Extension tube and teleconverter

It might be pretty simple to someone professional, but I'm not a pro, so I still have some questions about the issue. Correct me if I'm wrong in any of the statements below:

 

I know that "extension tube" is just a hollow tube which sits between the body and the lens and allows you to make closer shots. Sort of cheap option for macro shots. Have done it with M42.

 

I also know that "teleconverter" incorporates an optical element, sits between the body and the lens and allows you to amplify lense's focal length by 1.4x, 2x or whatever it is, while minimum focal length remains the same. So theoretically one could use a t-con instead of a tube for closeup shots, at certain level (also have done it once or twice with Olympus OM).

 

I know both items do bad things to light, meaning actual working aperture is also multiplied and for example, instead of f/2 min. aperture you get f/4, need longer exposure and so on. But what happens to DOF in this case? If I set the aperture at 2, but light passing through it is equivalent of f/4, will the depth of field also be same as f/4?

 

I understand that buying a dedicated macro lens is the best option since they produce good edge-to-edge sharpness of the image which isn't achievable in other cases, but I'm asking out of curiosity, so please don't dodge my question and don't tell me to go and buy a macro lens :D

Depth-of-field will be whatever the effective aperture and focal length of the lens dictate.

 

A simple extension tube will reduce the effective aperture, but not alter the focal length. The effect is greater with increasing length of tube.

 

A teleconverter will change both the aperture and focal length, regardless of focus distance. The aperture will also reduce as the lens is focussed closer.

 

For example: A 2x teleconverter turns a 50mm f/2 lens into a 100mm f/4 lens, and any aperture number set on the lens needs to be multiplied by 2. Set the lens to f/2.8, and its effective aperture is f/5.6.

The depth-of-field will then be exactly the same as a non-converted 100mm lens at f/5.6.

 

In other words the lens + converter actually become a lens with twice the focal length and having a maximum aperture number twice that of the prime lens.

 

The same applies with a 1.4x, 1.7x or whatever converter. Both the focal length and aperture number are multiplied by the converter power.

 

Any decent depth-of-field table or calculator automatically takes into account the change of effective aperture with focus distance. So all you need to do is dial in the effective focal length, effective aperture and focus distance.

 

The same with an extension tube, except you just use the marked aperture and focal length, together with the focus distance.

 

There is a slight catch however. Conventionally, focus distance is measured from focal-plane to subject, and that's how the focus scale on lenses is marked. Some depth-of-field tables/calculators will use the distance from lens to subject. This doesn't matter for long distances, but for close-ups it becomes vital to know how the depth-of-field is calculated.

 

An easy check is to put less than 4x the lens focal length into the calculator. If this 'breaks' the calculator, then you'll know it correctly expects the focal-plane to subject distance. If not, then it (incorrectly) uses the distance from lens to subject.

 

In any case, depth-of-field is very subjective, and a calculator or screen preview will only give you a rough guide as to how a finished print or screen view will look.

 

BTW. An aperture of f/2 would be referred to as the maximum aperture of the lens, not minimum. The aperture size gets smaller with increasing number, and it's the physical diameter of the aperture we refer to as maximum and minimum. Biggest 'hole' is maximum; smallest 'hole' is minimum.

Edited March 22, 2018 by rodeo_joe|1

That's one way of thinking about it. Here's my way:

 

Depth of field is determined by what happens in front of the lens. For each point in the image, there's a cone of confusion with a base at the effective entrance aperture of the lens, which comes to a point in the focal plane in the scene; the "cone" continues behind this. Anything intercepting this cone influences the image (unless it's blocked by something else). So long as the focal plane is defined, nothing behind the lens (whether that's a teleconverter or a different sensor size) affects the what's captured.

 

However, that tells you what's affecting a given area of the image. A teleconverter effectively gives you a crop of the image you'd have captured otherwise, enlarged to fill the sensor area. If you look at a "native" image from a sensor, it will have more depth of field than a shot through a teleconverter or (equivalently) a crop from the image, because - for the same final image (print or screen) size, you are making less of an enlargement to fill the frame. The more you zoom in on detail, the smaller the depth of field gets.

