How big a mirror for wide lens ?

[Les Dittert]

I may be resorting to a beamsplitter rig until Red gets the sensor flip going again.
Just playing with the Tokina 11-16mm and a mirror, it seems like the beamsplitter would have to be HUGE to allow the 11mm side of that lens to be used.
Not only is the field of view wide, but something I didn't think of : The lens can't be close to the mirror, or else it will 'see itself' .
So my question: are people avoiding wide lenses, or what ?

Another Q: how well does a 1/4 wave plate help get rid of the polarization problem ? I hear it is helpful.

 

[Rich Schaefer]

I have found about a 14mm @ 4kHD to be about as wide as possible before you start seeing everything on a mirror rig (ET). More Rez like 5k... means same angle of view but more millimeters on the lens say like 17 or 18mm

I have tested the 1/4 wave, it helps a tad but only a tad, there are still disparities between the reflections. If you do some more testing and find definitive results, please share!

Prost,
Rich

 

[Michael Hastings]

I may be resorting to a beamsplitter rig until Red gets the sensor flip going again.
Just playing with the Tokina 11-16mm and a mirror, it seems like the beamsplitter would have to be HUGE to allow the 11mm side of that lens to be used.
Not only is the field of view wide, but something I didn't think of : The lens can't be close to the mirror, or else it will 'see itself' .
So my question: are people avoiding wide lenses, or what ?

Another Q: how well does a 1/4 wave plate help get rid of the polarization problem ? I hear it is helpful.

Les: remember the mirror is at a 45 to the lens, so it shouldn't see itself.

With scarlet/epic and our standard 14"x10" mirror on the Aero3D and a 50mm interaxial (the wider the iA the closer you get to the sides so you can't go as wide angle of view) we were able to get:

12mm @ 4KHD
12.5 @ 4K
and about 15.5mm at 5K

One nice thing about the Aero3D design is that it is fairly easy to make another set of the 3" Aluminum crosspieces to allow a wider mirror.

BTW the Aero3D was kind of put on the back burner for a bit while we were trying different glass manufacturers and then a fairly long lead time once we chose, but we are moving it to the front burner again and should be able to ship new units next month. The glass we came up with is made to the same specs as ET and is extremely good as I think quality of the beamsplitter glass is one of the critical elements.

 

[Les Dittert]

Hi Michael,

I understand how the mirror is at 45 degrees to the lens, but the lens has a 104 degree view.
So if I have the edge of the lens close ( touching ) the mirror, it does see itself.
How far away do you have to keep the lens so that isn't happening ? My rough tests with a hand held mirror seemed really far, like 8" or so.
I know that with a dive housing you are really trying to keep it small, so if anybody knows this it would be underwater guys like you.
Maybe I just didn't get the mirror at the correct angle, I was just doing a quickie gut feeling test is all.

I don't imagine you have no polarization issues under water either, right ?

 

[Michael Hastings]

Hi Michael,

I understand how the mirror is at 45 degrees to the lens, but the lens has a 104 degree view.
So if I have the edge of the lens close ( touching ) the mirror, it does see itself.
How far away do you have to keep the lens so that isn't happening ? My rough tests with a hand held mirror seemed really far, like 8" or so.
I know that with a dive housing you are really trying to keep it small, so if anybody knows this it would be underwater guys like you.
Maybe I just didn't get the mirror at the correct angle, I was just doing a quickie gut feeling test is all.

I don't imagine you have no polarization issues under water either, right ?

Now you have me confused ;-) we run the lens up to like a half inch or less away from the mirror with no issues.

The Aero3D base configuration is pretty inexpensive at $3995 and very solid and lightweight with the cnc machined aluminum sides and crosspieces but I also recently made some plastic side version for some testing purposes that you might be interested in trying out if you are in the experimental stage. It has the same high quality glass and the crosspieces and mounts are the same - you just don't get the aluminum side pieces but it is fully upgradeable to the aluminum sides later, and although it hasn't been advertised I've sold a few for $2995. The only issue is if you are doing handheld and do rapid moves you can torque and twist it a little bit. (I'm not sure if it isn't just as solid as a film factory or other 80/20 types of rigs but it isn't as rock solid as the aluminum Aero3D.) You also don't get the side flags but those are easy to make DIY out of plastic or even foam board - you do get the hinges.

Why the plastic? Well the aluminum I use is a special cast and ground tooling plate that has been stress relieved so it won't warp and is pretty expensive, I already have the aluminum in house but I wanted to do some plastic versions to test other configuration stuff before I commit to a final setup on the CNC and make 60 of them.

 

[Jason Goodman]

Les,

I cannot speak to your underwater questions but I'll give you my input on a few of the others. I agree with Michael that you should be able to get very close to a beam splitter mirror without having the lens see itself. One issue you can run into is one lens seeing the other lens and this is definitely related to FOV and proximity to the mirror. 21st Century 3D's current BX3.5 has been tested with lenses as wide as 15mm at 4K 16x9 and there is no vignette. Same lens at 5K on EPIC does see the left edge of the mirror box. BX4 is on the way but I will reserve information on final specifications for now.

As for quarter wave retarders they help much more than a tad. It should be noted that you will not always be in situations with polarized light, but when you are, they help. Quarter wave retarders were in place for the overwhelming majority of shots on The Amazing Spider-Man and after initial testing with our optics, stereoscopic sup on that film determined they " looked great and seemed to eliminate 80-90% of light differences presumed to be due to polarization". I am uncertain about the underwater factor but quarter waves are a good idea.

The mirror you use is also important. A good quarter wave retarder cannot solve all the problems of a bad mirror.

 

[Les Dittert]

Thanks for the info Jason. If I wanted to experiment with a 1/4 plate, do you know of any sources that are 'reasonable' as far as pricing? It's interesting that circular polarizers all have a 'depolarizer' on the back, but I can't find one with just that and not the linear polarizer glued on the front !

 

[Jason Goodman]

Les,

I am pretty sure there is no such thing as a "depolarizer" this is a marketing term 21st Century 3D was using when we first implemented the use of teh 1/4 wave retarder in an effort to keep it as a trade secret.

 

[Dave Blackham]

If your lens is over 82 degrees angle of view I think the corners of the image will be too soft due to mirror issue's also even if it did work the mirror box would be huge making it impractical.

 

[Maciej Orman]

I may be resorting to a beamsplitter rig until Red gets the sensor flip going again.
Just playing with the Tokina 11-16mm and a mirror, it seems like the beamsplitter would have to be HUGE to allow the 11mm side of that lens to be used.
Not only is the field of view wide, but something I didn't think of : The lens can't be close to the mirror, or else it will 'see itself' .
So my question: are people avoiding wide lenses, or what ?

Another Q: how well does a 1/4 wave plate help get rid of the polarization problem ? I hear it is helpful.

It works well for narrow view angles but at wide angle the retarder would have to be shaped spherical or else it will not work in corners and off axis.