Kenko Teleconverter Test

[Patrick Sparkman]

As noted in this thread http://www.birdphotographers.net/for...-III-Questions, we have been trying to find a teleconverter that works with the 5DIII and allows autofocus with the 800 f5.6. I purchased the Kenko 1.4X MC4, and the 2X MC4 to test, along with the Tamron AF 1.4X and SP AF 1.4X. The Kenko 1.4 MC4, and both Tamron Teleconverters will not work with the 5DIII and 800 combo. All three cause the camera to lock up which means that the camera thinks it is receiving an error from the lens. But the Kenko 2X works with the 5DIII, and reads out the correct aperture and focal length in the Meta data. It will even autofocus through the viewfinder in good light. Both the 1.4 and 2X Kenko's work fine with the 5DII and the 800, and the 1.4 works with the 70-200 2.8, and 70-300L. Overall a very frustrating experience, but I did want to do an optical test of my own to see how they stack up against the Canon Teleconverters.

Here is how I tested. I put two dollar bills on a flat piece of gaterboard, and used the LensAlign base to make sure the board was perpendicular to the camera. The pictures below show the results of the bare lens along with the Kenko 1.4, 2X, and both Canon 1.4III and 2XIII teleconverters. I used the 5DII for this test, since the 5DIII won't work with the Kenko 1.4. This shouldn't be an issue as these were shot at 200 iso, and they perform practically identical at that iso. Everything was of course tripod mounted outside in variable overcast conditions and I moved the target to get the same framing with each combination. I don't know the exact crop shown, but it was to get an idea of the relative sharpness. These images were imported into Lightroom 4 with the neutral camera profile and everything set to zero including sharpening and noise reduction.

As expected adding a teleconverter reduces sharpness and contrast. The Canon and Kenko 1.4's look almost identical in these crops, but the Kenko does vignette pretty significantly. If only it worked with the 5DIII, I think we would have a winner as it autofocuses well with the 5DII. In fact, it seems to work as well or better than I remember the 1DIV with Canon 1.4III combo. Things change when we get to the 2X however. The Canon 2X is a little worse than the 1.4's as expected but the Kenko 2X is significantly less sharp, especially at the edges. I don't think that I would be happy with the results from that combination.

So in summary, I am returning the Kenko 2X, and both Tamron's. One note: the Tamron's appear to be almost identical to the Kenko's and probably have the exact same circuit board. I am keeping the Kenko 1.4 though to use with the 5DII/800 combo as well as the 70-300 5DIII combo which I will talk about in another posting. Here are my conclusions in nice table form:

Canon 1.4 III - No autofocus with the 5DIII and 800 - (unless you tape the pins then it is somewhat unstable) Sharpness is good
Kenko 1.4 MC4 - Locks up 5DIII with the 800 but autofocuses with the 5DII Sharpness is good
Canon 2XIII - No autofocus Sharpness is pretty good
Kenko 2X MC4 - Works with 5DIII and 800 Sharpness is not very good

Please let me know if you have any questions, and if anyone else has any other ideas. I might order the Sigma 1.4. Does anyone have any experience with the Sigma 1.4 and the Canon 5DIII with Canon lenses?

[Roger Clark]

Please let me know if you have any questions, and if anyone else has any other ideas.

Patrick,
If you have a 7D, it would be interesting to show it with the 800 bare compared to the 800+1.4xTC+5D3, both imaged from the same location, same shutter speed:

7D + 800 no TCs, f/5.6 (wide open)

5DIII +800+1.4x wide open (f/8).

Detail will be close to the same, but I predict 7D will be better and faster AF.

Roger

[Patrick Sparkman]

I might be able to borrow one to compare image details, but the 5DIII will not autofocus with the 1.4 Kenko, and you have to tape the pins on the Canon tele and it hunts when it focuses. So for sure the 7D with autofocus better with the bare 800.

[Jason Holzworth]

Patrick,

Thank you for the test. This may be a dumb request but in your opinion how does the bare 800 w 5d iii look cropped and uprezed compared to the 5d iii + 800 + 1.4x ?

Jason

[Arthur Morris]

Patrick, Bummer on the no-AF news. I e-mailed the supposed Kenko reps but they remain mute. Your link at the original post linked back to the original post. I will repair it now :).

Any clue as to why the apparent differences in brightness/exposure in the test frames?

[Vikram Agrawal]

hi patrick, thanks for doing this test for everyone's benefit. Do you know if the kenko 1.4s work with the 5dIII and 400/5.6 combo? Apologies if your thread was targeted at the 800mm lens only.

[Doug Schurman]

hi patrick, thanks for doing this test for everyone's benefit. Do you know if the kenko 1.4s work with the 5dIII and 400/5.6 combo? Apologies if your thread was targeted at the 800mm lens only.

Yes, that combo works great.

[Vikram Agrawal]

Yes, that combo works great.

thx doug. i should also have asked, which particular version of the kenko works well with the 5d3+400/5.6.
many thanks

[Doug Schurman]

thx doug. i should also have asked, which particular version of the kenko works well with the 5d3+400/5.6.
many thanks

Kenko TelePlus MC4 AF 1.4X DGX Teleconverter

It also works well on the Canon F4 IS + Canon 1.4 extender to give you about 1000mm focal length.

