PDA

View Full Version : Anamorphic lenses for 16x9 sensor?



Gregor Hagey
02-23-2008, 09:25 PM
I don't know if there's a thread on this yet, but are there Anamorphic lenses for a 16x9 sensor? Instead of a 2x squeeze it should be a probably be a 1.33 squeeze to get 1.78 to 2.39. I guess this would introduce a host of problems for exhibition though.

I Bloom
02-23-2008, 09:46 PM
I don't know if there's a thread on this yet, but are there Anamorphic lenses for a 16x9 sensor? Instead of a 2x squeeze it should be a probably be a 1.33 squeeze to get 1.78 to 2.39. I guess this would introduce a host of problems for exhibition though.

Hawks is building them. That's all they would tell me.

IBloom

Jeff Kilgroe
02-23-2008, 10:29 PM
I don't know if there's a thread on this yet, but are there Anamorphic lenses for a 16x9 sensor? Instead of a 2x squeeze it should be a probably be a 1.33 squeeze to get 1.78 to 2.39. I guess this would introduce a host of problems for exhibition though.

Why do you say it would introduce a host of problems? Dalsa is using 1.2X anamorphics...

There is also nothing keeping you from using 2:1 anamorphics on RED, you just have to crop more data off the sides. It's been discussed off and on here, especially lately with the introduction of 3K modes, but having dedicated anamorphic shooting modes in-camera. Jim has said that anamorphic functions are coming. So I'm expecting to eventually see support in camera for 1.33X and 2:1X anamorphics where monitor output is stretched to compensate and 2:1 modes will not record the excess data on the sides of the frame, which could give longer record times.

On RED, shooting anamorphic really has no benefits for resolution or detail. People will shoot it for "the look" of anamorphics. The stretched bokeh, elongated flares, etc..

There are various productions that have already made use of 2:1 anamorphics on RED.

David Mullen ASC
02-23-2008, 10:58 PM
There's no 'release problem' in the sense that you wouldn't release an image with a 1.34X squeeze anyway, you'd convert and transfer your 16x9 recording with a 1.34X optical squeeze to 35mm 4-perf anamorphic with the standard 2X squeeze, or you'd make a digital projection master with the squeeze removed, with a 2.39 aspect ratio.

It's just like shooting 3-perf 35mm -- it's not a projection format, just a shooting format. You'd convert it optically or digitally to 4-perf 35mm for release printing. Same goes for shooting Super-35 and framing for 2.39 -- you'd convert it to 4-perf 35mm 2X anamorphic for release printing.

The problems I see are more with monitoring on set (and in the viewfinder) and in editing, how to get rid of the 1.34X squeeze. Possible, of course, just not commonplace.

Gregor Hagey
02-24-2008, 09:42 AM
If cropping the Red sensor to 2.39 makes use of 4096x1719 pixels (7 megapixels) wouldn't a 1.33 squeeze of the full 4520 x 2540 sensor (11.5 megapixels) give noticeable resolution advantages? I realize that the picture would be unsqueezed and shrunk down to fit the DCI spec of 4096x1716 for exhibition in the end, but wouldn't the extra resolution help maintain better quality through post?

Kenn Christenson
02-24-2008, 10:36 AM
You'd have to account for the inherent loss of resolution when shooting through, what is essentially, an astigmatism.

David Mullen ASC
02-24-2008, 11:01 AM
There is some image improvement from using a larger area overall, but it is weighed against some optical loss from the anamorphic elements. I would describe the look as "smoother" and "cleaner" in some ways than I would say "sharper". It makes more sense in 4-perf 35mm where you have a nearly square negative format, so the 2X anamorphic format uses nearly twice as much negative than cropping to 2.39.

So the benefits of a 1.34X anamorphic on a 3-perf 35mm negative or 16x9 sensor like the RED is less significant, but I wouldn't say it was insignificant -- for one thing, a 1.34X anamorphic lens should be sharper than a 2X anamorphic lens, and if it is a front anamorphic design, you should still see classic anamorphic artifacts just as you do in old Ultra Panavision 65mm movies (1.25X squeeze) and Technirama movies (1.5X squeeze.)

Gregor Hagey
02-24-2008, 01:44 PM
Thanks for your answers. I guess time will tell how Anamophic lenses will be used with 16x9 digital sensors.

As a slight tangent, is it possible that 2.39 will ever be referred to as 64x27? I thought is was ridiculous when 1.33 and 1.78 were replaced with their fractional equivalents (what video engineer came up with this idea?).

William Wedig
02-28-2008, 09:37 AM
There is also nothing keeping you from using 2:1 anamorphics on RED, you just have to crop more data off the sides. It's been discussed off and on here, especially lately with the introduction of 3K modes, but having dedicated anamorphic shooting modes in-camera. Jim has said that anamorphic functions are coming. So I'm expecting to eventually see support in camera for 1.33X and 2:1X anamorphics where monitor output is stretched to compensate and 2:1 modes will not record the excess data on the sides of the frame, which could give longer record times.

