Underwater 3D

[Ross Isaacs]

Has anyone shot 3D underwater using 2x Red One cameras? I am interested in building an underwater rig to accommodate 2 Red One cameras if anyone has any technical information on this subject or know where I can find it?

[Clark Dunbar]

DJ Roller of Liquid Pictures has.....

RED Underwater Stereoscopic/3D (http://www.liquidpictures3d.com/Liquid_Pictures_3D.html)

[Pawel Achtel]

Has anyone shot 3D underwater using 2x Red One cameras? I am interested in building an underwater rig to accommodate 2 Red One cameras if anyone has any technical information on this subject or know where I can find it?

hahaha, sounds like a good way to get hernia. :biggrin5:

[Ross Isaacs]

"my lips are sealed" I hear you shout Pawel

[Michael Hastings]

Has anyone shot 3D underwater using 2x Red One cameras? I am interested in building an underwater rig to accommodate 2 Red One cameras if anyone has any technical information on this subject or know where I can find it?

As he said, DJ has done it with a beamsplitter. A monster, but friends that have seen the footage say it is phenomenal.

Epic/Scarlet should have some major advantages for 3d.

[Ross Isaacs]

As he said, DJ has done it with a beamsplitter. A monster, but friends that have seen the footage say it is phenomenal.

Epic/Scarlet should have some major advantages for 3d.

Thanks Michael
I contacted DJ but its inhouse gear. I am planning some deep dive projects and need something not too logistically cumbersome. Seems like the twin scarlet idea might work although I am not sure the IO distance would work. I don't see how a dome port would work.....complex stuff

[Michael Hastings]

Thanks Michael
I contacted DJ but its inhouse gear. I am planning some deep dive projects and need something not too logistically cumbersome. Seems like the twin scarlet idea might work although I am not sure the IO distance would work. I don't see how a dome port would work.....complex stuff

What kind of subjects? We have been playing with some Canon single chip 1080P camcorders in side by side that allow a 3.1" intraocular (including wide angle conversion) and will be trying some new ones that allow less than 2.7" (so probably significantly smaller than Scarlet would allow.) And remember AVATAR is all 2K (just slightly more than 1080P) and supposedly the two images create a sharper overall impression.

The rule of thumb I have seen is your main/normal subject distance should be about 30x the intraocular. That equals 7.7 feet for 3.1" or 6.7 feet at the 2.7" intraocular. In practice we seem to be fine at half of those distances (and sometimes even less) but so far we have only been showing 3 minute pieces of some shark footage I shot over at Stuart Coves on a vacation trip so it is probably better not to push it constantly on longer pieces. But the effect is very impressive. For deep wrecks or something like that it would probably be excellent where using a beamsplitter rig (even a smaller one that could be made for scarlet/epic) might be a nightmare.

Dome ports are probably only feasible for long shots (15 feet plus) where you could separate the cameras by six inches or more to accommodate the domes, but I will probably make a housing like that eventually. Trying to shoot through a single large dome with both cameras would create aberrations in opposite directions and probably not worth the trouble.

[Ross Isaacs]

Thanks Michael
so does that mean the underwater beam splitter style rig would cover everything from close ups to our wides?

[Markus Stone]

Hi Ross,

Just to chime in here - we've been looking into this for a client. Essentially, if you want to get close to your subjects, you'll likely require a beamsplitter rig. 2 Reds on a beamsplitter are by their very nature buky - I can't see any way around that - it'll be like swimming around with a small fridge attached.

The alternatve is to use small rigs designed to work with small cameras, like the SI-2K mini's (or other 'sensor on a wire' type cameras), but you need to have a lot of electrical connections passing in and out of the housing as they do not record locally, which is the trade off for having something truly managable in terms of size. The need to use c mount lenses to fit wthin the rig can also be an issue for some.

So you can get smaller, but it presents its own set of issues.

Regards

-Markus

[Ross Isaacs]

Hey thanks Markus
I am thinking that the Red Scarlet 3D setup might be the way to go although it seems the IO distance may only be suitable for medium distance shots. I am planning a deep dive film max depth 120 meters so dragging a fridge around might not be such a good idea. Perhaps a beam splitter rig with 2 scarlets might be an eventual solution?

[Markus Stone]

I am planning a deep dive film max depth 120 meters so dragging a fridge around might not be such a good idea.

...yes, but it does keep the beer cold :-)


Perhaps a beam splitter rig with 2 scarlets might be an eventual solution?