 

With an extension tube, you are moving the focal point in the scene closer to the lens than would otherwise be the case. That also reduces depth of field (the "cone of confusion" has a more obtuse angle, so it gets wider faster as you move away from the focal plane), and it also gives you a different perspective on your subject - the background will be smaller relative to your subject if you use a wider lens close up than if you use a longer lens from farther away.

 

To an approximation (because actual macro lenses have floating optics and complicated stuff going on), there's a thing called the "thin lens formula". If you want to do a back-of-the-envelope calculation about what your lens is doing, that's a good place to start. An extension tube adds to the distance between the lens and the sensor (which moves the subject distance closer to the lens); a teleconverter effectively makes the focal length longer, which is equivalent to having a smaller image area. You can work out your field of view by drawing straight lines through the lens's optical centre from the limits of the sensor, and by drawing in cones from the bounds of the aperture, you can see what's happening to the depth of field. There are better diagrams on these links than I'd likely manage to produce, so I hope it becomes clear as you look at it.

 

As Joe says, it helps to think in terms of the aperture as a fraction. Actually, it is: f/2 means the (effective) entrance aperture of the lens is half the focal length of the lens. This is why, for example, a 50mm f/2 lens tends to have a front element about 25mm across, and a 200mm f/2 lens has a front element 10cm across. Similarly, at the same focal length, an f/4 lens has an aperture half as wide as one at f/2. The "front element" argument breaks down for wide angles, since not all the front element is contributing to all of the image, but it's a good approximation for longer glass (where the light that contributes to the image enters the front element fairly parallel to the optical axis). If you never want to get confused by "bigger number is a smaller aperture" again, realise that 1/2 is bigger than 1/4, so f/2 is bigger than f/4. And then you know how big the base of your cone of confusion is. :-)

 

I hope that helps.

It certainly helps. However, I'm afraid now I have a cone of confusion in my head :oops:

If you have a camera with TTL metering, then any change in effective aperture will be dealt with by the camera. Doesn't matter if that's due to use of a teleconverter, extension tube, or just plain focusing extension.

 

As for depth-of-field; it is what it is.

The aperture used is more commonly dictated by ISO or shutter speed requirements. And where you do have the luxury of selecting an aperture, it's usually chosen from experience, not by consulting a depth-of-field table.

If you have a camera with TTL metering, then any change in effective aperture will be dealt with by the camera. Doesn't matter if that's due to use of a teleconverter, extension tube, or just plain focusing extension.

 

As for depth-of-field; it is what it is.

The aperture used is more commonly dictated by ISO or shutter speed requirements. And where you do have the luxury of selecting an aperture, it's usually chosen from experience, not by consulting a depth-of-field table.

That's what I usually do - shoot in aperture priority. And if I want some closeup shots, I go as low as f/8, or even more. But I wanted to dig deeper into the matter. It was a bad idea.

"But I wanted to dig deeper into the matter. It was a bad idea."

 

- Gaining knowledge is never a bad idea. It's just that the subject of depth-of-field is a little bit complex.

 

It took me over a week to refine a spreadsheet for the purpose of DoF and other optical calculations. It was an exercise in trigonometry and maths rather than being of much practical use.

 

You're probably better off doing online reading into the subject of photo optics. Asking questions here can open up cans of worms you didn't even know existed!:confused:

Mikheil: Here are some doodles, which may or may not help more than doing maths:

 

These are all side-on views of your sensor, your lens, and what you're shooting. The side view of a cone is a triangle, which makes this easier to draw...

 

In the top diagram (a), you've got how one would normally expect to use a lens: the sensor is a lot closer to the lens (S2) than the subject is (S1), borrowing the terminology from the Wikipedia article I linked above. The "thin lens formula" says that 1/S1 + 1/S2 = 1/f, where f is the focal length of the lens. As the subject gets more distant (S1 tends towards a very big number), 1/S1 tends towards zero - so when you focus at infinity, S2 is roughly f. Since light travels in straight lines and you can think of a lens as a special kind of pinhole, you can work out your field of view for a lens by drawing lines from the edge of the sensor into the scene - as I've shown in red. You can't really separate out the question of field of view from that of depth of field: what is "equivalent" depends on what exactly you're comparing, and if you're looking at a smaller section of the subject or moving the lens, this has a different effect than changing the aperture but staying in the same place.

