Zoom mm and f range

[Jeremy87]

Hi

I've noticed that zoom lenses have a range of focal length and f-number. So far so good.
However, keeping the aperture size constant, shouldn't the f-number decrease as much as the focal length increases?

Say the focal length is 18-200mm and the f-number is f/3.5-5.6.
At 18mm and f/3.5, the aperture is 5.1mm.
At 200mm and f/5.6, the aperture is 36mm.
Why the change? Why not upen up to 36mm even at a wide angle?

Is this because the aperture stop isn't actually at the first lens, but deeper inside the camera and moves less (or sometimes not at all) with the zoom? In that case, why build it in such a wasteful way?

[jcuknz]

This is something that the still photographer has had to put up with since time immoral while it would be completely unacceptable to the cinematographer. You can get "constant' lenses for still cameras which maintain their basic aperture through the zoom range I have a constant f/4 90-230 lens.... but they tend to be expensive and bigger and today there seems to be a pre-occupation with compactness .... nice to have if you can tolerate the loss of aperture at the long end of the zoom.

[Jeremy87]

That doesn't really answer the question. I know they exist, but why? Constant f-number is ideal, sure, but the physical size of the lens system is unfortunately constant, so why not give the user the full range of f-numbers available?
Sure, you don't want it to change during a zoom when filming, but you might want to have an even bigger aperture when zoomed out IF you know you won't be zooming in.

For example your constant f/4 90-230mm lens. At 230mm the aperture is 57.5mm in diameter. What is preventing you from using that 57.5mm aperture at 90mm focal length (f/1.6) other than the manufacturer limiting your choices?

[rpcrowe]

Jeremy, it is very simple. It costs more to manufacture a lens which has a constant f/stop across the focal range than it does to manufacture a lens that has a variable f/stop.

That is because, the lens that maintains a constant f/stop across the focal length, actually has a variable aperture which gets larger as the focal length is increased and smaller as the focal length is decreased.

The "variable" aperture lens actually has a constant aperture, one which doesn't vary as the focal length is increased and decreased.

Since the f/stop is determined by the relationship between the size of the aperture and the focal length, a longer focal length needs a larger aperture to maintain the same f/stop.... and a shorter focal length needs a smaller aperture to maintain that same f/stop...

Obviously, the engineering required by adjusting the aperture to get larger and smaller as the focal length is changed thus maintaining a constant f/stop, not to add in the optical characteristice needed, is more complicated and expensive than engineering a lens in which the size of aperture ramains constant throughout the zoom range resulting in a variable f/stop..

Why would I pay more for a lens which has a constant f/stop than for a lens with a variable f/stop?

1. My 70-200mm f/4L IS lens has a constant f/stop throughout its zoom range. Than means that when I shoot at 200mm, my maximum f/stop is still f/4 rather than, say f/5.6. I need the largest f/stop possible for achieving a shutter speed fast enough to hand hold the camera/lens.

I am going to answer the question I think you will ask... Why not just buy a lens that has an f/2.8 f/stop available? I selected the 70-200mm f/4L IS lens over the 70-200mm f/2.8L (series) because the f4 L IS lens is considerably lighter in weight and more compact than any of the f/2.8 (series) lenses. That is an important attribute since I travel with this lens and like to carry it everywhere. The fact that it is a lot less expensive than the f/2.8L IS ii lens is simply icing on the cake for me...

2. As mentioned, when shooting video, you need a constant aperture to zoom while you are shooting.

[Jeremy87]

OK let's try this one more time...

You have a 70-200mm f/4 lens. So when you shoot at 200mm, you have f/4, all is good.
Now you zoom out to 70mm. Why would you want to limit yourself to f/4, when a 200/4=50mm aperture already exists? Your f/4 lens could be f/1.4 when at 70mm, and I see no reason why it couldn't be because the hardware is already there. What am I missing?

You ask why I would want to pay more for a fixed f-number. I don't want to pay more for it. I'd pay more for a variable f-number, because then it would make more efficient use of what aperture size it has for all zooms.

[herbert]

Hi Jeremy,

The aperture is not actually the size you think it is. The physical size of the smallest hole through which the light passes does not change. What changes is the effective size. The effective size of the hole (the entrance pupil) is the physical size multiplied by the magnification of all the glass in front of it.