Yes, that's what I'm thinking too. 2 scarlets recording to CF could be a great solution.

Cheers

Markus

[graemeduane]

Michael
How does the footage look from the two Canon's? Doesn't it look washed out behind flat glass? Did you lanc control those units? Did you flip one camera?

I agree that SS is easier underwater, but getting the lenses close is a real mission.

[Michael Hastings]

Michael
How does the footage look from the two Canon's? Doesn't it look washed out behind flat glass? Did you lanc control those units? Did you flip one camera?

I agree that SS is easier underwater, but getting the lenses close is a real mission.

It doesn't look washed out, it just isn't quite as good as the same lens through a dome because you lose 25% due to magnification and therefore shooting through that much more water so you lose that amount of clarity and saturation. We try to compensate by using wider lenses than we might otherwise.

Why would you flip the camera on the side by side version? - with the cameras we are using the limiting factor is the wide angle lenses. It was a little over three inches on the ones I shot with in November and February - the ones I am building now are 2.65 IO.

The current cameras don't have LANC. One of the new Canon models has LANC so will be implementing that when it is available.



Thanks Michael
so does that mean the underwater beam splitter style rig would cover everything from close ups to our wides?

Yes, but understand that the wide shots are easy even on side by side (anything say 3-4 feet and beyond - you can cheat closer for a short time.) It is the macro stuff that demands the beamsplitter. Once you have the beamsplitter you would probably use it in wide mode too because then you could get a little closer and also you could use a dome.

[Ross Isaacs]

Yes, but understand that the wide shots are easy even on side by side (anything say 3-4 feet and beyond - you can cheat closer for a short time.) It is the macro stuff that demands the beamsplitter. Once you have the beamsplitter you would probably use it in wide mode too because then you could get a little closer and also you could use a dome.

Hey Michael
it is my understanding that you can't use a dome port with a beam splitter rig underwater as the 2 x optical axis would not line up with the centre of the dome port.

[Ross Isaacs]

Hey Michael
it is my understanding that you can't use a dome port with a beam splitter rig underwater as the 2 x optical axis would not line up with the centre of the dome port.

Hey Michael
I was more asking you a question not making a statement to the above :smile5:

[Dave Blackham]

Hey Michael
it is my understanding that you can't use a dome port with a beam splitter rig underwater as the 2 x optical axis would not line up with the centre of the dome port.

It may be possible with a common objective system but only for macro which is were common objective works best. However we think have a very specialist optical solution for twin lens fixed IO cameras that are appearing on the market. Its in test right now and we hope to have more info soon.

[Ross Isaacs]

It may be possible with a common objective system but only for macro which is were common objective works best. However we think have a very specialist optical solution for twin lens fixed IO cameras that are appearing on the market. Its in test right now and we hope to have more info soon.

Thanks Dave...it was my understanding (please correct me if I am wrong) that to create stunning 3D images which really work an IO is required to produce a parallax separation for the eyes. These single lens system eg. Sony's new single lens system might not be as effective in creating the illusion of 3D as a twin lens or beam split system.

[Michael Hastings]

It may be possible with a common objective system but only for macro which is were common objective works best. However we think have a very specialist optical solution for twin lens fixed IO cameras that are appearing on the market. Its in test right now and we hope to have more info soon.

alright, you guys are confusing me in several ways :-)

First, why would you want to use a dome port for macro? The basic idea of the dome is to correct two things 1) to correct for the 25% magnification you get with a flat port which makes your lens less wide and the idea of being as wide as possible is that to get the same size subject in the frame you can move closer thus reducing the amount of water you are shooting through and increasing both sharpness and color saturation. You also get the added benefit of reducing the effect of unwanted camera motion and your lights are effectively stronger because the light doesn't travel as far through the water so maintains more of its color and brightness. and 2) the dome corrects for some of the loss of edge sharpness and chromatic aberrations that happen with wides behind flat ports because the light rays from the edges of the frame are hitting the flat port at extreme angles whereas if properly aligned all of the light rays hit the dome perpendicularly.

One thing to remember though (that is important when we talk 3d) is that the dome creates a curved virtual image and some "fisheye" distortion. The problem with 3d is that the lenses can't be centered so any distortion is in opposite directions - and the images don't overlap properly.


