Using teleconverters

[Bart_van_der_Wolf]

Are you sure about this, Bart? Here's an extract from the USA Canon site at http://learn.usa.canon.com/resources/articles/2014/ef_extenders_pt2.htmlp

"Change in lens AF speed with EF Extenders

Because AF systems are essentially computer-controlled to read and react to focus distance changes, the information must be modified so that the focusing movement (or sensitivity) compensates for the added presence of the extender. In the Canon EOS system, this is done by deliberately reducing drive speed when an extender is detected.

Before you immediately conclude that this is a problem, understand that this reduction in drive speed now corresponds to the effective speed you would achieve with the same EF lens alone. It compensates, automatically, for the reduced distance lens elements in the lens’s focusing group(s) need to move to refocus on a subject, with either EF Extender in place. Accordingly, overall AF performance remains essentially unchanged with an EF Extender attached, versus the lens’s AF speed without an extender."

Hi Ray,

It admittedly gets a bit complicated. There is a difference between precision (which is what Canon refers to), and accuracy (which is what I refer to). What Canon is referring to is that the reduced effective aperture (giving more DOF) is compensated by the longer focal length (giving less DOF). Therefore, the focusing mechanism (which operates in focus steps of a given size) reaches the same effective precision in placing the DOF around the optimal focus plane. The lens elements can only be moved within the step size limits that the lens offers. The darker image may also push detection capability a bit more to its limits.

However, the increased magnification of the image will allow to focus more accurately (within the limits of the AF detector lens aperture) when the micro adjustment is available (the focus step chosen is more accurate) for the lens+extender (within the precision of the focus mechanism), and when contrast detect focusing is used (on some models) because the image is larger.

Image stabilization will allow any focus method to acquire more accurate focus, because the image is less motion blurred, also at the moment of focus detection (although custom functions allow to ignore/override accurate focus confirmation).

Cheers,
Bart

[BJL]

OK, let's start afresh: 400mm f/5.6 and 200mm f/2.8 with 2X teleconverter will produce exactly the same optical result full open. That is really all there is to it.

DOF comparisons get controversial partly becuase different people assume different comparison conditions, so I can see tha 2x covertor either halving DOF or not changing it:

a) If you print the entire image recorded by the sensor at the same total size, the 2x TC image will have half the DOF: the image of the subject will be twice as big, and the circle of confusion at each point of the image will also be, so you only have to look at something half as far from the plane of critical focus in order to see OOF effects.

b) If for some reason you display the images with the subject the same size (TC print half as big in each dimension, so as if you had just cut down the print of the non-TC version) and view from the same distance, DOF will appear the same.

c) But if you then view those different size prints at distance proportional to print size, you are back to half the DOF with the TC version.

c) And in the new-fangled style of comparing at equal PPI and equal viewing distance, it is again half the DOF with the 2x TC.


So to my preferred way of comparing (version (a)), a TC reduces DOF by the magnification factor of the TC.

[dwswager]

DOF comparisons get controversial partly becuase different people assume different comparison conditions, so I can see tha 2x covertor either halving DOF or not changing it:

a) If you print the entire image recorded by the sensor at the same total size, the 2x TC image will have half the DOF: the image of the subject will be twice as big, and the circle of confusion at each point of the image will also be, so you only have to look at something half as far from the plane of critical focus in order to see OOF effects.

b) If for some reason you display the images with the subject the same size (TC print half as big in each dimension, so as if you had just cut down the print of the non-TC version) and view from the same distance, DOF will appear the same.

c) But if you then view those different size prints at distance proportional to print size, yo are back half the DOF with te TC version.

c) And in the new-fangled style of comparing at equal PPI and equal viewing distance, it is again half the DOF with the 2x TC.


So to my preferred way of comparing (version (a)), a TC reduces DOF by the magnification factor of the TC.

Let us simplify.  A 400mm lens shot at f/5.6 and location X focused on subject at location Y on a camera with sensor W will have the same DOF as a 200mm lens shot at f/2.8 with a 2x TC (effective aperature of f/5.6) and location X focused on subject at location Y on the same camera with sensor W.   

All the rest is useless mubmo jumbo.

