Super Fast Lenses (<T1.0) for S35 (or larger) Sensors

[Mathew Joki]

Of course most of us know about the Super Fast T0.95 lenses by Voightlander, and the new NOKTOR SLR Magic Hyperprimes, that are only compatible with M4/3 sensor sizes.

My questions are:
1) Are there ANY prime lenses for FF or S35 sensor sizes that are faster than f1.2? Or better yet, faster than f1.0?
2) If not, what is it about larger image (sensor) formats that makes it either impossible, or unfeasible to manufacture lenses to meet this low of a T/f-stop?


I ask mainly out of curiosity. I realize there's likely a lot of technical reasoning behind this dealing with optics, etc. which is something I don't know enough about. Any insight would be greatly appreciated, as I'm always interested on learning how the optics, and other processes that create the image on screen work, in order to exploit their capabilities and push their limits.

So far the fastest I've seen are the Canon EF 50mm & 80mm f/1.2 L USM, which are FF photography lenses, which has made the wonder that if there is some sort of correlation between image (sensor) size and minimum aperture, why aren't there any S35/ASP-C lenses faster than (or equal to) f1.2? Not just asking about Cinema primes here, but Photo primes as well.

Thoughts, oppinions, and explanations greatly appreciated!

 

[Jacek Zakowicz]

If you put a focal reducer on the 50 f1.2 it would become a 35-40 f1 lens for S35/apsc format

 

[PatrickW]

leica .95 noctilux ?

 

[Timur Civan]

Panavision 50mm f1.0...... Yeaaaaaaaaaaaaaa

 

[Christopher Mills]

yes, physics does govern the coverage and stop of lenses, and thus the size..
Here's a link to the equations for measuring light transmission (and other optical things)..

http://pleasemakeanote.blogspot.com/2010/10/mathematics-of-fstop-aperture-numbers.html

I especially note that for every doubling of light transmission, the aperture opening must be another 41% larger ..

IE.. an iris opening for f 1.4 is 41% more open on the same lens than that for f2... so .95 is another 41% than f 1.4?

it looks like the short answer is not really... which led to a filmstocks CCD and Chip development becoming more and more light sensitive...

and oh, yeah.. don't forget that chromatic abberation gets more prevalent at wide apertures.. which then leads to corrective lenses being required, which affect light transmission.. etc


when you also take into account the circle that must be projected to cover the target (frame of film or sensor) and the distance from the focal plane needed to get focus on acceptable distance objects, and achieve a useful focal length.. and oh yeah, be practical to use on a portable camera... well.. engineering hits some end points of practicality and cost..

I did leave out light transmission of the optical elements as well.. sorry...

 

[Nick Gardner]

The practical limitations of DOF come into play as well. Trying to follow focus on any kind of a real shot like you would do in a movie with camera and talent moving at anything wider than an f2 gets damn tricky. It's damn near impossible in a CU at an f1.3, being in noctilux world most ACs would stab you with a fork.

Nick

 

[Tom.Wong]

The practical limitations of DOF come into play as well. Trying to follow focus on any kind of a real shot like you would do in a movie with camera and talent moving at anything wider than an f2 gets damn tricky. It's damn near impossible in a CU at an f1.3, being in noctilux world most ACs would stab you with a fork.

Nick

still done all the time though, AC's have to go with it. just did a commerical with master primes and the whole thing was shot a 1.3

been seeing it more and more, stacks of ND indoors just to get the shallow dof the DP wants. they just had to get a lot of marks and time everything carefully. tricky yes, impossible no, it's just really rough when you don't let the AC do what they need to do to be able to hit that focus properly rushing everything.

 

[paul ross jones]

Panavision 50mm f1.0...... Yeaaaaaaaaaaaaaa

i like the canon better in this example.

paul

 

[Roger Valdes]

Because the faster the lens, the harder (and more expensive) it is to correct for chromatic aberration and other lens imperfections.
Bigger formats also means bigger lenses, more weight and higher cost. Which is one of the reasons why there are plenty of fast c-mounts lenses around. They're very small.