On RED, shooting anamorphic really has no benefits for resolution or detail. People will shoot it for "the look" of anamorphics. The stretched bokeh, elongated flares, etc..

There are various productions that have already made use of 2:1 anamorphics on RED.

When is this happening? When did Jim mention that? I bet it's not a huge rush for the RED Team to get 2:1 Anamorphic support (as it wouldn't be used by many productions) but it would really help me with my production that's coming up in August. I really love the flares and bokah of Anamorphics myself though...

Perhaps a 4k [4,000 by 2,500 (roughly)] sensor cut in cam to a 2k with the complete vertical [2,000 by 2,500 (roughly)] would be a nice range for 2:1 Anamorphics... :biggrin:

Alexander Nikishin
02-29-2008, 12:50 AM
Hawks is building them. That's all they would tell me.

IBloom

:greedy:

Come to papa.

William Wedig
02-29-2008, 12:24 PM
Hate to be the annoying guy bumping his own question...

Again, does anyone know if Red has or has not actually stated that they want to eventually support a 2:1 Anamorphic lenses? I have a way to mount them, but it'd be nice if the output would unsqueeze the image for the eye piece and the lcd/monitor.

I's really like to have the blue horizontal flares, even more shallow DOF and the oval bokah.

Thanks

James T Mather
02-29-2008, 12:29 PM
The blue horizontal flares are more of an optical artifact of the C and E series panavision anamorphics really.

Fredrik Callinggard
02-29-2008, 01:35 PM
The blue horizontal flares are more of an optical artifact of the C and E series panavision anamorphics really.

Actually i find the E-series to be the blue one (Terminator, Lethal Weapon etc)

The C-series are more pink, I'd say (Blade Runner, Alien etc).

My preference being the C-series


fred

David Mullen ASC
02-29-2008, 02:42 PM
Anamorphic prime lenses with front cylindrical elements get the horizontal flare -- how blue it is depends on the coatings used -- I find the Primo anamorphics to be a quite deep blue, almost a neon blue, whereas C-Series has a warmer blue (if that makes sense).

Hawks put their anamorphic element in the middle of the lens so have less of that effect.

You can see some flare tests here, though the colors are too washed out to tell much:

http://www.claudiomiranda.com/Alpine.html
http://www.claudiomiranda.com/Alpine75mm.html

Here is a frame from "Akeelah and the Bee" that shows Primo anamorphic flares:

http://www.davidmullenasc.com/akeelah7.jpg

James T Mather
02-29-2008, 03:08 PM
Actually i find the E-series to be the blue one (Terminator, Lethal Weapon etc)

The C-series are more pink, I'd say (Blade Runner, Alien etc).

My preference being the C-series


fred

Sure, absolutely - my point was really more that those lenses specifically exhibit the horizontal streaking that he is talking about (C's and E's seem to be the most popular due to their tendency to flare easily) - The new anamorphics may not have this artifact and this may not be what he is looking for - (I'm hung over today so wasn't putting my all into the response - apologies ;)

William Wedig
02-29-2008, 03:46 PM
Gotcha, thanks for the tips!

But again, I'm more wondering more about whether or not Red is planning on supporting it or not. Do things like auto unsqueeze for monitor outs, crop the edges to save hard drive space, etc..

Jeff Kilgroe mentioned that Jim had talked about it. I'm guessing for now it's not a huge priority, but it would be nice to have it in the next few months possible. Perhaps build 20? 21?

Brandon Fraley
03-10-2008, 10:55 PM
i remember them RED saying they were working on it. And Soderbergh has already shot a large production using it. I dont see why the wouldn't pass it on to the rest of us. Maybe one if the big features of build 16, who knows.

Brandon Fraley
03-13-2008, 01:57 AM
i remember them RED saying they were working on it. And Soderbergh has already shot a large production using it. I dont see why the wouldn't pass it on to the rest of us. Maybe one if the big features of build 16, who knows.

or 15 lol :)

Nova Invicta
03-17-2008, 11:24 AM
The use of anamorphic on a 16x9 area calls for 1.34 squeeze and at this ratio you do not get enough of the anamorphic effect to really justify the investment in the glass. Joe Dunton built 1.5 squeeze lens prototypes but they never really took off most people want the full fat version and long may that continue because film should coexist with digital not be wiped out otherwise we should get Ford to start building again the Model T in only black!

Brandon Fraley
03-17-2008, 02:58 PM
The use of anamorphic on a 16x9 area calls for 1.34 squeeze and at this ratio you do not get enough of the anamorphic effect to really justify the investment in the glass.

I agree. I'd rather use normal Ana glass and take the res hit