 

In the second diagram (b), we've added a (2x) teleconverter. That doubles the effective focal length of the lens f, and for focus at roughly infinity, that also doubles S2 - so your field of view is halved (and the subject gets bigger). A lens that's natively twice as long has the same effect.

 

Back to not having a teleconverter. For a simple lens, as you get closer to the subject (S1 gets smaller), the lens extends away from the sensor (S2 gets larger). At 1:1 magnification, S1 = S2, and the lens is twice as far from the sensor - that's diagram ©. True (more than 1:1) macro lenses, such as the recently-announced Laowa 2.5-5x macro, have to extend a very long way from the sensor, which is why it's relatively long for a 25mm lens and a relatively short focal length is still useful. Not only are you getting a reduced field of view of your subject because you're closer to it, but because the sensor-to-lens distance is increasing. An extension tube doesn't make this happen (it'll happen just as the lens focusses), but an extension tube does let you move the lens farther from the sensor than the lens alone would allow. It also typically stops you from getting the lens as close as f to the sensor, so you can no longer focus at infinity.

 

Note that all of this assumes the focal length of the lens doesn't change. In practice, this isn't the case: macro lens working distances don't end up like this, and many lenses show some form of "focus breathing" where the effective focal length changes. A famous example is the Nikkor 70-200 VR II, whose "200mm" end is actually nearer that of a 130mm lens at its shortest focus distance (and even then isn't all that "macro"). Modern lenses tend to be quite "telecentric", so the light doesn't come from where you expect - notably if the lens is retrofocal (for a wide angle) or telephoto (for a longer lens). But I think of that as a small matter of engineering, and you can still use this to reason approximately about lens behaviour.

 

Okay, on to diagram (d). I'm trying to show depth of field for the original lens at a moderate focus distance. For each point that makes up the image, there's a single point in the focal plane in the scene. I've shown green in the middle of the sensor, blue at the top of the sensor and red at the bottom of the sensor (which are mirrored in the scene). Just looking at the green triangles, that's a side-view of the cone of light (as wide at the aperture at its widest point) coming from the lens aperture towards the sensor, which hits the sensor at a point in the centre. There's a corresponding cone in the scene with its tip at the centre of the subject, on the focal plane: light from that point radiates in all directions, and the light rays which hit the lens aperture are those within the green cone. This green cone continues behind the point - any light travelling towards the front of the lens within that cone will contribute to the image. The red and blue triangles show the equivalent cones for the top and bottom of the subject area.

 

There's an alternative way of looking at it: Any point in the scene within the field of view which can see the front of the lens will contribute light to the image. If you work out the distance from that point to the lens, you can work out where the lens will focus that point - and there is a cone with its base at the aperture and its tip at the focus point. Unless that point is on the sensor (if the point isn't on the focal plane, it'll be in front of or behind the sensor), the cone draws a "circle of confusion" on the sensor, which is why out of focus things look blurry.

 

With the teleconverter (or just a longer lens) we have diagram (e). Our physical lens aperture didn't get any bigger - it can't capture any more light - so its f-stop is effectively halved. The sensor is farther from the lens - the cones of light are more acute than before, and the effective aperture of the lens as seen by the sensor reduces. (If you're farther from a circle of light, you receive less light, like being at the bottom of a well. Since for exposure we want to know how much light is hitting the subject, this is why we work with f-stops: the aperture diameter divided by the distance from sensor to lens gives us the effective aperture in f-stops, and for a lens used near infinity where S2 is roughly f, we can describe the light gathering ability of a lens with f = 100mm and an aperture of 50mm as having an aperture of "f/2".) If you moved the camera back so that you had the same field of view at the subject plane (which means you have less field of view behind it and more field of view in front of it), you'd have more depth of field because the lens would look smaller from the subject's perspective. Therefore the "cones of confusion" would be more acute and scene elements would go out of focus more slowly as we moved away from the focal plane.