Here's a list of lenses of f1.0 and below that cover atleast full frame. MF stands for Medium Format

Fujinon IDEAX 125mm f0.67 (MF)
Kinoptik 60mm f0.7
Zeiss Planar 50mm f0.7
Aerojet Delft 150mm f0.75 (MF)
Super Farron 76mm f0.87
SLR Hyperprime 50mm f0.92
Astro Berlin 52mm f0.95
Canon 50mm f0.95
Perkin Elmer 114mm f0.95
Leica 50mm 0.95
Wild Heerbrugg Reconar 98mm f1.0 (MF)
Canon 50mm f1.0
Panavision 50mm f1.0
Pacific Optical 25mm f1.0 (MF)

http://www.flickr.com/photos/76802428@N04/7705425000/in/photostream

 

[D Fuller]

Because the faster the lens, the harder (and more expensive) it is to correct for chromatic aberration and other lens imperfections.
Bigger formats also means bigger lenses, more weight and higher cost. Which is one of the reasons why there are plenty of fast c-mounts lenses around. They're very small.

Here's a list of lenses of f1.0 and below that cover atleast full frame. MF stands for Medium Format

Fujinon IDEAX 125mm f0.67 (MF)
Kinoptik 60mm f0.7
Zeiss Planar 50mm f0.7
Aerojet Delft 150mm f0.75 (MF)
Super Farron 76mm f0.87
SLR Hyperprime 50mm f0.92
Astro Berlin 52mm f0.95
Canon 50mm f0.95
Perkin Elmer 114mm f0.95
Leica 50mm 0.95
Wild Heerbrugg Reconar 98mm f1.0 (MF)
Canon 50mm f1.0
Panavision 50mm f1.0
Pacific Optical 25mm f1.0 (MF)

http://www.flickr.com/photos/76802428@N04/7705425000/in/photostream

That is quite a collection! Are you really selling them? Thet must hurt.

 

[Nick Gardner]

still done all the time though, AC's have to go with it. just did a commerical with master primes and the whole thing was shot a 1.3

True, I did two commercials last week that were the same - master primes wide open. Of course you can get away with shit on commercials that you can't get away with on a movie, but I really don't see much need for faster lenses. If you are looking for ultra shallow dof, just shoot swing shifts. I'm sure all the rental houses that bought a set in the 90's would love that ;-)

Nick

 

[Tom.Wong]

you should talk to deakins about that then, he's been shooting wide open on masters a lot more as of late

but at his level, it's a aesthetic choice, not really cause there isn't enough lights to jam on set. the way it should be imho.

 

[Ryan De Franco]

There are so many blisteringly fast 16mm lenses it would be impossible to start counting. The smaller format does you three favors: the lenses are smaller, cheaper, and none involve a fork in the arm from the AC. Shooting 35, you need a 50mm lens to achieve the field of view a 25mm lens gives you on 16. A 1.0 lens might only give you a few inches of acceptable focus, but shooting 16, you've got double the distance compared to 35.

At the risk of sending prices through the roof (again), there are Angeneiux, Kern-Switar, Kowa etc primes that deliver incredible quality wide open. and i'm talking textural quality, not optical quality, I've never measured these things with charts just blown away by the images.

PS:
To bring a bit of Deakins forum-lurking to reduser.... he loves the master primes for their cleanliness. while many of us would take superspeeds for their character, he sold his set because he didn't like the small flares off point sources, especially head lights

 

[Ryan Patrick O'Hara]

There are so many blisteringly fast 16mm lenses it would be impossible to start counting. The smaller format does you three favors: the lenses are smaller, cheaper, and none involve a fork in the arm from the AC. You need a 50mm lens to get the same field of view a 25mm gives you on 16, giving you twice the depth of field to pull with.

We all know what you mean, but some noobies might read this the wrong way. I hope you don't mind but I'd just like to clarify what you are saying, since at first I read this as saying you need a 50mm lens to get the FoV of a 25mm lens when using 16mm... but realized that is not what you were saying.

To clarify: "You need a 50mm lens (on 35mm) to get the same field of view a 25mm gives you on 16mm"

 

[Mathew Joki]

Wow. Lots of great responses and information here. Thanks a lot everyone.

 

[Ryan De Franco]

Ah thanks, tried to make that a little more clear now. Sorry to take us into 16mm land. Can't resist:

Dragon is 30.7mm. The 16mm image circle is 14.5mm diameter... will we get 3K (almost) out of S16 lenses? Enough to debayer down to 2K or 1080? Forget proxy recording, this is how you cut down on recording time and storage space... go shoot documentary, concert, wildlife etc with a 10x T/2.2 zoom weighing 4 lbs that sells for $400 on ebay now....

And of course all that fantastic .95, 1.0 glass. Right, back on track, sorry folks.

 

[Jeff Kilgroe]

Dragon is 30.7mm. The 16mm image circle is 14.5mm diameter... will we get 3K (almost) out of S16 lenses? Enough to debayer down to 2K or 1080?