[Ray]

I'm still waiting for someone to post comparison images showing how much sharper hand-held shots can be when using a teleconverter, than the same scene shot without teleconverter, after cropping, interpolating and appropriately sharpening the image without teleconverter, or downsampling the image from the teleconverter.

I appreciate the argument that a very high quality lens used with a high quality converter may rival the quality of a mediocre telephoto lens of the same equivalent focal length, and if one is satisfied with the performance of mediocre lenses of longer focal length, then it's much more convenient to buy a good extender for use with one's super-great prime which one already owns and is very proud of.

However, one can't deny that the better quality lens that one uses with the extender can also produce better quality images without the extender.

[telyt]

I appreciate the argument that a very high quality lens used with a high quality converter may rival the quality of a mediocre telephoto lens of the same equivalent focal length

Is the Leica 400mm f/4 APO a mediocre lens?  Compared with the Leica 280mm f/4 APO (a very high quality lens) used with the Leica 1.4x APO extender (a high quality converter), the only advantage of the 400mm f/4 APO is one stop faster maximum aperture.  The 280mm lens with 1.4x extender out-resolves the 400mm f/4 until the 400 is stopped down to f/8.

This is a crop from a photo made with stacked extenders - the 280/4 APO with 1.4x and 2x APO extenders.  I won't make big prints of this photo; I think this is where the prime plus extender(s) rivals the quality of a mediocre 800mm lens.

[Ray]

Is the Leica 400mm f/4 APO a mediocre lens?  Compared with the Leica 280mm f/4 APO (a very high quality lens) used with the Leica 1.4x APO extender (a high quality converter), the only advantage of the 400mm f/4 APO is one stop faster maximum aperture.  The 280mm lens with 1.4x extender out-resolves the 400mm f/4 until the 400 is stopped down to f/8.

This is a crop from a photo made with stacked extenders - the 280/4 APO with 1.4x and 2x APO extenders.  I won't make big prints of this photo; I think this is where the prime plus extender(s) rivals the quality of a mediocre 800mm lens.

I have no experience with those Leica lenses you mention, Doug. However, logically, if you describe the Leica 280mm as a very high quality lens which can out-resolve the Leica 400/F4 when a 1.4x converter is attached, then one might reasonable conclude that the Leica 400mm lens is either not a 'very high quality' lens, or that you just happen to have a copy which is rather low in the acceptable QC range.

I would like to see comparisons of the same image, with and without extender, which include the disadvantage of one stop more noise, camera hand-held. The reason I would like to see such a comparison is because I don't own any 'very high quality' primes that can fit to the two extenders I have, so I cannot test this myself. All the tests I've done with merely 'good' lenses indicates there's no worthwhile image quality improvement, although I accept there can be an advantage in being able to see more clearly what one is shooting as a result of the greater magnification.

[telyt]

I have no experience with those Leica lenses you mention, Doug.

You've said it all right there.

[Ray]

You've said it all right there.

No, I said more than that. 

I said I would like to see comparisons of the lens with and without converter. Your picture of the bird is very nice and certainly acceptably sharp at the size presented, but gives me no idea of the relative quality of images with and without converter in circumstance when one cannot avoid the f/stop disadvantage with the converter.

[BJL]

A 400mm lens shot at f/5.6 and location X focused on subject at location Y on a camera with sensor W will have the same DOF as a 200mm lens shot at f/2.8 with a 2x TC (effective aperature of f/5.6) and location X focused on subject at location Y on the same camera with sensor W.   

Agreed: a TC changes both the focal length and the aperture ratio from that marked on the lens by the same factor, because the aperture ratio is focal length divided by effective aperture diameter (= entrance pupil size) and the latter is not changed by the TC. But ...

All the rest is useless mubmo jumbo.

... given that the original question was about how attaching a TC affects DOF, the rest is actually relevant to answering that question!  Especially since there is so much confusion about how DOF (how visible OOF effects are) is affected by choices of how to view the image (how big, from how far away, etc.), not solely by traditional "guideline" formulas using only focal length, aperture, subject distance and such.

[kers]

...I'm still waiting for someone to post comparison images showing how much sharper hand-held shots can be when using a teleconverter, than the same scene shot without teleconverter, after cropping, interpolating and appropriately sharpening the image without teleconverter, or downsampling the image from the teleconverter...

i do not exactly know what you mean, but if a converter is made for the specific lens ussually the quality of the lens is very good and the converter makes sense...
In all other cases it does not work...at least that is what i have noticed with this specific converter: the nikon 2x versionII.
I have found out it works really well with the new Nikkor 300PF ; a lens that is very sharp and the center can handle a 135MP full frame.
I have added some 100% crop-samples ....