Try this. Look into your lens with the zoom at its widest. Then zoom the lens and you will see the hole appear to get bigger. Nothing is actually changing shape inside the lens. The change is because the lens is magnifying the size of the hole. Unfortunately on cheaper lenses it does not magnify it enough. So the effective aperture reduces in proportion to the focal length.

Fixed aperture zoom lenses are a lot more complex (more glass in front of the aperture) allowing the aperture to be magnified the same at all focal lengths. So the effective aperture is a constant ratio.

Hope this helps.

Alex

[Brocken]

The scientific name for the f number is focal ratio.

It is the ratio of a lens's focal length to the diameter of the aperture or entrance pupil.

If we had an aperture of fixed size - a simple hole drilled in a metal plate - as was common many years ago, then you will see that if we changed the focal length, while the physical size of the aperture remained the same, then the ratio between them would change like this:

Imagine a lens which has a focal length of 50mm, and a fixed aperture which is 10mm across.

The focal ratio is the ratio of 50mm to 10mm which we could write as:

f number = f/D

where f is the focal length, and D is diameter of the aperture. So 50mm divided by 10mm is 5, and that is the f number of the lens.

Now, if we increase the focal length to 100mm, but still have our fixed diameter aperture of 10mm, then the f number will be 100mm divided by 10mm which is 10.

If we zoom to an amazing focal length of 1000mm, but still have our fixed 10mm hole, then the f number will rocket to 100.

Suppose we wanted to keep the same focal ratio of f/5 as we zoomed from 50mm to 100mm. 100 divided by 5 gives us an aperture diameter of 20mm. And if we did our mega zoom to 1000mm, 1000 divided by 5 is 200. Yes, a 1000mm f/5 lens would need an aperture 200mm across and you would need a wheelbarrow or perhaps even elephants to cart it around with you.

[Mike Buckley]

Thank you to Jeremy -- and welcome to CiC -- for asking the question and thank you to everyone else for providing such informative responses. I had never thought of asking the fundamental question that Jeremy posed and I also had no idea about much of the information provided in the various responses.

[IanCD]

I've wondered about this before... vaguely!
Alex and Christopher's answers don't seem (to me) to be saying the same thing... Christopher's response being consistent with Jeremy's original thinking, which was previously mine also... I'd not heard of magnification by lens elements being taken into account in determining aperture before this, so thanks to Alex for sending me off to look into this in more detail.
I wonder if this article might help? Wikipedia may not be a definitive source but I thought their explanation might be the most straightforward.
Ian

[darkslide]

This is both a fascinating subject, now that it's come up, and a fascinating article (thanx IanCD) although I have to admit to being stumped on the first line - I'm learning what "disambiguation" means at the moment - I'll get to the rest of the article later

Many thanks for the bright ones amongst us/you for chipping in - keep it up chaps

[herbert]

I've wondered about this before... vaguely!
Alex and Christopher's answers don't seem (to me) to be saying the same thing... Christopher's response being consistent with Jeremy's original thinking, which was previously mine also... I'd not heard of magnification by lens elements being taken into account in determining aperture before this, so thanks to Alex for sending me off to look into this in more detail.
I wonder if this article might help? Wikipedia may not be a definitive source but I thought their explanation might be the most straightforward.
Ian

The wiki article talks a little bit about the effective aperture. It does not define it. Wikipedia (or perhaps me) should update its links to this other article:

http://en.wikipedia.org/wiki/Entrance_pupil

Which states:

In an optical system, the entrance pupil is the optical image of the physical aperture stop, as 'seen' through the front of the lens system.

From the original aperture article:

The lens aperture is usually specified as an f-number, the ratio of focal length to effective aperture diameter

So the actual hole is not changing size when you alter the focal length (i.e. zoom the lens) but the hole appears to change size when you look into the lens from the front and zoom. And it is the effective aperture that is used to define the f-number.

The difference between a variable aperture and a fixed aperture zoom is that the magnification of the aperture is scaled in proportion to the focal length on the fixed aperture lens. This is may or may not be more complex to design and make but given that the lenses are usually heavier and bigger it must involve some extra glass in front of the aperture.

The entrance pupil article goes on to contain our complete explanation:

In photography, the size of the entrance pupil (rather than the size of the physical aperture itself) is used to calibrate the opening and closing of the diaphragm aperture. The f-number ("relative aperture"), N, is defined by N = f/EN, where f is the focal length and EN is the diameter of the entrance pupil. Increasing the focal length of a lens (i.e., zooming in) will usually cause the f-number to increase, and the entrance pupil location to move further back along the optical axis.

