Extenders compared: Canon 1,4 III vs Kenko 1,4 DGX MC4 vs no extender

[photojoern.de]

I just received the Canon extender 1,4 Mark 3 version and hoped to improve my focal length of my combo 5 DS R and Canon 100-400 II L f4,0-5,6 to 560 mm f8,0. I compared photos with the Canon extender to photos with the Kenko 1,4 extender versus crops without the extender? Which combination is the best? I would say: none of the extenders did the job properly.
What do you think?
Did I miss anything? Do you have a better experience with any of the converters? See sample pictures and my methodology here:
http://photojoern.de/gear/canon-extender-14-real-world-test-canon-14-iii-vs-kenko-14-dgx-mc4/

 

[hendrik-sg]

I did similar Trials with a 7dii, 1.4iii and 100-400ii, with the 7dii having similar Pixel density as your 5ds.

I hardly did get more Resolution than with the bare lens. with extender there is a loss from the extender, much decreased AF Performance, half shutter Speed or 2x ISO, and difraccion should begin to show up even at f8.0. On top it's more difficult to compose a Picture specially under time preassure.

So i got better results with the bare lens, just because the the risk of ruining an shot for any of the above reasons is much less. On the small display of the camera the exeeded range is looking spectacular of course.

generally, People (or at least me) tends to use extenders to much, in all than perfect lighting conditions cropping is less spectacular, but gives better results, at least for my possibly poor skillls.

Differet game was with a loaned 500 4.0 ii of course, here the 1.4x and even the 2x extender gives a benefit, but this is a different class of lens.

I kept the 1.4x for my 5d3, there it keeps is's promise, and replaces the (loaned) 7dii, at least for static subjects, and emergency use, but not for tracking BIF of course.

 

[candc]

the canon extenders are good with the supertele primes. i have several brands of extenders that i have used with a variety of lenses and i have gotten mixed results. i don't have the combo you are using but i would have expected a bit better results with the canon 1.4xiii. there are copy variations with extenders just like lenses. with most lens/extender combos you will get significantly better iq if you stop down 1 stop.

 

[Don Haines]

If you are using a crop camera, or a FF camera with a high pixel count, you need an insanely sharp lens (big whites) to get more reach with a teleconverter. As nice as the 100-400 II is, it's not good enough so your results are what I would expect.

 

[jasny]

Was AFMA done WITH the extender?

 

[AlanF]

I've tested the 100-400mm II ± 1.4xTC III ad nauseam on the 5DS R and 7DII. I've used charts, sitting birds, and brick walls and there is no doubt for me the extender does increase resolution.

Here are some typical shots of mine from the 5DS R, combining a brick wall and a chart. I think my bare lens is sharper than yours and you can see from the charts that the resolution increases with the 1.4xTC. Having said that, I tend to use the bare lens because I am an opportunistic bird photographer and put on the TC only when I think it will help. Top one is 400mm, bottom 560mm.

 

[Loswr]

Was AFMA done WITH the extender?

And without, for that matter...

If you're going to rely on AF for a lens sharpness (generally a bad idea, although I acknowledge the 'real world' intent), you should at least start with properly adjusted lenses.

 

[AlanF]

Here are some bird shots. I regularly shoot a peregrine falcon which roosts about 45m up the University Library at about a distance of about 60m. Here in descending order the 5DS R with my second copy of the 100-400mm II, the 300/2.8 II + 1.4x TC, on the same occasion, and the 100-400 at 560mm and the Sigma 150-600mm at 600mm on its very first outing. They are A1 Servo but that is all part of the real world test (and hand held, to boot, which favours the shorter lenses).

 

[Mt Spokane Photography]

Your test method did not mention manual focus or live AF, so we assume autofocus. If that's the case, you are testing autofocus accuracy, not the lens or TC.

A test needs to remove all variables, have accurate exposure, and no vibration. That is much more difficult than it seems at first, trucks and automobiles, even at a distance shake the ground and houses.

I've been involved with labs that sink piling into bedrock 200 ft down out in the desert away from traffic in order to reduce vibration levels for sensitive tests. That's extreme, but it also shows that vibration and movement is always there. My lab used a 4 X 10 ft X 6 inch thick plate mounted on special shock absorbers for optical measurements with our bench.

