3D Shooting

Craig Howie · July 5, 2007

hello everyone.

i'm about to start experimenting with 3D filming. however, i'm in need of some advice. since i want to shoot 35mm, it would be ideal to use a single camera method (save $$). i've read a bit about shooting with a single camera into two angled mirrors with a set up such as this:

http://www.the3drevolution.com/3dbuild.html

but i have a few questions before starting:

1. after i digitize all the footage, how do i separate the side by side images and overlap them again? can this be done with FCP?

2. i've never shot anamorphic before and was wondering when the 'unsqueezing' was done. at the transfer stage? or again is this something that's done with FCP?

3. anyone know a way of utilizing a single camera that doesn't require buying an anamorphic lens? i suppose i could just use a normal lens but then the aspect ratio would be vertical rather than horizontal. that wouldn't look so great.

any help appreciated. thanks!

Nick Mulder · July 5, 2007

3. anyone know a way of utilizing a single camera that doesn't require buying an anamorphic lens? i suppose i could just use a normal lens but then the aspect ratio would be vertical rather than horizontal. that wouldn't look so great.

any help appreciated. thanks!

Hmmm - I've always wondered if up and down parallax would yeild a 3D image to viewers once it was combined for the standard L&R viewing ...

If it did, you could shoot with your camera sideways and youd have a landscape aspect again ...

time to experiment!

Leo Anthony Vale · July 7, 2007

Hmmm - I've always wondered if up and down parallax would yeild a 3D image to viewers once it was combined for the standard L&R viewing ...

The veiwer would have to be laying on his side.

Chris Keth · July 8, 2007

Hmmm - I've always wondered if up and down parallax would yeild a 3D image to viewers once it was combined for the standard L&R viewing ...

If it did, you could shoot with your camera sideways and youd have a landscape aspect again ...

time to experiment!

I think it would make people sick, or not work at all.

Nick Mulder · July 8, 2007

If I watch normal anaglyph 3D - and then rotate the screen 90 deg, its still 3D ... (No vomit ...)

So I dont think the laying on the side thing is relevant - the difference is the input parallax vs. output parallax

Any reason why it 'wont work at all' ?

up and down is a rotation of 90deg - I suppose a rotation of 180deg would be the same as swapping L with R - ie. inverse perspective - so 90deg is somewhere inbetween the two... a flat field ?

Really need to stop the theory and resort to some blatant empiricism ... Building a rig at the moment, 2x16mm (Sync'd bolexes) want to have it so there is no convergence like you would have with the mirror set up... (Unless instead of angling the mirrors you offset one further behind the other ?)

3D is hedfuq :blink:

Craig Howie · July 9, 2007

Building a rig at the moment, 2x16mm (Sync'd bolexes) want to have it so there is no convergence like you would have with the mirror set up... (Unless instead of angling the mirrors you offset one further behind the other ?)

hi nick. you've obviously given 3D shooting a lot of thought.

just wondering what you mean exactly by the 'convergence of a mirror set-up'? is this a major drawback i haven't considered?

cheers

Nick Mulder · July 9, 2007

hi nick. you've obviously given 3D shooting a lot of thought.

just wondering what you mean exactly by the 'convergence of a mirror set-up'? is this a major drawback i haven't considered?

cheers

A bit of thought, but I'm no expert - for me at least it seems like the kind of thing where real world experience craps all over whatever theory you can think up ... Just it takes time & $$ to set things up to get this real world experience.

I'm building some stuff at the moment but its a two cam set up, where I can set the separation and parallax up easy as 123 I hadn't thought of the one cam mirror set-up, but its not really super cool for 16mm - resolution counts huh

Anyhoo, parallax 'problem' >> I'll try to find it again, but there was this site with a trillion bits of info re. this that and the other - and I remember reading that parallax isn't so fancy pants after all, which went against my intuition - I mean we all do parallax don't we ? - but then I got to thinking that on a 3D film we are in essence forced to watch the 'correct' parallax of one point - if we were to wander our eyes elsewhere on the screen our eyes aren't going to be able to converge to get the parallax of that point, they can only converge on the screen. So as film makers I assuming we have to make damn sure we catch peoples attention in the areas that the parallax is correct - now so back to zero parallax - i.e.. parallel cameras... maybe indeed, this solves the problem ? the site I mentioned seemed to be heading that way, but I didn't read into it with busy times and various illnesses at the mo' ...

I am certainly looking forward to test it - also testing low DOF with tele lenses (which will have to converge or be so close as to fuse into each other - your mirror cam 35mm will do a much better job with that)

And making a rotating cam set up - so the cameras will maintain a horizon with respect to the world (i.e.. a rotation around the lens axis) but will also rotate around a common point 1/2 bet them... the parallax will go from correct to up and down to inverse all in one shot. I'll combine it in standard L-R anaglyph and watch it - I'll take a little luncheon satchel and some MacGyver type tools/tapes in case It zaps me forward, reverse or sideways in the space-time continuum ..