Camera manufacturers charge a premium for fixed aperture lenses (bigger, heavier, more complicated == expensive) but it does have benefits when setting exposure manually since you can zoom and not have to change settings.

Alex

[rpcrowe]

I need to mention that present technology precludes a constant f/2.8 f/stop in any lens with a focal range of over 3X.

That means that you can get a 70-200mm lens with a constant f/2.8 (200 divided by 70 = 2.86 X ratio).

However, you cannot get a constant f/2.8 f/stop in a wide range lens such as an 18-200mm (200 divided by 18 = 11.11 X ratio). These lenses have a variable aperture. You also cannot get a 24-105mm lens with a constant f/2.8 aperture (105mm divided by 24mm = 4.16 X ratio). The closest to this is the Canon 24-105mm f/4L IS lens.

This is not to say that new technology will not make constant f/spotps available in wide range zoom lenses. However, judging by recent entries into the lens field, indicate that this advance in technology may not be what camera companies are aiming at.

http://en.wikipedia.org/wiki/Aperture

[jcuknz]

Still photographers can change settings when they zoom, between shots ... the cinematographer cannot within the shot working with a fixed 'shutter speed' of 1/32 or 1/50 or 1/48 depending on which fps they are shooting at. I think I have read of exceptions to this but the details escape me. I think I had a variable shutter on my Bolex H16 way back. But one cannot increase the shutter speed for ordinary shooting as while the still photographer usually wants crisp shots the movie guy normally wants subject blur when there is subject movement to provide smoothness to the movement between frames.

[Manfred M]

Alex - I'm not sure that your definition of effective aperture is quite correct. It's not what you see while looking at the iris from the front of the camera, but rather the impact on the light falling on camera's sensor that is relevent. When a lens is focused at infinity, the actual and relative aperture are equal, but as the subject gets closer to the lens, the subject is magnified and relative aperture comes into play; and relative aperture is defined as N' = N*(m+1) where N' is the effective aperture, N is the actual aperture setting, m is the magnification).

A good real life example is a macro lens where this is quite noticable, so looking at two conditions:

1. When focused at infinity, m = 0, so the effective aperture is N' = N*(0+1); i.e. N' = N as I stated previously, the effective aperture and actual (geometric) aperture are the same.

2. If we focus our macro lens down to a 1:1 ratio, i.e. magnification is 1 then the equation reads N' = N*(1+1); or N' = 2N. In this example while the f-stop reading on our lens is, say f/11, the amount of light reaching the sensor is 2*11 or f/22. Notice that effective aperture is independent of focal length. A 50mm or 100mm lens would have exactly the same aperture at the 1:1 magnification.

Constant aperture is not all that important in most still photography, as we can usually increase exposure or ISO setting to compensate. Where it is important is in film and video work, where we don't want the scene to get lighter or darker as we follow focus on a subject.

[jcuknz]

Before I had TTL metering, what a giant step forward for photographers that was , one had to do calculations when going in close as to focus close the focal length of the lens changes, increases. Measure the extension and divide by aperture.
We are getting away from the original question which seems to be an acceptance of reality versus an ideal perhaps.

[Dave Humphries]

An interesting thread.

When I was in TV production, many years ago, part of lens line up was to set the 'ramp limit switch', something that stopped the camera operator zooming in fully if the aperture (controlled by the 'racks engineer' (that was me)) in use for the scene was such that it would get darker. If you had a bright outdoor scene, you could zoom in 20 x, if dim indoors, rather less. So they weren't constant aperture lenses, but under the right circumstances, could be made to behave as such.

The fact that 20:1 was the normal range for a TV zoom lens in those days shows how long ago it was (and they cost the price of a small house in the UK!). These days double, or four times, that range (e.g. 86:1), are in use daily where I work and there are even longer ones - but probably houses are (relatively) more expensive these days

I have no idea whether these days, with CCD/CMOS sensors in TV cameras, they can alter the iso to compensate for ramping, the 86:1 isn't much bigger than the old 20:1s, so it must also ramp I assume - perhaps I'll ask Gary, the guy in our "Len's workshop" tomorrow

[Manfred M]

I have no idea whether these days, with CCD/CMOS sensors in TV cameras, they can alter the iso to compensate for ramping

Dave - I do / have done a fair bit of shooting with semi-pro Panasonic and Sony video cameras. The first thing you learn is that while these cameras do have autofocus and auto iris; you use these only to set up your shot. Auto iris is highly reactive and it is very obvious when it clicks in. Let's just say that the resulting video looks very amateurish. Same comment goes for autofocus; we've all seen video where the camera locks in on a passing car, rather than the main subject.

