f/0.7: A Look at Lenses With Some of the Most Extreme Apertures Out There

The photography world certainly loves its wide-aperture lenses, with most full frame consumer lenses topping out at f/1.2 and a few extreme lenses pushing the limits at f/0.95. However, there is a world of lenses that go far beyond f/0.95, and this interesting video takes a look at that extreme glass.

Coming to you from Media Division, this fun and informative video takes a look at extreme aperture lenses. In it, they modify a Zeiss Biotar 100mm f/0.73 lens to shoot with. If you are wondering just how extreme f/0.73 is, it is about 1.88 stops faster than an f/1.4 lens. Even compared to some of the extreme f/0.95 lenses available today, it is about 0.76 stops faster. The 100mm f/0.73 is absolutely massive, weighing almost 18 lbs (8,100 g)! Generally, such lenses have a fixed aperture and fixed focus and were made for x-ray machines. There are other examples, however; for example, Stanley Kubrick used three Zeiss Planar 50mm f/0.7 lenses to film "Barry Lyndon" in candlelight, the same lenses NASA used to film the far side of the moon. If you are interested in playing with such extreme optics, you can generally find used x-ray machine lenses on eBay. Check out the video above to learn more about this fascinating world. 

Log in or register to post comments
Matt Williams's picture

On the one hand, I think ultra shallow DOF is a silly pursuit (though that obviously wasn't the objective of some of these, e.g. Kubrick).

On the other hand, the optics/mechanics nerd in me loves this crazy stuff, especially the modifications. And these rare lens designs are really cool no matter how you spin it to me.

Some day I'll finish my modification of a Zeiss 35mm Tessar from a dead Yashica T4. I've done a number of other mods - like a Minox 35 lens to M-mount - but all of those are comparatively simple.

Jonathan Brewer's picture

".....I think ultra shallow DOF is a silly pursuit....."

I've actually based/invested more that 20 yrs. work into that kind of photography. Using/modifying lenses to get the extremely shallow DOF which is a bank shot introduction into the transitions from representational parts of an image/vision to the abstract/surreal all in the same frame.

Better than verbalizing this, I invite anyone to peruse some of my images like this one (the bottom image) on my discussion venue "wideopen" where I intentionally went from representational to a complete obliteration of the background via a Taylor-Hobson 125mm F1.5 projections lens where I'm using only the front lens group/front cell, on a Speed Graphic(to use it's rolling shutter).

This lens sells for a fortune now, when I bought it in the late nineties I bought it for 25 bucks.

I was able to have an adaptor machined to mate the threaded casing of the front cell to a Speed Graphic lens board.


Hopefully this makes sense; my thinking in executing this image was "more than a photograph, more than a painting", because it had the feeling of both, or at least to me, something beyond the tools I was using.

I studied to be a painter, but ultimately, I've gotten tired of, making photographs that look like photographs, and paintings that look like paintings.

This site, and my Fine Art site......


were both created in the late nineties when this kind of photography was dismissed as worthless/trash/terrible/a waste of time, but there was a boatload of good things that became available from the tool box of this type of photography.

Alexander Petrenko's picture

I don’t understand how T0.95 is possible.

Matt Williams's picture


Alexander Petrenko's picture

How do you bring to the sensor more light than enters the lens?

Matt Williams's picture

You don't.

Whatever your f-stop, your theoretical maximum t-stop is the same, though that never bears out in reality because no lens has 100% transmission.

T-stop is simply the f-stop divided by the square root of light transmittance.

So, if your F-stop is f/0.95 and you transmit 95% of light: 0.95/sq root of (0.95) = 0.975 = your t-stop

Any lens with a T=0.95 would have a slightly larger F-stop (the exact number depending on your light transmittance).

Matt Williams's picture

Probably shouldn't have used 95% and f/0.95 at the same time if I didn't want to make that confusing. If your transmittance was 90% and f/0.95 then it would be 0.95/square root of 0.90 = T/1.0

Philipp Schmid's picture

And f-number (or t-number) below 1 just means that the aperture diameter is larger than the focal length. F/0.0 would be an infinetly large aperture. A lens that produces more light than it receives has a t-stop lower than its f-stop but can still have a t-stop above 1

Matt Williams's picture

(and the latter of which does not and cannot exist)

Robert Edwardes's picture

it's f0.95 and 0.7 they never say the t-stop which is always higher then since f-stop which is a simple equation that doesn't take into account light absorption which t-stops does

Tom Reichner's picture

I am not sure what it is about this that you don't understand. Perhaps you don't know what the f stop really is, or how it is calculated. Is that possible?

The f-stop is the ratio between the focal length and the diameter of the aperture. If the focal length of a lens is 50mm and the aperture is 50mm, then the f-stop is 1:1.0

If the focal length is 50mm and the aperture is 55mm, then the f-stop is 1:0.90

If the focal length is 50mm and the aperture is 100mm, then the f-stop is 1:2.0

Does that help you to understand how an f-stop of 1:0.95 is possible?

This also explains why f-stop is always expressed as a ratio (when it is properly or formally expressed). To be technical and nit-picky, we should not say that an f-stop is f2.8, or f4.0. We should say that an f-stop is 1:2.8, or f1:4.0. That is why it is written that way on your lenses. Not just a number, but a ratio.

You must not try to think in terms of how much light is being transmitted - that is NOT a factor in the way f-stops are calculated. Light transmission is merely a by-product of the f-stop; a result of it, not an essential factor or component of the f-stop quantification.

Alexander Petrenko's picture

Thanks, it’s a long text, but not about T-stop, I suppose.

Nikolas Moldenhauer's picture

Let me try... T1 doesn't mean 100% transmission. T1 on a f1 would be 100% or T1.4 on a f1.4. The T-Stop is the f-stop corrected by the light loss by reflections and other imperfections of the lens. No lens can practically have 100% transmission. The R-Biotar in this test will have something like T0.85

Ben Harris's picture

And off to eBay I go!

Nikolas Moldenhauer's picture

I'm sorry ;-)

Spy Black's picture

I'm surprised they didn't use the original Noct Nikkor instead of the Canon. The Canon was about as good similar lenses like the Minolta, but the Nikkor was in a league of it's own.

Ben Harris's picture

Agreed, it was an odd omission. The Noct is something unique.

Nikolas Moldenhauer's picture

the simple answer... the Canon FD is over 10x cheaper. It would be "odd" to pitch a 2–3K lens as the "budget" option... what the last lens entry is all about.

Spy Black's picture

The standard f1.2 Nikkor was also better than the old Canon and Minolta 1.2s.

In the end he pointed out that it was essentially futile.

Nikolas Moldenhauer's picture

If you say so....... but still not "budget"

Spy Black's picture

Of course it is. Used Nikkors are going for pretty much the same price as the FD on eBay, and you don't have to worry about adding an optical interface that degrades the image quality further.

guy bellingham's picture

f/0.7! I designed and built a video camera that shoots with an effective DoF of less than half this. ... https://youtu.be/d1H6fHCodKM

Nikolas Moldenhauer's picture

Our test was not about getting the least DoF, you can built that quite easily by filming off the ground glass of a large format camera or with tit shift. It is about fast aperture lenses and the physical limits. Faster than f0.5 is physically impossible, faster ten f0.7 is practically pretty much impossible.