PS
i have used a dNikon d810..
the lens is only 755 ; gram the converter is about 450 gram... a 600mm F8 at about 1200 gram... not bad...

[BJL]

I would say that extenders make a lot of sense with high quality optics that are outperforming the sensor. In real world I would say that shooting a cropped sensor camera without extender could be better than using an extender on full frame.

Yes, or just using a sensor that allows such cropping: having a sensor with twice the pixel count than one needs for long telephoto shots allows a 1.4x crop (to half the image area) as an alternative of using a 1.4x TC.  And with a lot of modern sensors, the full resolution is mostly of value with stationary or very brightly lit subjects (landscapes, architecture, studio portraits, etc.) so for telephoto shots of wildlife, sports and such, half the total pixel count is often enough. This cropping approach allows "loose framing", to be able to correct the framing in the case that a bad prediction of subject motion causes subject to ends up framed poorly.  On the the hand (as indicated in Reply #25 above by dwswager) the cropping approach can have the problem of a far smaller OVF image of the part of the scene that you actually care about, and can mess up light metering.  That is why I like the "digital teleconvertor" mode of some EVF cameras like the EM5, which can in principle also do light metering based only on the intended crop (but I do not know if the EM5 meters this way).  This mode only crops the in-camera JPEG to what is seen in the EVF; the raw file still records the whole image, so preserves the ability to correct the framing if needed.  Unfortunately the EM5 only offers a 2x digital TC; I would like a 1.4x option.

On another point, it seems that most comments in favor of TCs are about 1.4x ones, while most poor experiences reported are with 2x TCs.  That fits what I have often heard: that 2x TCs are often it disappointing.  But 1.4x is where I am comfortable instead with having enough sensor resolution to support cropping!

Ray: that 4 stop increase is correct in the worst-case scenario of a 2x TC and hand-holding (with no or insufficient IS), which is the scenario where the shutter speed would also need to be doubled.  If instead the shutter speed needed is dictated mainly by subject motion, there is not necessarily any need to increase shutter speed.  For example, the same shutter speed that works to freeze motion of a subject with a 200mm focal length at 50 meter range also works with a 400mm focal length at 100 meter range (these alternatives giving about the same framing of the subject).  So in the best-case scenario of 1.4x TC and camera on tripod, or with any stabilization system good enough that only subject motion is a problem, then only a doubling of exposure index is needed, and the increased noise effect from that is probably comparable to getting the same pixel count on the subject by halving pixel area instead of using a 1.4x TC.

[Ray]

i do not exactly know what you mean, but if a converter is made for the specific lens ussually the quality of the lens is very good and the converter makes sense...
In all other cases it does not work...at least that is what i have noticed with this specific converter: the nikon 2x versionII.
I have found out it works really well with the new Nikkor 300PF ; a lens that is very sharp and the center can handle a 135MP full frame.
I have added some 100% crop-samples ....

PS
i have used a dNikon d810..
the lens is only 755 ; gram the converter is about 450 gram... a 600mm F8 at about 1200 gram... not bad...

Thanks,
In the examples you show above, the EXIF data indicates that both shots are at F16. That suggests you used the 300mm lens at F8 with 2x converter, then compared the result without converter, with the lens set at F16. The fine detail in the F8 shot is of course sharper, as you would expect any good lens to be sharper at F8 than at F16.

If you are going to be objective with such tests, using this approach, you should choose a scene with great depth, and then also compare the out-of-focus parts. You might then see that the OoF parts in the shot without converter are more significantly sharper than the in-focus parts are sharper in the teleconverter shot. 

[Ray]

Ray: that 4 stop increase is correct in the worst-case scenario of a 2x TC and hand-holding (with no or insufficient IS), which is the scenario where the shutter speed would also need to be doubled.  If instead the shutter speed needed is dictated mainly by subject motion, there is not necessarily any need to increase shutter speed.  For example, the same shutter speed that works to freeze motion of a subject with a 200mm focal length at 50 meter range also works with a 400mm focal length at 100 meter range (these alternatives giving about the same framing of the subject).  So in the best-case scenario of 1.4x TC and camera on tripod, or with any stabilization system good enough that only subject motion is a problem, then only a doubling of exposure index is needed, and the increased noise effect from that is probably comparable to getting the same pixel count on the subject by halving pixel area instead of using a 1.4x TC.