 

[AlanF]

Come on guys! These aren't measurements to a thousandth of 1% but estimations of the usefulness of a 1.4xTC in practice. I use vibration-free optical benches for our ultra-sensitive experiments, but I don't take one out with me on a hike. Here we are talking about practical photography in the real world. When I do test lenses, I always take at least 10 shots to look at the variation - repeat measurements are crucial in all testing.

 

[Kerry B]

I have done a fait bit of testing the Canon 1.4 + 2 x extenders mkiii. Here are the results of tests undertaken on a 7dmkii. I mounted the camera on a tripod and used the Canon 300f2.8ii lens. I kept the same distance between all shots. I used A1 servo mode with high speed framing to replicate what I shoot normally. I am pleased with the results.

 

[LSeries]

I have done a fait bit of testing the Canon 1.4 + 2 x extenders mkiii. Here are the results of tests undertaken on a 7dmkii. I mounted the camera on a tripod and used the Canon 300f2.8ii lens. I kept the same distance between all shots. I used A1 servo mode with high speed framing to replicate what I shoot normally. I am pleased with the results.

The 300 f/2.8L II is known to take extenders really well.

 

[Ozarker]

Here are some bird shots. I regularly shoot a peregrine falcon which roosts about 45m up the University Library at about a distance of about 60m. Here in descending order the 5DS R with my second copy of the 100-400mm II, the 300/2.8 II + 1.4x TC, on the same occasion, and the 100-400 at 560mm and the Sigma 150-600mm at 600mm on its very first outing. They are A1 Servo but that is all part of the real world test (and hand held, to boot, which favours the shorter lenses).

Beautiful bird. My favorite.

But, I think the test is more towards your holding technique rather than a comparison between the various combos.

I think your holding technique is far better than mine.

 

[AlanF]

The Swedish site objektivtest.se measures MTFs on an optical bench like lensrentals does. This is the type of test the purists like, with no artefacts from a camera. It has measured the MTFs of the 100-400mm II and 300mm f/2.8 II ± 1.4xTC III.

http://www.objektivtest.se/tester/canon-ef-100-400-mm-f45-56-l-is-ii-usm-test/
http://www.objektivtest.se/tester/canon-extender-ef-14x-iii-test/

There is only a 10% hit for a 40% gain in focal length.

The 2xTC on the 300mm is not as good a result. The contrast is hit by some 20% - the 20 cycles/mm. But there is a huge hit at 30 cycles/mm, which is a measure of resolution, of about 30%

http://www.objektivtest.se/tester/canon-extender-ef-2x-iii-test/

My 300/2.8 II + 2xTC was brilliant on my 5D3 but not on my 7DII and 5DS R, and I prefer to use it with the 1.4xTC. I am looking forward to a 5D4 with larger pixels for the 600mm lenses I have.

 

[photojoern.de]

Thanks for the replies so far.

If you are using a crop camera, or a FF camera with a high pixel count, you need an insanely sharp lens (big whites) to get more reach with a teleconverter. As nice as the 100-400 II is, it's not good enough so your results are what I would expect.

Sorry, but I do not get that point. Doesn´t sound logical for me. And the resolution if the 100-400 II is already really great, only inferior to very few prime big glasses.
In general, I believe from the replies and tests, that either photographers are also not really happy with the extenders, or they use them with results that show minimal improvements.

Regarding the point of "camera shake" etc.: Yes, I do sports photography with the 100-400, usually not landscape. So there is always minimal camera shake, but then there is exposure time of 1/320 and shorter for a non-moving object, a monopod like support and the 4-stop IS. It won´t get sturdier in real world applications. My opinion.

Re autofocus: yes, I used the autofocus, not the manual focus. And yes: I did not have the time of adjusting the autofocus to the combo of 100-400 plus extender. So there is a possibility that the extender needs a different adjustment for the autofocus, other than the bare lens.

Still, my conclusion is for now: using 400mm and then cropping gives me more speed / less ISO plus more flexibility compared to extender usage.

 

[AlanF]

I did not have the time of adjusting the autofocus to the combo of 100-400 plus extender. So there is a possibility that the extender needs a different adjustment for the autofocus, other than the bare lens..