[Patrick Sparkman]

Ok, so I am trying to shoot a "Dollar Bird" and it will only let me get to a certain distance then moves away. Also it is jumping around, so I need a reasonably fast shutter speed.

What to do?

1) Shoot with the 5DIII and 800 f5.6 bare, then crop and uprez to the size I want.
2) Shoot with the 5DIII and 800 f5.6 plus 1.4XIII teleconverter
3) Shoot with the 7D and 800 f5.6 bare.

The light was pretty bright, so I settled on a shutter speed of 1/2500 of a sec at the maximum aperture for each combo (f5.6 for the bare lens, and f8 with the teleconverter), this meant an ISO of 400 for the bare lens 5DIII and 7D and ISO 800 for the 1.4X teleconverter shot.

Here are the results shot using the same setup as above.

As Roger has told us many times, the 7D works better if you take into account the lower ISO advantage it gives in a fixed shutter speed situation. Also, the lower ISO gives the bare lens/(5D no teleconverter) some help in the cropped situation. It is still not as good, but not as far off as I expected. So if I can't get closer to the bird, I should pull out a 7D instead of adding the teleconverter to the 5DIII. If I can get closer, the 800 bare/5D combo would definitely win.

[Patrick Sparkman]

By the way, I just got the Sigma APO teleconverters in today and did a quick test. The 1.4X does not lock up the 5DIII/800 combo like the Kenko, and it reads out aperture properly and even shows a 1.4 Teleconverter attached in the focus micro-adjust screen. However, it also disables auto focus like the Canon 1.4. Taping the pins makes it work just like the Canon teleconverter. That is to say, focus oscillates around your subject before locking in. The 2X operates just like the Canon. I did a sharpness test, and the Sigma's are not as good as the Canons (I will show the results when I can post another picture here). So the Sigma's were a disappointment and will be going back.

[Roger Clark]

Hi Patrick,

Great test. I agree that if you can move close and fill the frame of a full frame (FF) sensor, the FF image will be better.

Your posted images show that the 7D +800 bare has the finest detail, even over the 5DIII + 800 +1.4x. The AF performance should be a lot better too. I assume these were all taken from the same distance.

Roger

[arash_hazeghi]

Hey Pat,

I think your TC is bad or it didn't focus right. The digitally up-rezed 5D image almost looks the same as the one with TC. (up-sampled actually looks better in the corner). That means the TC is actually reducing the detail in those areas! something is wrong

If my 1.4XIII was as bad as that I would trash it right away and just up-rez in PS for free. Or maybe the 800 does not take TCs well?

[Patrick Sparkman]

Thanks Roger,

The camera was never moved. It was on a tripod of course and I just changed the cameras/focused and took the shot. As for autofocus, since the 5DIII 1.4X combo oscillates with the pins taped, the 7D 800/Bare does autofocus much better. The 5DIII 800/Bare autofocuses the best of course.

I know Arash I was surprised as well. Perhaps I messed up the test, but I did use Live View focusing to reduce focusing error as much as possible. I will try again today and see if it repeats. Remember that the uprezed version gets the benefit of ISO400 vs 800 with the teleconverter. The Teleconverter shot does look a little sharper to my eye, even in the corner. But no doubt that the lower ISO helps the uprezed shot and the contrast is impacted by the teleconverter. Perhaps the teleconverter is bad, but it consistently out performs all the others I have tried and was equal or slightly better than the version II one I had when I got it. Also, I have gotten a lot of good non "Dollar Bird" shots with the combo, so this is really a pretty harsh pixel peeping test.

Below is the test with the Sigma Teleconverters. They are not as sharp as the Canon's, but probaby comparable to the Kenko's. They do show a lot more vignetting like the Kenko's. The 2X is quite a bit better than the Kenko 2X though at the edges, and they report correctly to the camera body. Personally I am sticking with the Canon's.

[Patrick Sparkman]

Well I did some more testing today, repeating the test described above at constant distance. As Arash noted, the results seemed strange before. Well with my repeat test, things make sense now. The 5DIII with the Canon 1.4X Teleconverter definitely looks better than the uprezed bare lens. Now also, the 7D images seem to match the 5DIII with teleconverter combo. Also included the 5DIII with the 2X compared to the 7D with the 1.4X. Those images were shot farther away than the 5DIII/1.4X and 7D bare lens images to give close to the same pixels in the crops. What does it all mean. Well for me the work was interesting, but I am right back where I started. With the 5DIII, and taped pins on the 1.4X Canon teleconverter. Now if you will excuse me, I am really tired of all of this pixel peeping!

[Roger Clark]

Hi Patrick,

Much cleaner. I do see a slight difference in the sharpness in the two 7D images you've posted (first 7D image is slightly sharper).

My take on the results:

1) Small focus errors really rob fine detail and contrast.

2) When one needs more reach, instead of considering adding a TC, consider changing bodies to a camera with smaller pixels. A camera with smaller pixels is like having a built-in TC without the loss of AF that a TC would give (and in some cases, without the complete loss of AF).

Roger

[Gary Kinard]

Well I am late to this party. But thanks Patrick. A lot of work and extremely interesting.
It really clears things up for me.