also be pretty helpful with hardware selection if you talk to the technicians. Since ServoCity hadn't been mentioned thus far, it seemed worth mentioning for anyone who wasn't familiar with them. Someone here might not have the knowledge about the available products which you obviously have. That is why it is a moderated open forum. You are not the moderator, only the OP, so your gripe at me about "keeping things constructive", isn't your call, nor constructive.

Grant
 
If you want to call me a troll that's fine, I don't much care. But before you do, you might want to check through my post history for both quality and attempting to help others as they have helped me. For sure I get cranky when someone calls me out for a thread contribution which doesn't fit your particular constructive posting criteria.

Peace
 
I just got around to test my microservos which i wanted to first use for the zoom, and i have to admit, they deserve a special place in hell. These things are so choppy that you would have to do black magic in order to get a decent flow on the zoom. http://hobbyking.com/hobbyking/stor...MG_Servo_13_5g_1_5kg_0_08s_EU_Warehouse_.html it's these guys, and, if you want to do some diy, i strongly recomend not to buy these, or similar small servos. i'll take wild guess and say that most of them are terrible. I think this week i'll be able to order the micromotors, encoders and arduino and hopefully that setup will work smoothly. If anyone has a better solution, please do write it down here. cheers, and productive tinkering.
 
Well, if you read post 3, you could have saved yourself some time :). I know, I know, you have to try this stuff yourself to see what it is all about. I do the same thing. Now you know.

If you want to do this right you need to use small DC motors that are driven by a variable DC voltage; not PWM. You can try the Pololu motors with the encoders and drive them with the PWM. But I've been down that road (without the encoders) and I found the results disappointing.

Save yourself some time and learn from the professionals who already figured this out. Go buy a cheap (under $100) Fujinon or Canon 2/3" ENG lens with the 'servo' on it. There's lots of them on ebay. I bought a lot of 4 for $200 to take apart. Open it up and see what's inside. See how they drive the motor to get that ultra-smooth drive. They can make those zooms creep so slow you can't hardly notice it. Yet they can also zoom very fast. Study it and copy it. The problem is that I haven't found anyone selling the electronics necessary to duplicate it. I'm not interested in making my own circuits, so I stopped there.
 
well. i searched on the voltage regulators, and they seem to be very inefficient. probably that's why they're not really in use in the rc world. The formula : (Vin-Vout) x Iout =P is the amount of power it consumes. for these small motors it's next to nothing, but when you move on to a bit larger motors, 100, 200 W of power turned into heat for no reason, is not a good thing. PWM controllers are much more efficient, since they just switch on and off really fast, and do not rezist the electrical flow. Thanks for the input, but i think the pwm solution is better(no waste, does not require a large heatsink). With the encoders i am also hoping to make it run really slow, it should be just a try and err thing. Oh, yeah, I know you told me, but i had already ordered the servos :)) . The good thing is that they are extremely cheap, so i can use them on other , less precise projects.
 
Yes, efficiency is exactly why PWM is used. DC drive stinks for efficiency, but it delivers the performance you are looking for. Batteries aren't a big deal now. So waste the power and get the performance you need.

But try the PWM and learn about that. I will be wanting to copy it if you get it to behave well.
 
Looks good. So you are having the turnwheel encoders behave as absolute position wheels? I.e 'n' turns or some percentage of a turn of the focus wheel equals end-to-end movement of the focus (same for zoom)?

Having the arduino let's you make the turnwheels behave any way you like. Nice.
Curent sense - nice

Do you have enough inputs to the arduino nano to add more buttons/pots? Once you have this working, you may want to be able to set programmed moves of set points.

Example - for focus, it would be nice to be able to program focus stop points. Zoom would also be nice to have something similar. More sophisticated ENG lenses have the zoom set points to be able to snap-zoom to a focal length.
 
so, again in order :
encoders&stuff, spot on. I found high resolution optical encoders, 500ppt, and 300 ppt, i'll probably get the 300 one, it's a bit cheaper.
the current sensing is there so the receiver arduino can figure out where the end points of a lens are. I found non invasive hall effect sensors to measure the current. it will have a set current value at the motor start, and when it starts searching for the limits, current goes up as motor stalls and struggles, i think 10 miliseconds is enough, arduino senses that, and saves the motor position as an endpoint, searches for the other one.
I have no idea on the nano. from what i know, i need four interrupt pins for the 2 decoders. If nano isn't ok, i'll use a larger one, it's not an issue.
For the programmed focus points you mean memorizing and recalling focus or zoom position by pushing a button ? I am imagining , i don't know, five buttons, each for a point. Please elaborate on this subject.
 
Focus points - on a film camera, it is useful to have pre-defined focus points at the actors marks in a scene. So the focus puller can track the actor and start and stop at specified points that have been worked out before filming. So I see this working two ways:

- The focus-points act as hard stops, similar to those provided on a mechanical focus puller. The focus puller starts and the first hards stop but manually pulls focus with the encoder until it hits the second pre-programmed stop.

- the focus points act as start/stop points for a pre-programmed focus pull. The start and stop point are pre-programed. The focus change rate is pre-programmed. The focus operator then hits a button during the scene and the focus automatically changes from A to B at a pre-specified rate.

On some ENG/EFP lenses, the operator can pre-define zoom focal lengths. They can then snap zoom to those points by hitting a button. For film, a pre-programmed zoom rate from focal length A to focal length B at a specified zoom rate would be handy.
 
You must log in or register to reply here.
Back
Top