Do (all) cinema lenses get slower as you focus closer?

[Matt W.]

Was just discussing this today, curious to hear what more knowledgeable people have to say about the subject and (more importantly?) what you do about it.

As we know, focusing closer with most still lenses slows them down a bit because they breathe as they focus closer (focal length is essentially the distance at which a given lens focuses at infinity, after all). So this is the genesis of breathing (as you focus closer you also zoom in) and the darkening effect from which most macro lenses seem to suffer from dramatically because it's really pronounced on a macro lens (if the focal length changes and the aperture remains the same, the f stop also changes). From my days (okay, minutes) shooting macro and large format, I remember there's some formula for exposure compensation that you have to use and it's actually pretty significant, a matter of two stops or so by life-sized, I think?

A lot of older lenses don't have good close focus (Zeiss superspeeds in particular) and I speculated that this was so they would stay smaller and a larger assembly didn't have to be built around the movement of the glass, but a friend speculated that it was to limit the necessary correction of any floating elements that would account for both breathing and stopping down. But the close focus is particularly bad on longer lenses and I know that still lenses from that area (and many modern lenses that aren't retrofocus) lack floating elements...so I'm not sure.

So: do modern lenses account for breathing and this darkening effect through the same "floating element" assemblies that both zooms out a bit and thus changes the effective f-stop so that it's constant? Do older cinema lenses? And at what point were these changes made?

Does anyone compensate for this when metering for lenses that aren't corrected as such? Open up a bit for the CU? Have you seen your exposure change when pulling focus?

Do most diopters have filter factors?

Thanks, just curious because I am bored at work.

 

[Phil Holland]

It largely depends on lens design and how far the elements travel within the tube. As you said it's more noticeable with macro lenses losing around 3/4s of a stop or sometimes more and close focus. It's essentially like using extension tubes where that additional dead space creates a loss of light, but allows you to focus much closer.

On most lenses/focal lengths the exposure variation is extremely slight and sometimes non existent. Where you would really see it is pulling focus on say a 100mm or up macro lens from close to infinity. In that situation there would likely be a reason to pull the iris as well. That's a pretty extreme and rare situation to be in.

Usually there are specifications on diopters that tell you how much light is being lost. It's usually no more than a quarter of a stop with modern optics and that's something you can certainly meter and calculate. Same deal for wide angle adapters, teleconverters, and anamorphic adapters. More glass or more tube equals more light loss.

 

[Jacek Zakowicz]

It's all about the magnification

It's all about the magnification

Actually.....
It has to do with the magnification.
2:1 magnification causes 1 stop loss and 1:1 is 2 stops. Pretty simple.....:wink5:

 

[Phil Holland]

well yes it magnifies during that

 

[Matt W.]

That makes sense re: diopters having a set filter factor to compensate for whatever range they allow focus in. Thanks! Never worked with a high end one.

Actually.....
It has to do with the magnification.
2:1 magnification causes 1 stop loss and 1:1 is 2 stops. Pretty simple.....:wink5:

Right, I remember now.... A 150mm lens was 1:1 at full bellows extension (300mm) with two stops of light loss.

So, with a modern lens like a master prime, does it have some sort of optical assembly that allows it to zoom out a bit as you focus closer, both correcting for breathing and light loss, or are they separate issues? At what point was this kind of thing introduced into cinema lenses?

And why do the Zeiss superspeeds have such bad close focus? And do they have floating elements in the longer focal lengths or are they optically similar to contemporaneous still lenses (which, if I remember correctly, only have floating elements to correct for aberrations in retrofocus designs e.g. 28mm and wider?). A friend and I were arguing over whether or not a cinema lens had to have the same t-stop over its focus range; I said I didn't think it did, he said it would have to, but it could just be a case of the difference being trivial.

 

[Jacek Zakowicz]

That makes sense re: diopters having a set filter factor to compensate for whatever range they allow focus in. Thanks! Never worked with a high end one.
These are very good questions- congrats



Right, I remember now.... A 150mm lens was 1:1 at full bellows extension (300mm) with two stops of light loss.

So, with a modern lens like a master prime, does it have some sort of optical assembly that allows it to zoom out a bit as you focus closer, both correcting for breathing and light loss, or are they separate issues? At what point was this kind of thing introduced into cinema lenses?

And why do the Zeiss superspeeds have such bad close focus? And do they have floating elements in the longer focal lengths or are they optically similar to contemporaneous still lenses (which, if I remember correctly, only have floating elements to correct for aberrations in retrofocus designs e.g. 28mm and wider?). A friend and I were arguing over whether or not a cinema lens had to have the same t-stop over its focus range; I said I didn't think it did, he said it would have to, but it could just be a case of the difference being trivial.

These are all very good questions- congrats:thumbsup: The short answer is that these are all separate issues:
Yes the MPs optical design (most likely) has a zooming element to compensate for breathing- that's how breathing is handled. You pay for it dearly with the complexity- both optical and mechanical. And , consequently -the purchase price.

The internal focus design deals with another- more practical issue: it keeps the front of the lens stationary- not rotating or telescoping. Makes fitting a matte box much easier.

SS are not focusing close out of designer's choice- with helical focus control design you have to choose between close focus ability and legible proper focus marks. Especially with small lens barrels. They chosen the legible focus marks spread properly. For close focusing they offered whole set of macro lenses- I believe 35, 50, 60, 100- maybe more.
The opposite approach would be the RPPs- they chosen close focus and, consequently, had to sacrifice some focusing marks especially further out since they got very close together. Even with the much bigger RPP barrels.