Constant aperture, parfocal lenses were the mainstay in the days of regular video. The sensors in those cameras were quite small compared to what we see in current HD cameras, and these features were likely easier to design and manufacture. If you look at the pro HD productions, the current trend seems to be to shoot with primes. A set of good primes from Zeiss, Cooke, or Leica. Even the relatively "cheap" Red lenses seem to be in the same price range as a new car.

http://www.bandpro.com/products/lens...&product_id=62

http://www.zeiss.com/cine

http://www.cookeoptics.com/cooke.nsf/products/s4i.html

http://www.red.com/store/lenses

[Dave Humphries]

I'm talking about real lenses

http://www.usa.canon.com/cusa/profes...igisuper_86_xs

Not them dinky little things LOL

(I don't mean to be rude btw)

[herbert]

Alex - I'm not sure that your definition of effective aperture is quite correct.

Hi Manfred, I think we are both correct.

My discussion of effective entrance pupil is for a constant subject distance and an altered focal length within a zoom lens. The effective entrance pupil is the size the aperture appears to be when projected through the lens in front of it. In essence if you have a very powerful lens in front of a hole the hole will look bigger as you move the lens away. In reverse you can think of a powerful lens as being able to capture light from a larger diameter circle and focus it through the hole than a weak lens.

The upper limit on the effective aperture is not the physical size of the lens in front of the hole, it just needs to be wide enough to capture all the light that can be focused. The limit to this is governed by refraction of an air-glass surface. I.e. how much you can bend light. I think the Wikipedia aperture article puts this limit for lenses to f0.5. With exotic materials like diamond it could be lower. The article states the lowest f-number lens used in a film was a NASA built f0.7 lens used by Stanley Kubrick. It must have had an interesting look.

Your discussion of effective aperture is for an altered subject distance and so an altered focus distance. It is very valid for macro photography as you point out because magnifications are high. I try a brief thought experiment to explain how the effective aperture fits in. With the subject at infinity it is defined using parallel rays of light hitting the top and bottom of the lens. We can assume they are refracted and bent into a cone that focuses into the lens and through the aperture. If the subject comes closer then the rays of light are no longer parallel. They are angled out in a cone from the subject. Assuming the lens can bend the light the by the same amount then the cone of light focused by the lens on the other side will be wider. Given that the physical aperture hole is the same size some light will no longer be focused through the hole (it will hit the black insides of the lens barrel). This reduces the amount of light that gets through the hole and so the effective aperture is smaller.

This leads to the interesting thought that although the effective aperture is reduced when subject distance changes the physical aperture size is not. So assuming the focal length is unchanged will there be any more diffraction? That is can we use the effective aperture calculation in macro photography to help set flash exposure but not worry about diffraction limited resolution. Does anyone have the answer to that? This is clouded by the fact that the focal length will probably change due to focus breathing but I would like to find out anyway.

On a less technical note I do think that fixed aperture zoom lenses have a place in still photography. When manually setting ISO, aperture and shutter speed for the light conditions it is nice to be able to zoom for composition and not have to constantly adjust. Perhaps this just reflects my preferred photographic process but I find it quite common in flat lighting to walk around with my exposure set for a few minutes taking photos of the subject in various compositions knowing that the settings are those I am comfortable with. I can then review them all and know the exposure was the same.

Alex

[Manfred M]

I'm talking about real lenses

http://www.usa.canon.com/cusa/profes...igisuper_86_xs

Not them dinky little things LOL

(I don't mean to be rude btw)

If you look at the specs, the lens is built for a 2/3" imager. The image size is 9.59 mm (W) x 5.39 mm (H).



The imager it is designed for is extremely small when compared to normal HDV sensors. For example, my Panasonic AG100 has a 17.8mm x 10.0mm sensor and it does a 1920 x 1080 native image.