 

Bottom diagram (f) - we're back to the shorter lens at a close focus distance. This is the cones of confusion again, but for a macro situation. Again, because the sensor moved farther from the lens in order to focus closer, it's at the same distance as with our teleconverter. However, the subject is closer to the lens, and this gives you more obtuse cones of confusion in the scene: the depth of field reduces, because the amount of the scene in front of and behind the focal plane that contributes to a pixel increases more quickly than with the longer distance.

 

Back to the original question. A teleconverter effectively increases f. This means that, to focus at the same distance from the lens, the distance from the sensor to the lens optical centre increases, and you have a narrow field of view of the scene, even at a longer subject distance. An extension tube increases S2 beyond the maximum that the lens would normally allow, allowing you to make a corresponding decrease to S1 by getting closer.

 

I hope that helps. And if it doesn't, I hope I can reuse the diagram on someone. :-)

Andrew, very well designed response! I'm impressed (and better informed).

 

Pulling back to the practical, as opposed to the esoteric: Since the OP explicitly asked that we NOT tell him to buy a macro lens, I won't. In fact, I played with extension tubes and tele-extenders for years before I could afford a high-quality macro lens. Before buying the Micro-Nikkor 105mm/2.8, I only had the Micro-Nikkor 55mm/2.8 and a selection of extension tubes and tele-extenders to play with on my other lenses. Since the vast majority of my macro work was done using flash, the changes in effective aperture were of relatively little import, since I generally shoot at the smallest workable aperture in the interest of DoF. I did notice several other major tradeoffs: First, adding an extension tube to an already close-focusing lens (like the 55mm/2.8) meant that the lens-to-subject distance became too short for effective work. I frequently could not get enough light on the subject with that subject essentially flush with the front of the lens barrel. Second, using extension tubes on longer focal length (and focus distance) lenses was very effective in making for a larger projected subject at the sensor. However, the effective magnification also magnified every tiny flaw and distortion in the lens, resulting in a lower IQ than I had hoped. Still, this was a cost-effective approach until I could afford purpose-built solutions. I just had to adjust my expectations to match the capacity of the equipment. Third, adding a tele-extender to a macro lens can result in usefully higher magnifications. However, this is more true for older, lower-resolution sensors than today's 24+ Mp sensors. Adding more elements, frequently of lower quality and/or less precisely tuned construction, to the light path can add more distortion than would be seen by simply cropping the original image with the non-extended, high-quality lens. I've tried it both ways. In the case of my two Micro-Nikkors, differences in perceptible detail between the cropped and 2x extended images are essentially indiscernible. Meanwhile, the images made using the 2x tele-extender have perceptibly lower contrast and vibrance. (Those two lenses, the Micro-Nikkor 55mm/2.8 and 105mm/2.8 are the two sharpest lenses I have ever owned, so they are good baseline comparators.) However, if you add the 2x tele-extender to the 55mm.2.8, and compare the result to the baseline 105mm/2.8, the non-extended, longer native focal length lens has clearly and substantially higher IQ.

 

I bought extension tube sets for both of my daughters as an economical intro to macro. I haven't bought them tele-extenders, since even today's cheap zoom lenses frequently outperform high quality telephotos/zooms with a tele-extender added. (My relatively cheap Tamron 150-600mm zoom outperforms my Nikkor 70-300mm with 2x tele-extender added.)

Edited March 23, 2018 by DavidTriplett

Thanks to all, I understand a lot more now than I did before. So, to summarize, I'm not burning with the need of taking macro shots. I will get extension tubes with aperture control and use my 18-55 kit lens for closeup work on my Pentax, but I'll also look for Pentax 50 mm f/4 macro, since it's cheaper than other options and will allow me to perform stop-down metering. I guess there's no escape from dedicated macro lenses in a long-term perspective.

If you can spend the extra money, look into a Pentax 50 f/2.8 A series macro. I have had both, and the 2.8 version is a bit sharper, the viewfinder is brighter and it will meter at full aperture with a Pentax DSLR. Stop down metering with Pentax DSLRs is a bit frustrating--sometimes it is pretty accurate, sometimes not in my experience.

Just a thought. Depending on exactly what you are shooting in macros, a 50mm lens might not cut it. For instance, small animals are easily spooked by having a lens 2-10 inches from them. In these cases a longer focal length macro lens (or teleconverter combination) is best because it allows one to be further from the subject with less chance of spooking it. The most common ones are in the 90-105mm range. I use a 50, 90, or 105 depending on my subject and camera body.

