www.ssontech.com

[drkollossus]

Hey, all! So... I'm doing a motion test for an upcoming spot. I have to track a phone as some one is texting on it. The shots have not yet been designed (that is why I'm doing the test), so I don't yet know the range of motion I'll have to cover. But what I'm thinking is placing three dots on the back side of the phone, and making sure the performers hands don't cover them. I also have a small stick with a suction cup to stick to the face of the phone. if it turns sideways.
Do you think that will be enough to get a solid track from? I'm worried that the phone is so small that it will be hard to get a good track from.
Any ideas comments or help would be much appreciated.

Cheers

[himkhu]

I liked the idea of sticking a stick on the phone for the depth but i also suggest u to also take the measurement between the markers and stick
and make a 3d model out of the measurement also if u know the distance of the object from the camera will be great for adding some more constrain.
you can bring that 3d model in syntheyes for geometry tracking

[Toadguy]

Hi,

I had to track a mobile phone about 18 months ago. It was to replace the handset with a CG model - it had to be unbranded but they wanted to keep the real screen. We put quite a lot of markers on the phone - but then it didnt matter because we were getting rid of the real thing completely.

One thing that will really help is if you also shoot reference frames of the phone, run the camera and then slowly rotate the phone about so you get various other angles on the markers. You only need to extract single frames to use in tracking.

You then use them at the end of the sequence to get more information about the relationship between the markers - basically follow the tutorial by rtep in this thread

http://ssontech.com/phpBB2/viewtopic.php?t=727

just using the reference frames of the phone at the end instead of what he does with his markers.

Hope that helps.

Toadguy