Although we often shoot macro with zoom lenses through dome ports -because it is easier than swimming to the surface to change to a flat port - it isn't ideal. You generally shoot macro with telephoto lenses, so you don't have the problem of steep off axis light rays and you don't need to be 25% wider to get closer. You generally choose your shooting distance to macro subjects based on how close you can comfortably work - often a foot or less - so you aren't shooting through a lot of water.

So shooting 3D macro through a flat port is just fine anyway.

[Ross Isaacs]

alright, you guys are confusing me in several ways :-)

First, why would you want to use a dome port for macro? The basic idea of the dome is to correct two things 1) to correct for the 25% magnification you get with a flat port which makes your lens less wide and the idea of being as wide as possible is that to get the same size subject in the frame you can move closer thus reducing the amount of water you are shooting through and increasing both sharpness and color saturation. You also get the added benefit of reducing the effect of unwanted camera motion and your lights are effectively stronger because the light doesn't travel as far through the water so maintains more of its color and brightness. and 2) the dome corrects for some of the loss of edge sharpness and chromatic aberrations that happen with wides behind flat ports because the light rays from the edges of the frame are hitting the flat port at extreme angles whereas if properly aligned all of the light rays hit the dome perpendicularly.

One thing to remember though (that is important when we talk 3d) is that the dome creates a curved virtual image and some "fisheye" distortion. The problem with 3d is that the lenses can't be centered so any distortion is in opposite directions - and the images don't overlap properly.


Although we often shoot macro with zoom lenses through dome ports -because it is easier than swimming to the surface to change to a flat port - it isn't ideal. You generally shoot macro with telephoto lenses, so you don't have the problem of steep off axis light rays and you don't need to be 25% wider to get closer. You generally choose your shooting distance to macro subjects based on how close you can comfortably work - often a foot or less - so you aren't shooting through a lot of water.

So shooting 3D macro through a flat port is just fine anyway.

precisely what I thought Michael

[Michael Hastings]

precisely what I thought Michael

Great minds think alike :-)

(Just thought I'd bump the thread and see if anyone is thinking 3D this summer) been doing inexpensive 3D with Canon and Sony HIDEF single chip camcorders (trying to figure out how to sync the Panasonic 700 series camcorders since they are 3 chip and do 1080/60P at 28Mbps H.264) - housing in the $2K-$3K range)

[Clark Dunbar]

YES - 3D!! this summer.... We've launched the MammothHD-3D stock footage website (www.mammothhd-3d.com) and having just added AJA as a production partner/sponsor... we ready to crank!

[Dave Blackham]

alright, you guys are confusing me in several ways :-)

First, why would you want to use a dome port for macro? The basic idea of the dome is to correct two things 1) to correct for the 25% magnification you get with a flat port which makes your lens less wide and the idea of being as wide as possible is that to get the same size subject in the frame you can move closer thus reducing the amount of water you are shooting through and increasing both sharpness and color saturation. You also get the added benefit of reducing the effect of unwanted camera motion and your lights are effectively stronger because the light doesn't travel as far through the water so maintains more of its color and brightness. and 2) the dome corrects for some of the loss of edge sharpness and chromatic aberrations that happen with wides behind flat ports because the light rays from the edges of the frame are hitting the flat port at extreme angles whereas if properly aligned all of the light rays hit the dome perpendicularly.


Yes agreed, for macro a flat port should work just fine with common objective.




One thing to remember though (that is important when we talk 3d) is that the dome creates a curved virtual image and some "fisheye" distortion. The problem with 3d is that the lenses can't be centered so any distortion is in opposite directions - and the images don't overlap properly.


What we are aiming to do is create an optic so that both the lenses can be properly centered and angle of view preserved, any "fisheye" distortion will match in both eyes.



Although we often shoot macro with zoom lenses through dome ports -because it is easier than swimming to the surface to change to a flat port - it isn't ideal. You generally shoot macro with telephoto lenses, so you don't have the problem of steep off axis light rays and you don't need to be 25% wider to get closer. You generally choose your shooting distance to macro subjects based on how close you can comfortably work - often a foot or less - so you aren't shooting through a lot of water.

So shooting 3D macro through a flat port is just fine anyway.

But shooting 3D WA could be much better shot through specialist optics.......

[Michael Hastings]

Dave:

Not sure what you are driving at here - there is no problem making two side by side dome ports for two cameras, but it would be pretty far apart unless you had really small cameras and really small domes (which creates other issues).

Are you talking trying to make something for side by side or for beamsplitter (beamsplitter obviously has size issues along with a bunch of others.)