A teleconverter, aka telextender, is a camera equivalent of a Barlow lens on an amateur telescope = that is, it is a negative lens arrangement....

 

see In Search of Reach: Tele Converters

for a capsule history and comparison of some of the better ones

If you can spend the extra money, look into a Pentax 50 f/2.8 A series macro. I have had both, and the 2.8 version is a bit sharper, the viewfinder is brighter and it will meter at full aperture with a Pentax DSLR. Stop down metering with Pentax DSLRs is a bit frustrating--sometimes it is pretty accurate, sometimes not in my experience.

It sounds good, I'll keep my eye on those. As for the metering, I have only used Pentax M 50 f/2 on my digital and metering was indeed sometimes a bit off.

Just a thought. Depending on exactly what you are shooting in macros, a 50mm lens might not cut it. For instance, small animals are easily spooked by having a lens 2-10 inches from them. In these cases a longer focal length macro lens (or teleconverter combination) is best because it allows one to be further from the subject with less chance of spooking it. The most common ones are in the 90-105mm range. I use a 50, 90, or 105 depending on my subject and camera body.

I know what you mean, had that exact problem with my Olympus DSLR using Zuiko 35 mm f/3.5 which sometimes used to hit front element on the photographing surface when going down to 1:1. Gosh. That was one hell of a lens. Slow and noisy, but razor sharp from edge to edge! If only Evolt cameras were good enough...

To add just a little more, some lenses are designed to be used with extension tubes.

 

The Nikon 55/2.8 is a well known example. It was designed with 1:1 in the calculation, but without extension tubes only goes down to 1:2.

 

With the lensmakers formula, 1/i + 1/o = 1/f (i is image distance, o is object distance)

and the (de)magnification ratio is o:i, you can figure out the extension, either from the lens

focus helix or extension tube is needed.

 

For 1:1, i=o=2f

for 2:1, o=2i=3f, or i=1.5f

 

The extension tube to get from 1:2 down to 1:1 is f/2.

 

Lenses not designed for macro use are not optimized for these distances.

Glad to help, David. :-)

 

Another thing to throw into the mix: some non-macro lenses are better than others at close distances. Sigma have historically described several budget zooms as "macro" because they'll focus closer than most competing zooms, although not necessarily with particularly great quality. But I agree with others here about working distance (which is a major reason I have a 150mm Sigma macro and why I'm not happy about my old 90mm Tamron macro having such a heavily inset front element); if you don't actually need 1:1, some lenses are much better than others at getting close. A well-known case in point is that Nikon's 300mm f/4 AF-S gets to 0.27x magnification (about 1:4 - the latest version is slightly worse), which has been known to be plenty for getting, for example, dragonfly shots, while being at roughly 5' from the subject. A true macro that could get a lot closer would likely spook the insect. Other lenses are much worse at getting close, especially fast ones - for example, my 200mm f/2 doesn't even reach 1:8. Sadly I'm less familiar with the Pentax range. Unfortunately, the longer your lens, the less effect an extension tube has.

 

Have we mentioned diopters ("close-up filters") yet? While there are those who argue that autofocus is of limited used with macro (though it does offer cheap focus stacking and on a good day can help with moving subjects), this approach does at least keep the camera connection to the lens unmodified. Unfortunately, I believe diopters tend to come in two forms: the vaguely decent but expensive (typically cemented doublets) and the cheap but optically iffy. They do offer an even cheaper alternative to extension tubes and macro lenses, though - and they're not system-specific.

I deliberately avoided the subject of close-up lenses, cause even though I have never owned them, the results I've seen from others are unimpressive. It's like putting on the eyeglasses on the lens - center of the image is more or less acceptable, with strong distortions and loss of quality everywhere else. Probably those were the cheap ones.

It's like putting on the eyeglasses on the lens - center of the image is more or less acceptable, with strong distortions and loss of quality everywhere else.