You should repeat your measurements having done AFMA with the extender on. Here are the values for my two copies of the lens with two extenders, in both cases there is a noticeable 5 AFMA units difference:

Old 100-400 Old TC
400mm 3
560 mm -2

New 100-400 New TC
400mm 0
560 mm 5

You should have been able to see from my posting of both charts and birds that you really do get extra resolution with the TC on. There are, indeed, several reasons for using the TC only when strictly necessary, but when the conditions are right, it is a very useful plus.

A about comparisons, which is often ignored: you will see an improvement in resolution only when there are fine details that are just beyond the resolution of the bare lens but are resolved with an extra 40% of reach. It's obvious, but you often see photos of aircraft etc with no fine detail used to "prove" a point that thye never could.

 

[Loswr]

Re autofocus: yes, I used the autofocus, not the manual focus. And yes: I did not have the time of adjusting the autofocus to the combo of 100-400 plus extender. So there is a possibility that the extender needs a different adjustment for the autofocus, other than the bare lens.

However, you stated that the main question that you set out to address was, "Do I get a better image quality with an extender compared to cropping the photo?" Unfortunately, your testing methods precluded you getting a robust answer to that question.

FWIW, my 1.4xIII combos require AFMAs of 2-4 units different than the bare lenses, and that's certainly enough to make a difference in pixel-level sharpness. Using the extender in reach-limited situations definitely increases resolution with my 600/4 II and 70-200/2.8 II.

Still, my conclusion is for now: using 400mm and then cropping gives me more speed / less ISO plus more flexibility compared to extender usage.

I think that is quite evident without needing shots of a brick wall. Of course, if those are your main considerations then you'd be better off getting rid of the 100-400 II and using the 70-200/2.8 IS II instead – two stops more light / less ISO and even more framing flexibility!

In your blog, you concluded, "When I compare the cropped photo with the extender photo, I tend to favour the cropped version. I think it has more sharpness, more contrast, better colours. At least the Canon extender does not give me any (significant) advantage. Therefore I will send the Canon extender back and stay with my unextended combo..." Perhaps a premature decision, since the extender combo might be capable of better sharpness and contrast than the bare lens if focused properly.

The moral: poor methodology yields bad data which lead to unreliable conclusions.

 

[mnclayshooter]

Once again, I have no photos available at my office to share, but I just received a 2x III extender last week on Friday... in the excitement to try it out, I popped it on the OLD 100-400 and walked to the back yard.

It's manual focus only at f11-ish... but... I can visibly see that it gave me considerably more reach and in a situation where I have time to really check the manual focus over, it will give me great results. I believe, without much way of knowing for sure, that I have one of the "sharp" copies of the old 100-400, as I've never really had concerns over sharpness that's often complained about on this and other sites. So, it's possible that there's no great argument to be had in my post. My only other prior experience with extenders is an old 1.4 tamron that have, that is honestly a hunk of junk, but it was a couple bucks at a garage sale, so meh... why not?

Back on topic: is it "better" than the bare lens... no. Is it bad enough to not ever bother trying it again, not in the least. I get fine detail with the outer edges of the frame getting a little soft or CA'd. I'd venture to say that it's not that important as usually we're after the bokeh in that area of the frame anyways. The CA is correctable in post from what I've seen so far.

All in all, I've been pretty happy with the limited experience I've had. Now, to complete the kit with the 300 2.8!

 

[AlanF]

I'll trundle these out again as they do show how well the 100-400mm II/7DII takes extenders, first posted about 20 months ago). There are 4 shots of a Swarovski target for testing their telescopes. It was 240 metres away and was 1mx1.3m. Top is the target. Below is the 100-400mm II on the 7DII at f5.6, next below + 1.4xTC III at f/8 and bottom +2xTC III at f/11 and liveview. Before saying they are awful read: at 400mm, the target, only 1mx1.3m is 240m away and fills only a 300x524 crop from the centre of the 20 mpixel image, at which size it is impossible to resolve most of the target because the details are below the Nyquist limit. The 2xTC III brings features above the Nyquist limit and you can resolve details in the round section of the target. Those images show that the 2xTC is able to increase resolution without badly degrading the image - you can easily read the the 6 10 14 18 in the circles, which are only smears without the TC.

 

[Loswr]

Those images show that the 2xTC is able to increase resolution without badly degrading the image - you can easily read the the 6 10 14 18 in the circles, which are only smears without the TC.

I quite like that fact that the extenders are actually allowing you to 'See the Unseen'.