Capture volumes larger than 10 feet

[bsteagal]

Good morning!

Has anyone tried to get a capture volume larger than the standard 10 feet with a 12+ camera setup? I have someone that is looking to get about 15-20 feet if possible. Would adding IR lamps to pump up the amount of light to the markers do the trick? (similar to what some recent productions like Beowulf have done?)

Just curious if this could be done and if anyone has attempted it..

Thanks!

Bryan

[Birch]

Bryan,

We have been testing with large capture volumes lately and have found that a 20x20 camera rig (resulting ~10x10 capture volume) is where the current configurations top out. We have only been using the on-board IR illumination though.

It might be possible to add additional illumination from behind each camera, it would need to be co-axial with the camera to work best with retro-reflective markers. You would want to be attentive to how much heat the light sources put out if they are near the cameras.

[bsteagal]

Birch:

I would probably be using something like the IR LED lamps used for surveillance and night vision cameras, which don't output any heat... so that shouldn't be too much of an issue
My thought would be installing 4-6 of these ir sources around the capture volume to flood the area, much more than what is done with the standard light source... I'm just wondering if this would give me the additional volume that I'm looking for..

Anybody else try this? Brad?

Thanks!

Bryan

[Birch]

Flooding the volume from arbitrary light locations would probably increase the ambient IR and reduce your signal to noise ratio (i.e. bad since markers illuminated from camera will have less contrast with background light). It might also create more false markers. It would be much better to mount one IR light source directly next to each camera (as close as possible).

You might want to get a single light source, attach it next to a camera and test to see what kind of range you can get before getting more.

Brad's blog entry on this has some good information about retro-reflectivity :
http://fie.us/blog/2008/01/how_moap_wor ... d_ret.html

[leith]

Its worth noting that when I've tried augmenting the light source in the past, I've been caught by the fact that the lense face, aperture and sensor are so small. you have to get the light source REALLY close to the lense for it to pick up as a result, and the best location for that is kind of already taken by the existing LED ring. In a comparison to other cameras, the smaller sizes do seem to be a large factor. Working with a SLIM camera and making a different LED ring might be easier at some point.

That being said, I have had some limited success with mounting halogen projector bulbs about a foot behind the cameras. I suspect an augmentation such as that will work better the farther the marker is from the camera but I've not been able to confirm that just yet.

I fear though, that we may be bound by the inverse square law and its pesky inverse nature with regard to light and camera. The key will be to cleverly get around it. And so far that has been challenging. It may turn out to be more about the other characteristics of the light source rather than its general ability to put out light. Such as say, the compactness or sparseness of the light face (clear glass vs. frosted). The extreme example would be bare bulb vs. softbox. Retroreflection is similar to specular reflection and therefore it may make more sense to think about the brightness as an actual reflection of the light face rather than bounced incident light. Anyhow, food for thought. I've not finished my experimentation in this area yet. Sorry.