BJL,
We hear a lot about the additional shutter speed that a high-resolution sensor requires to take full advantage of that increase in pixel count. Michael mentioned this in relation to the D800 when it first became available.

A 2x converter effectively quadruples the pixel count, compared with the same FoV from the lens without converter. Regardless of whether the movement is due to subject movement or camera shake, I would think that in order to get the maximum resolution advantage when using a converter, one should always increase the shutter speed beyond what one would use shooting the same scene without converter, when the camera is hand-held.

The question is, by how much? Imagine using a particular lens on a 9mp camera, then upgrading one's camera to 36mp and using the same lens. In order to see the full advantage of the increased resolution potential of the 36mp camera, would one not be advised to quadruple shutter speed?

I'm not sure why you are comparing in your above comment, a 200mm lens at 50metres and a 400mm lens at 100 metres. If one uses a 200mm lens with 2x converter it is presumably to get an enlarged image (greater pixel count) from the same distance.

[kers]

Thanks,
In the examples you show above, the EXIF data indicates that both shots are at F16. That suggests you used the 300mm lens at F8 with 2x converter, then compared the result without converter, with the lens set at F16. The fine detail in the F8 shot is of course sharper, as you would expect any good lens to be sharper at F8 than at F16.

If you are going to be objective with such tests, using this approach, you should choose a scene with great depth, and then also compare the out-of-focus parts. You might then see that the OoF parts in the shot without converter are more significantly sharper than the in-focus parts are sharper in the teleconverter shot.  

Dear Ray, don't believe everything you read ( in the EXIF)  I made a double layer photoshop file in order to straighten the photographs for your convenience..
As you understand both layers have now the same exif....  
If i would like to foul you i would do so... but this is meant to be an informative forum....
So i try to be informative... that is all...

..

back to the real world..

If you want to know how the shots are taken please ask me and i will tell you ..

The shot with the 300mm PF is made at 1/13 sec;   f/16;   ISO 64

the shot with the 300PF + 2x converter is made at f5.6 1/400  64 asa...
I tried to get the shots as steady as possible- because i was interested in the optical performance.
As you can see the d16 shot + converter contains much more information..
That is the point  ( my unbiased opinion) i was trying to make... and i believe i made that clear to anyone but...  

[dwswager]

But ...
... given that the original question was about how attaching a TC affects DOF, the rest is actually relevant to answering that question!  Especially since there is so much confusion about how DOF (how visible OOF effects are) is affected by choices of how to view the image (how big, from how far away, etc.), not solely by traditional "guideline" formulas using only focal length, aperture, subject distance and such.

But DOF is a choice, or at least should be.  And it is a choice made based on the intended enlargement.  Calculating DOF with a single Circle of Confusion for all occasions is a mistake.  You are either going to be disappointed when enlargement ratios are large (too loose CoC) or be making inappropriate choices when enlargements ratios are going to be small because you will select too small an aperture (CoC too stringent).

This is one reason that when going large, I find viewfinder DOF preview inadequate.

[Ray]

If you want to know how the shots are taken please ask me and i will tell you ..

The shot with the 300mm PF is made at 1/13 sec;   f/16;   ISO 64

the shot with the 300PF + 2x converter is made at f5.6 1/400  64 asa...

That doesn't make sense to me, or have you just made a typo? Or perhaps you're just trying to have fun.  

Why are you comparing a lens without converter at F16 with the same lens at F5. 6 using a 2x converter which drops the F/stop reading to F11.
For both shots, the 300mm lens should be set at the same f/stop, whatever you choose, otherwise the DoF will be different.

It's been explained over and over again in this thread that a 300mm lens at F5.6 will have the same DoF as that same lens used at the same aperture with a converter. The effect of a 2x converter is to unavoidably drop the f/stop by 2 stops, ie. F5.6 becomes F11, but the aperture of the 300mm lens is still physically the same despite the reading being F11. The converter does not go in front of the lens but behind it. You must have noticed that.  