Actually we don't know the official specs and dimensions just yet. We know that the photosite or pixel size is approximately 5 microns and that we'll be able to shoot 5K within the S35 aperture. So far all the dimensions and resolution numbers I've seen posted have been labeled incorrect by Jim and Jarred. Anyway, we should be close, I think 30.7 is a bit on the large size, although safe to use for approximates right now. S16 lenses should yield 3K or at least 2.7K+ Definitely ample resolution to sample to 1080p/2K. For short-lived productions or timely documentary work this is a solution. Although in the past 5 years of shooting RED, moving to a lower resolution to increase space or recording time has never proven to be a valid decision for myself or anyone I've worked with. Too many compromises and wishing after the fact that it had not been done. The real solution is to bring more media or have a proper set-up that can offload and cycle media faster than the camera crew can fill it up. That can be difficult for owner/operators and productions on a shoestring budget, but it has proven to be the best course of action every time the choice has come up for us... Sometimes that choice didn't get made and there's always issues or regrets to follow. Always.



Generally speaking, F number and T scales do coincide by design. And most modern lenses marked in F numbers are using "calculated" F stop markings. Which is the same thing as T-Stops -- meaning you set the aperture to T2.8 (or F2.8 as the case may be) and the aperture actually becomes something larger so that you get the amount of light equivalent to 100% transmission at aF2.8.

 

[Mathew Joki]

Thanks for that clarification on f vs T numbers. I guess my misconception of there being T0.95 glass out there comes from misleading branding that I've seen recently with the "SLR Magic T0.95" lenses. Though I understand the basic conceptual differences be T/f numbers, I thought that this worked due to some sort of magic or something. haha

 

[Ryan Patrick O'Hara]

Actually we don't know the official specs and dimensions just yet. We know that the photosite or pixel size is approximately 5 microns and that we'll be able to shoot 5K within the S35 aperture. So far all the dimensions and resolution numbers I've seen posted have been labeled incorrect by Jim and Jarred. Anyway, we should be close, I think 30.7 is a bit on the large size, although safe to use for approximates right now.

That would be terrific if Dragon 6k wasn't all that much larger from our current Epic 5k.

S16 lenses should yield 3K or at least 2.7K+ Definitely ample resolution to sample to 1080p/2K. For short-lived productions or timely documentary work this is a solution. Although in the past 5 years of shooting RED, moving to a lower resolution to increase space or recording time has never proven to be a valid decision for myself or anyone I've worked with. Too many compromises and wishing after the fact that it had not been done. The real solution is to bring more media or have a proper set-up that can offload and cycle media faster than the camera crew can fill it up..

Might be the best quote I've read today. Totally agree. Even if finishing at 2k or 1080, I find shooting 4k (or 5k) and scaling down almost has infinite advantages than shooting at 2k. That said, I have some wonderful Cooke and Angie 16mm zooms that I've been fighting hard to rationalize using... and I've thought about the same thing.

On a thread-related note here. Keep F and T numbers straight. Technically a T1.0 refers to 100% light transmission, whereas an F1.0 aperture means that you have a 1:1 ratio between the aperture diameter (or equivalent, as it assumes a circular iris) and focal length measurements. So you can have an F number smaller than 1.0, but not a T number as that would imply your light being amplified within the optical path of the lens. T1.0 is pretty much in defiance of the physical laws of our universe anyway. Even the air we breathe doesn't allow for 100% light transmission, let alone chunks of ground and polished glass stacked together.

Again, great points. A perpendicular glass-to-air surface reflection is 5% light loss, greater at severe incoming angles. With special lens coatings, like the ones Cooke uses on their cinema lenses, that is reduced to .1%. But if a lens has 12 elements, that is still 2.4% light loss at technology's best.

One would have to custom make and design lenses that were for larger formats and had focal reducers/condenser lenses to increase light. It's really not easy, as these optical devices eat up light themselves and must be inserted within the typical lens assembly. That is from my experiments using them, and I've murdered many nikon primes in the pursuit of that knowledge.

 

[Roger Valdes]

Thanks for that clarification on f vs T numbers. I guess my misconception of there being T0.95 glass out there comes from misleading branding that I've seen recently with the "SLR Magic T0.95" lenses. Though I understand the basic conceptual differences be T/f numbers, I thought that this worked due to some sort of magic or something. haha

That lens is actually f0.92. The T0.95 number is correct. ;-)

If you scroll down on the page linked below you can find a more complete list of ultra-speed lenses
http://en.wikipedia.org/wiki/Lens_speed