Jump to content


Photo

What goes into making a fast lens?

fast

  • Please log in to reply
12 replies to this topic

#1 Phillip Mosness

Phillip Mosness
  • Basic Members
  • PipPip
  • 85 posts
  • Other

Posted 24 March 2014 - 02:49 AM

I've had a little trouble finding info on what is required to manufacture faster lenses.

Is it quality of glass? Less elements? More? Diameter of the elements?

Curious because we've all read about the F 0.7 Zeiss lenses from Barry Lyndon , but what did it take to make them?

It seems like most get to the 1.4 range but not much past that, so what's the hurdle?


  • 0

#2 Brian Drysdale

Brian Drysdale
  • Basic Members
  • PipPipPipPip
  • 5070 posts
  • Cinematographer

Posted 24 March 2014 - 03:02 AM

I suspect it's the cost and a limited market, the F0.7 lenses were made for NASA and also with increased sensitivity there's less need for lens speed. There are a number of f0.95 lenses around.


  • 0

#3 Zac Fettig

Zac Fettig
  • Basic Members
  • PipPipPipPip
  • 338 posts
  • Other
  • Boston

Posted 24 March 2014 - 07:03 AM

Mostly physical limitations.

 

The f number is a ratio of the focal length of the lens to the size of the aperture. For example, to make a 100mm f1.0 lens, you need a 100 mm aperture. So the back of the lens has to be more that 4" across, to accommodate it.

 

To make that happen, you need extremely good glass, and a flange mount free of obstructions. And extremely fine adjustments in the manufacturing process.

 

I want to say that the Zeiss f0.7 lenses were originally meant for a Hasselblad, then adapted to Mitchell BNC.
 


  • 0

#4 Adrian Sierkowski

Adrian Sierkowski
  • Sustaining Members
  • 7118 posts
  • Cinematographer
  • Los Angeles, Ca

Posted 24 March 2014 - 10:39 AM

Yep, the .7s were certainly an adaptation from a stills lens, and I often wonder what "T" they were. Sure they were an F.7, but I'd not be surprised if they were more like a T1.2

I should've checked them out over @ LACMA, but I didn't really want to be around people in a museum. I strongly dislike people at a museum.


  • 0

#5 Brian Drysdale

Brian Drysdale
  • Basic Members
  • PipPipPipPip
  • 5070 posts
  • Cinematographer

Posted 24 March 2014 - 10:58 AM

Here's an even faster lens although it seems to be more a PR f0.33 than a functional one. http://petapixel.com...lens-ever-made/


  • 0

#6 Adrian Sierkowski

Adrian Sierkowski
  • Sustaining Members
  • 7118 posts
  • Cinematographer
  • Los Angeles, Ca

Posted 24 March 2014 - 11:00 AM

That's a fake lens, "Made"as a joke. If you look it up, it never worked.


  • 0

#7 Brian Drysdale

Brian Drysdale
  • Basic Members
  • PipPipPipPip
  • 5070 posts
  • Cinematographer

Posted 24 March 2014 - 11:03 AM

As I said a PR lens.


  • 0

#8 Adrian Sierkowski

Adrian Sierkowski
  • Sustaining Members
  • 7118 posts
  • Cinematographer
  • Los Angeles, Ca

Posted 24 March 2014 - 12:00 PM

Ah so you did Brian; for some reason my brain didn't process any of that. Apologies.


  • 0

#9 Dom Jaeger

Dom Jaeger
  • Basic Members
  • PipPipPipPip
  • 1604 posts
  • Other
  • Melbourne, Australia

Posted 24 March 2014 - 08:03 PM

The f number is a ratio of the focal length of the lens to the size of the aperture. For example, to make a 100mm f1.0 lens, you need a 100 mm aperture. So the back of the lens has to be more that 4" across, to accommodate it.

 

That's not quite correct Zac, the f number is the ratio of the focal length to the diameter of the entrance pupil, which is not the same as the actual aperture size. It's basically the size of the aperture as seen through the front of the lens. A lot of lenses have front groups that act like magnifying glasses, so the aperture looks much larger than it is. For example, I'm currently working on a 180mm Cooke S4 which is T2 (lets say f/2 for simplicity). So the entrance pupil is around 90mm diameter (180/2), but the actual iris aperture itself is only about 40mm across. It just looks a lot bigger through the front. 

 

This is how a lot of zooms can maintain a constant f stop as the focal length changes - the magnification of the aperture changes so that the focal length to entrance pupil ratio remains the same.