There is another solution that allows for more flexibility- cam focusing design. Cooke and Leica Cine primes use this approach. By changing the slope of the cam Cooke was able to spread the further focus marks and maintain very respectable close focus ability.

Leica Summilux C took that approach even further and calculated the focus slope a it relates to focus ditance for each FL hence the "magical" identical focus scales down to 6 feet on (pretty much ) all lenses in the set. Pretty brilliant....:thumbup:

All that mechanical ingenuity pales in comparison to the flexibility of the electronic control.
With proper Electronic lens controllers coming to the market very soon- we are very close to the release now- your head will spin with all the (lens control) options possible at your finger tips . For small fraction of the cine lens price, fraction of the cost of FIZ..:coolgleamA:
And, with all these goodies I believe that even the "cine lens die hards" soon will forget about the only ONE diadvantage of using the electronic (still) lenses- that you can't zoom during the shot. It's not is style these days anyway:wink5:

All that does not change the physics: 1:1 magnification causes 2 stops loss on any lens...

 

[Peter Dmitriyev]

thanks for the informative thread. Did not know about the exposure variation at focal lengths..

 

[Jacek Zakowicz]

thanks for the informative thread. Did not know about the exposure variation at focal lengths..

OK I knew that was coming:
Magnification is NOT a zoom ratio- it has not much to do with zoom.
The first number is simply how big the object photographed is - in front of the lens. The second number is the size of the image of this object behind the lens.
1:1 simply means that the size of the actual object is equal to the size of its image generated by the lens.
Think closeups of bugs. Macro photography. In order to do that you have to get relatively close to it(the subject) and, consequently, much further from the image plane- hence the light loss...
Hope this helps....
Oh , and one more thing to be clear:
1:1 magnification will cause 2 stop light loss regardless of the FL - 100mm,200mm 300mm makes no difference

 

[Stephen Williams]

Diopters dont have a filter factor, other than the thickness of the glass.

What Jacek says about 1:1 magnification is basically correct, however they is a way around the proplem, which if Jacek scratches his head for long enough will figure out!

 

[Rob Ruffo]

These are all very good questions- congrats:thumbsup: The short answer is that these are all separate issues:
Yes the MPs optical design (most likely) has a zooming element to compensate for breathing- that's how breathing is handled. You pay for it dearly with the complexity- both optical and mechanical. And , consequently -the purchase price.

The internal focus design deals with another- more practical issue: it keeps the front of the lens stationary- not rotating or telescoping. Makes fitting a matte box much easier.

SS are not focusing close out of designer's choice- with helical focus control design you have to choose between close focus ability and legible proper focus marks. Especially with small lens barrels. They chosen the legible focus marks spread properly. For close focusing they offered whole set of macro lenses- I believe 35, 50, 60, 100- maybe more.
The opposite approach would be the RPPs- they chosen close focus and, consequently, had to sacrifice some focusing marks especially further out since they got very close together. Even with the much bigger RPP barrels.


There is another solution that allows for more flexibility- cam focusing design. Cooke and Leica Cine primes use this approach. By changing the slope of the cam Cooke was able to spread the further focus marks and maintain very respectable close focus ability.

Leica Summilux C took that approach even further and calculated the focus slope a it relates to focus ditance for each FL hence the "magical" identical focus scales down to 6 feet on (pretty much ) all lenses in the set. Pretty brilliant....:thumbup:

All that mechanical ingenuity pales in comparison to the flexibility of the electronic control.
With proper Electronic lens controllers coming to the market very soon- we are very close to the release now- your head will spin with all the (lens control) options possible at your finger tips . For small fraction of the cine lens price, fraction of the cost of FIZ..:coolgleamA:
And, with all these goodies I believe that even the "cine lens die hards" soon will forget about the only ONE diadvantage of using the electronic (still) lenses- that you can't zoom during the shot. It's not is style these days anyway:wink5:

All that does not change the physics: 1:1 magnification causes 2 stops loss on any lens...

I have seen Jacek's new system. All he says is true. You want one. I want one. No doubt at all.

 

[Matt W.]

Interesting....what exactly is the concept behind lenses that don't breath (do they simply zoom out a bit as you focus closer, or is it much more complicated?), btw, and when was that introduced? And how does cam focusing work versus a helical?

Electronic focus sounds exciting; the USB systems for the 5D are already pretty cool for no-budget work.

 

[Rob Ruffo]

Interesting....what exactly is the concept behind lenses that don't breath (do they simply zoom out a bit as you focus closer, or is it much more complicated?), btw, and when was that introduced? And how does cam focusing work versus a helical?

Electronic focus sounds exciting; the USB systems for the 5D are already pretty cool for no-budget work.

The 5D system is worse than focussing by hand as it steps like crazy. Jacek's system is a whole other thing. It's awesome.

 

[Jacek Zakowicz]

Thank you Rob, I'm glad you liked it:thumbsup:
Matt helical is based on threaded rings so that, by definition, is a fixed pitch.
Cam system is a slot machined in a ring and a key that fits in it and guides the glass housing. By varying the slope of the slot you can adjust pitch (travel rate) at will.
The cam system is a necessity for zoom lenses as the optical design determines the movement of the glass.
It was adapted for focus control because of the flexibility and longevity it offers....