[Bart_van_der_Wolf]

I tried to get the shots as steady as possible- because i was interested in the optical performance.
As you can see the d16 shot + converter contains much more information..
That is the point  ( my unbiased opinion) i was trying to make... and i believe i made that clear to anyone but...  

Hi Pieter,

Indeed. When we want to know if a, in this case 2x, converter or extender can deliver more detail to begin with, it only makes sense to eliminate as many of the other variables as possible. By now, we've (hopefully) established that the addition of an extender will not change the DOF, and that there is more detail to be had with good matching optics.

We can now concentrate on other factors that have to do with getting the shot, and that's about technique, not about the extenders per se. BTW it helps some folks to reduce the caffeine intake before shooting handheld telephoto shots, although image stabilization does also help in that case.

We hear a lot about the additional shutter speed that a high-resolution sensor requires to take full advantage of that increase in pixel count. Michael mentioned this in relation to the D800 when it first became available.

A 2x converter effectively quadruples the pixel count, compared with the same FoV from the lens without converter. Regardless of whether the movement is due to subject movement or camera shake, I would think that in order to get the maximum resolution advantage when using a converter, one should always increase the shutter speed beyond what one would use shooting the same scene without converter, when the camera is hand-held.

The question is, by how much? Imagine using a particular lens on a 9mp camera, then upgrading one's camera to 36mp and using the same lens. In order to see the full advantage of the increased resolution potential of the 36mp camera, would one not be advised to quadruple shutter speed?

Since a 2x extender magnifies the optical image on the sensor by a linear factor of 2x, it would make sense to also double the shutter speed (reduce the exposure time) to get the most out of the additional resolution. That not only reduces the effect of also magnifying camera shake (esp. without stabilization), but will also help to reduce the subject motion that was previously undetectable without extender.

One can always reduce the exposure time even more if subject motion is an issue, but that also has nothing to do with extenders per se. If one routinely shoots subjects in motion, at low light levels, it would be advisable to use a better lens than to use an extender, if only to gain one or two stops of exposure speed. Nobody denied that an extender will reduce the maximum amount of light available, but as the doctor said to the patient; If it hurts when you push there, then don't push there ...

Cheers,
Bart

[spidermike]

It's been explained over and over again in this thread that a 300mm lens at F5.6 will have the same DoF as that same lens used at the same aperture with a converter. The effect of a 2x converter is to unavoidably drop the aperture by 2 stops, ie. F5.6 becomes F11, but the aperture of the 300mm lens is still physically F5.6 despite the reading being F11. The converter does not go in front of the lens but behind it. You must have noticed that. 

Are you sure?
The physical aperture is the same, I agree. But the 'f stop' is the ratio of physical aperture size to focal length, so if you add a teleconverter to go from 400mm to 800mm, the physical aperture is constant so the f value changes accordingly.

[kers]

...
We can now concentrate on other factors that have to do with getting the shot, and that's about technique, not about the extenders per se. BTW it helps some folks to reduce the caffeine intake before shooting handheld telephoto shots, although image stabilization does also help in that case.
....
Cheers,
Bart

To begin with i have to say that i am not very familiar shooting long lenses , say 300mm and beyond... but already i noticed that shooting handheld is a skill. I have shot this PF lens sharp at 1/30s (VR on) but also had moved shots at 1/320.
One thing i know is that shooting with a mirrorless camera with silent electronic shutter makes a clear positive difference. I have shot this 300PF with a Nikon V3 and got definitely more sharp images. (even on a tripod)
To answer an other question :  gives the Nikon V3 ( 18MP - factor 2.7) more detailed images than the 2x converter; i say no, but it is close. It is because of the lesser quality of the sensor.
 
cheers PK

[Bart_van_der_Wolf]

The physical aperture is the same, I agree. But the 'f stop' is the ratio of physical aperture size to focal length, so if you add a teleconverter to go from 400mm to 800mm, the physical aperture is constant so the f value changes accordingly.

Correct, the f-stop is a ratio (focal length/diameter) and is therefore expressed as e.g. f/8 .
That can also be easily seen when looking through the lens from the rear, with and without extender.
The exit pupil of the lens with extender is proportionally smaller.

Cheers,
Bart