 

Yes, that's basically how they work. :-) We are talking about budget solutions other than a dedicated macro lens, of course. (One of the primary ways to get a cheap macro is definitely to buy an old one - they're usually still optically good, not least because you usually need to stop down to get adequate DoF and this hides a lot of optical aberrations, and manual focus isn't such a hardship. But that may be where you were looking already.)

 

I understand that the pricier diopters are decent, but they may also be a decent chunk of the cheapest used macro lenses. I have a couple of cheap ones, but just accept iffy corners when I'm using them - and they mostly got used on medium format (coincidentally, on my Pentax).

From what I've seen so far, the cheapest truly macro lens for Pentax is 50 mm f/4 (either K-mount, or the fabled M42 Takumar). Both can be acquired for 60$ if I'm lucky enough. Tubes usually cost six times less. I already have those tubes for M42 mount and I have used it on my K-x via M42 to K adapter, coupled with Soviet Industar 50 lens (Tessar design, 50 mm f/3.5). I doubt that Pentax's own K-mount 50 mm will perform much better with tubes than the above mentioned option, except it will have automatic aperture control.

The Pentax macro is designed to be used for close work, which is to say it is corrected for this. Your Industar would have been optimized for infinity. The Pentax will also focus continuously from infinity to 1:2 without hassling with removing the lens to insert another extension tube, etc. so it will also be a lot smoother to work with. An automatic diaphragm is a great convenience, especially if you're working at f/11 or f/16 as you frequently need to at high magnification to have any depth of field.

This is the usual conundrum of whether your time and convenience are worth more than the money you will spend to get the convenience. Only you can answer that question for yourself.

The Pentax macro is designed to be used for close work, which is to say it is corrected for this. Your Industar would have been optimized for infinity. The Pentax will also focus continuously from infinity to 1:2 without hassling with removing the lens to insert another extension tube, etc. so it will also be a lot smoother to work with. An automatic diaphragm is a great convenience, especially if you're working at f/11 or f/16 as you frequently need to at high magnification to have any depth of field.

This is the usual conundrum of whether your time and convenience are worth more than the money you will spend to get the convenience. Only you can answer that question for yourself.

In this case my wallet answers for me, but I completely agree with all of the points you've made.

Good quality enlarging lenses make excellent macro lenses. If you're seriously into macro, it's worth getting a bellows unit to take advantage of them.

 

With darkroom gear going for a song these days, you can pick up enlarging lenses for 1/10th of their new price. Schneider Componons and Rodenstock Rodagons are the ones to go for IME. For some reason El Nikkors hold their price, despite being average performers. Durst Neonons can be quite good as well.

 

Enlarging lenses generally have a 39mm Leica thread, and adapters to this size are readily available in most camera fittings.

 

Most top quality and recent enlarging lenses have an illuminated aperture scale, and this needs to be blocked off when used as a taking lens. Otherwise you'll get red or green flare in your macro shots.

Edited March 27, 2018 by rodeo_joe|1

Good quality enlarging lenses make excellent macro lenses. If you're seriously into macro, it's worth getting a bellows unit to take advantage of them.

 

With darkroom gear going for a song these days, you can pick up enlarging lenses for 1/10th of their new price. Schneider Componons and Rodenstock Rodagons are the ones to go for IME. For some reason El Nikkors hold their price, despite being average performers. Durst Neonons can be quite good as well.

 

Enlarging lenses generally have a 39mm Leica thread, and adapters to this size are readily available in most camera fittings.

 

Most top quality and recent enlarging lenses have an illuminated aperture scale, and this needs to be blocked off when used as a taking lens. Otherwise you'll get red or green flare in your macro shots.

Sadly my father's enlarger had Zeiss Biogon 35 mm f/2.8 which although M39, has a huge lens protruding from behind. No way I could install it on anything except mirrorless (which I don't possess). I think other lenses will have the same issue with flange distance.

 

Now, if I want to use that lens, I have to attach it to my old FED camera (which is a bad clone of Leica). No closeup shots or anything remotely similar.

I have found Sigma 50 mm f/2.8 macro lens with 1:1 ratio. It's a manual focus lens and I'm sure it will be outperformed by Pentax macro lenses, but it's too cheap to pass it by. Now if only my salary can arrive a bit sooner this month :D