  • 1

#10 Dom Jaeger

Dom Jaeger
  • Basic Members
  • PipPipPipPip
  • 1604 posts
  • Other
  • Melbourne, Australia

Posted 24 March 2014 - 08:21 PM

I've had a little trouble finding info on what is required to manufacture faster lenses.

Is it quality of glass? Less elements? More? Diameter of the elements?

Curious because we've all read about the F 0.7 Zeiss lenses from Barry Lyndon , but what did it take to make them?

It seems like most get to the 1.4 range but not much past that, so what's the hurdle?

 

A few things about fast apertures that might be relevant:

 

The larger the relative aperture the more aberrations that are introduced, so very fast lenses need more correction to produce an acceptable image. They had f/1.5 lenses back in the 30s but they weren't particularly great wide open. Back then more elements meant more light loss and less contrast due to internal reflections so they kept the number of elements to a minimum. These days multicoating, modern glasses and aspheric elements means better correction can be done without compromising speed, but it requires expensive and precise manufacturing. Zeiss managed to make some pretty compact f/1.2 lenses a few decades back in the form of Super Speeds, the modern versions are Master Primes which aren't much faster but are a lot bigger - the difference is mainly in better correction of aberrations wide open (plus better mechanics).

 

The main limiting factors in lens speed tend to be the angle of view and the size of the format that needs to be covered. So you'll find plenty of fast medium focal lengths for a particular format but not so many very fast wide angles or telephotos, and the larger the format the slower (and bigger) the lenses tend to be. Using Super Speeds as an example again, they are generally compact except for the 18mm wide angle which needed a lot more glass to achieve an acceptable image at f/1.2 and the series ends with the 85mm, whose front element only just fits in the barrel.

 

Very fast lenses are often made for particular tasks, such as radiology, aerial/space photography or reproduction, but for general photography and particularly cinematography the problems of a very very shallow depth of field make them almost unusable. I don't think too many people actually want lenses faster than Master Primes, certainly not focus pullers!


  • 1

#11 Dom Jaeger

Dom Jaeger
  • Basic Members
  • PipPipPipPip
  • 1604 posts
  • Other
  • Melbourne, Australia

Posted 24 March 2014 - 08:54 PM

The f number is a ratio of the focal length of the lens to the size of the aperture. For example, to make a 100mm f1.0 lens, you need a 100 mm aperture. So the back of the lens has to be more that 4" across, to accommodate it.

 

Sorry, I should add that Zac is right in terms of fast lenses and long focal lengths needing to be large, only it's the front element that needs to accommodate the entrance pupil. In the above example a 100mm f/1.0 lens needs at least a 100mm front element.


  • 0

#12 dan kessler

dan kessler
  • Basic Members
  • PipPipPip
  • 172 posts
  • Other

Posted 24 March 2014 - 10:38 PM

Dom said the magic words - lens aberrations.   Historically, these have been the chief obstacle to speed.

Visualize the cone of light passing through a lens with spherical surfaces. (most lenses)

It is an optical fact of life that the rays at the outer edges of the cone do not focus at the same point as

the rays nearer the center.  This gives rise to a number of image defects that lens designers have had

to battle for more than a century.  The larger the light cone, the harder it is to get all the rays of all the

colors to come to nice sharp points in a flat image plane.

This is why photographers know, or should know, that stopping down improves the performance of
any lens.  We are narrowing the cone of light, eliminating the rays that contribute the most to
aberrations.


  • 0

#13 Gregory Irwin

Gregory Irwin
  • Sustaining Members
  • 589 posts
  • 1st Assistant Camera
  • Work is based in Los Angeles but I live elsewhere.

Posted 24 March 2014 - 11:33 PM

Dom, I am so impressed with your knowledge. Bravo!

G
  • 0



Also tagged with one or more of these keywords: fast

FJS International, LLC

Aerial Filmworks

Abel Cine

Willys Widgets

Wooden Camera

Media Blackout - Custom Cables and AKS

Tai Audio

Opal

CineLab

Rig Wheels Passport

Gamma Ray Digital Inc

The Slider

Glidecam

Broadcast Solutions Inc

Paralinx LLC

Visual Products

CineTape

Technodolly

Ritter Battery

Metropolis Post

rebotnix Technologies

rebotnix Technologies

FJS International, LLC

Technodolly

Broadcast Solutions Inc

Tai Audio

Abel Cine

The Slider

Wooden Camera

Willys Widgets

Rig Wheels Passport

Visual Products

Paralinx LLC

Aerial Filmworks

Gamma Ray Digital Inc

Opal

Metropolis Post

CineTape

Media Blackout - Custom Cables and AKS

Glidecam

CineLab

Ritter Battery