Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Write me an iPhone App for...

Jason Beckwith

Jason Beckwith

Well-known member
Joined
Feb 11, 2007
Messages
443
Reaction score
0
Points
0
Location
Santa Cruz, CA
Could one of you smart people write me an iPhone App that uses the GPS compass and inclinometer functions to record the phones' movements at different time code bases like 23.976, 24, 29.97, 30, etc.?

I'd love to build a little rig out of Zacuto rail parts or something that could hold the phone on top of the camera. That way, every movement on tripod would be recorded. I don't think the GPS has fine enough detail to allow hand-held, but it would be awesome if it did. As it stands, it would even pick up dutch angles.

It would be even better if there were a way to genlock it to the camera, and verify time-code plus get start/stop info. That's hardware, and hoping for too much, I fear.

Even if it had to be started separately, the data could be slipped to match the footage pretty easily. Even if the data was a little off, it would still give a nice starting point to begin a track with a virtual camera.

Maybe an Epic module could handle this as well?

Or is this just a stupid idea that wouldn't work?

Opinions?

---Jas
 
Hmmmm... I wonder what 'any length of time is'...

I mean, will it hold synch for the amount of time it takes to fill an 8gig CF card?

I was just thinking getting data on set might save a couple of trips through PFTrack, boujou or synth-eyes!

There's gotta be someone on here that has the iPhone dev kit and can give it a quick test! It should just be a matter of API calls and recording against the internal clock on the phone, right?

---jas
 
You mean like for motion control tracking? I think you're greatly overestimating the accuracy of GPS data, especially on the iPhone. I'd suggest going with standard servo encoders or a regular motion control system.

-Noah
 
Well, yeah...

I'd much rather put my camera on a milo for anything critical.

I was just toying with a couple apps, like that bubble level thing and then searched the app store for a compass and inclinometer and saw that there were a bunch available. I was mostly thinking about pan/tilt data, but that tracks pretty fast anyhow.

We often use chroma green for our backdrop, and digital green for tracking markers. I often have to pull a key on the digital green and render before I can take it into synth-eyes. That extra render and track can take a pretty good amount of time and can quickly balloon a client's post budget over multiple shots. But then again, trying to match the data to each shot by slipping it in time, if there is no way to synch start/stop the app, could eat a bunch of post time as well.

I'll trust you guys, though. If you say it's not worth looking into, then so be it.

Just thought I'd treat the idea like spaghetti... Throw it on the wall and see if it stuck!

---jas
 
Jason Beckwith said:
Hmmmm... I wonder what 'any length of time is'...

I mean, will it hold synch for the amount of time it takes to fill an 8gig CF card?

I was just thinking getting data on set might save a couple of trips through PFTrack, boujou or synth-eyes!

There's gotta be someone on here that has the iPhone dev kit and can give it a quick test! It should just be a matter of API calls and recording against the internal clock on the phone, right?

---jas
Click to expand...

IPhone apps are very limited in how much control they can take of the device, and there are all kinds of other things that the operating system does to interrupt their continuous operation. Because well... it's a phone.

So touch screen and slick interface aside, you'd be better off with one of the many modular computing or robotics platforms that are sold as kits. I like these http://www.gumstix.com/.

IBloom
 
If you want motion capture on the incredibly cheap you would be best off with any of the following over GPS:

1) a real moco rig.
2) An optical 3D tracking system.
3) A camera tracking program like PFTrack.
4) An inertial difference tracker (Like a wiimote).
5) A guy with a good vantage plotting the position of the camera from above on an etch a sketch.
 
As others pointed out, the iPhone GPS isn't accurate enough for this sort of thing. Neither are the tilt functions. The tilt and accelerometer have really good accuracy within specific ranges, but would not serve very well for this sort of application. And with the GPS only accurate to something like 3.5 meters, give or take, you won't have any way of tracking accurate motion data.

The upcoming iPhone OS 3.0 will give more control over these functions, but we're still limited by the hardware. I think tracking accurate timecode may actually be possible, to some extent. But there are plenty of other OS issues to deal with, such as interruptive processes and sustained application focus. You would be better served to run the app on an iTouch -- at least then the app wouldn't be stalled while the phone notifies you of an incoming call.

If you have a specific need for a MoCo project / shot, etc.. I say just rent a proper MoCo dolly system. That's not cheap, but neither is custom software development -- especially developing software on a platform that isn't really right for the job and then trying to make it work. :)
 
The current iPhones don't have a compass chip. So horizontal bearing can only be approximated by using GPS movement triangulation.

The next iPhone generation will probably include a magnetic compass chip. This will allow you to to build this kind of app for static camera positions (pan/tilt/dutch angle). But I still doubt that you will be able to read accurate enough values from the compass and accelerometer in the exact timing you want.

And concerning moving cameras, GPS is not accurate enough for this kind of application.
 
You'll never get properly accurate timecode sync on a current iPhone.
Not while it uses a threaded processor, and the Apps run below the OS.

However conceivably, with the new iphone blue tooth integration and possible compass chip you could probably rig a makeshift solution out of something like a wii controller and a timcode generator, using the iphone to log the data, but you're probably far better off going the circuit board and solder method.

My 2cents.
 
Time: You would spend more time cleaning terrible data than using an application which actually works.
Money: Syntheyes costs $400. Something tells me you don't want to have your only phone strapped to a camera (We've actually done this before for a product shoot when the agency couldn't get us a phone in time. Pretty hillarious to see the DP trying to take calls from the agency on a phone mounted to a CStand.) So there is your $400 for another iphone right there. Also if you have to deal with the time to cleanup the noisy GPS data then you're spending money by the hour. Syntheyes would probably pay for itself over a free iPhone app in what 2 hours?
 
if you have a usable data, ready to use in seconds.. will it change ur perspective?
 
Yes.

But I'll believe it when I see it. :D

I've never seen an inertial tracker produce good enough data even with dedicated hardware. And GPS is noisy down to at best a foot which for most camera tracking is completely unacceptable.

What I would love to see is something where the two work together. Get a rough track based on inertia. Compensate for drift with GPS. Lock it down with an optical tracker using the Inertial tracker as a 'hint'.
 
You must log in or register to reply here.
Back
Top