Bare Lens vs. FE 1.4x & 2.0x Teleconverters

My current rig includes the Sony a7R III, the FE 100-400mm f/4.5-5.6 GM OSS and the FE 1.4x and 2.0x teleconverters. I've had the bare lens and the 1.4x since September and I've been pleased with both image quality and autofocus performance, with no noticeable slowdown in AF.

 

I received the FE 2.0x teleconverter last week and was very happy to report that it autofocuses on both the a7R III and a9 bodies, with very little, if any flow down. I thought, holy cow, if I can shoot this thing in good light, I've got 800mm of reach in a very small package.

 

In early testing, I was very happy to run across a kestrel, in great light, at a reasonably close range, but I was very disappointed that none of my shots were as sharp as I'd come to expect with my bare 100-400m and even with my 1.4x teleconverter attached. (My Canon equipment was also sharper). See example:

 

Untitled by David Stephens, on Flickr

 

I decided to test, in good light, using a static subject. I took shots with the bare lens, all from the same distance, and then cropped the bare lens shot and the 1.4x teleconverter shot to match the 2.0x shot. So, the 2.0x is uncropped, the 1.4x is cropped a bit and the bare lens shot is cropped 50%. Here are the results:

 

Bare Lens

Bare Lens - 3892x2596p to approximate 2.0x teleconverter by David Stephens, on Flickr

14x Teleconverter

1.4x teleconverter - 5481x3656p Crop by David Stephens, on Flickr

2.0x Teleconverter

2.0x Teleconverter - Uncropped 7952x5304p by David Stephens, on Flickr

 

I cropped further so that you can study detail. Keep in mind that I took these all at the same distance, so the DOF shrinks as the focal length goes up. To get the DOF equal, I would have needed to back up from the bare lens to get the perspective and magnification the same. So compare in-focus to in-focus. I want to use real, in the wild, elements, to see realistic detail. BTW, resolution on the camera is 42.4mp. Magnified results:

 

Bare Lens

Bare Lens - Cropped to 1955x1304p by David Stephens, on Flickr

1.4x Teleconverter

1.4x teleconverter - 2737x1825p Crop by David Stephens, on Flickr

2.0x Teleconverter

2.0x Teleconverter - Cropped 3912x2610p by David Stephens, on Flickr

 

Sorry for the flaw in my methodology. My focus was on the front of the bunch on top.

As expected, the bare lens is best, even with the severe cropping needed to get the subject the same size. The 1.4x is close and acceptable. I can't accept the 2.0x's performance and will send it back.

Now that I've seen the DOF issues caused by my methodology, I'll try again, later today, using my tape measure to adjust my distance to the focal length of the lens. So, for example, I'll shoot 400mm at 4-feet and 800mm at 8-feet.

I'm not surprised at all - a 2.0x extender on a variable aperture zoom lens is not a combo I expect to deliver outstanding results. A 1.4x will already affect the performance of the base lens but results can still be quite acceptable or even almost imperceptibly degraded. For years I have used Nikon's 1.4x and 1.7x (mostly the latter) until I had to realize that especially the 1.7x was not doing the AF-S 300/4 any favors (the 1.4x falls in the "still quite good category). I never even bothered with the 2.0x having seen samples that clearly demonstrated how bad things can become.

 

I never bother with a TC when using a zoom lens. In fact, I now would consider TCs only with high-end superteles (400/2.8, 500/4, 600/4) or when there is only an occasional need for one.

I'm not surprised at all - a 2.0x extender on a variable aperture zoom lens is not a combo I expect to deliver outstanding results. A 1.4x will already affect the performance of the base lens but results can still be quite acceptable or even almost imperceptibly degraded. For years I have used Nikon's 1.4x and 1.7x (mostly the latter) until I had to realize that especially the 1.7x was not doing the AF-S 300/4 any favors (the 1.4x falls in the "still quite good category). I never even bothered with the 2.0x having seen samples that clearly demonstrated how bad things can become.

 

I never bother with a TC when using a zoom lens. In fact, I now would consider TCs only with high-end superteles (400/2.8, 500/4, 600/4) or when there is only an occasional need for one.

 

Dieter, I generally agree; however, Canon's EF 2.0x TC-III has excellent IQ and I don't hesitate to use it on my EF 500/f4 II. The prior version of Canon's TC was a bust and I wouldn't use it in any circumstances.

 

Canon and Sony have exceptional 100-400mm zooms. You can see from my examples that my Sony is excellent at 400mm and almost as good with the FE 1.4x teleconverter attached. The only reason I take it off is when there's advantage in going from f/8 to f/5.6. With the Canon, even the 1.4x TC-III slows autofocus noticeably, where I can't detect a slow down or inaccuracy with my Sony. I would use my EF 1.4x TC-III on my EF 100-400mm quite often, BUT it did noticeably slow down AF, so I mainly used it in good light and not for fast BIF. (An EOS 1DX-II would fix that AF issue).

 

With my Canon 500/f4 II, I keep the EF 1.4x TC-III on almost all the time, yielding a 700mm that's worth carrying, in addition to my Sony 100-400mm, which is 140-560mm with the 1.4x teleconverter. I take the 1.4x off the Canon only when I need f/4.

 

Still, IQ trumps focal length. You're better off cropping an image with superior IQ than sticking an inferior teleconverter or lens on an excellent body. Also, pixel-density matters. Shooting with the Sony a7R III, at 42.4mp, I have the luxury to crop and still have plenty of pixels on the subject. Crops sensors for birds and wildlife make sense, because most have even high pixel-density than the a7R III. In good light, the superior dynamic range of the Sony full-frame will not matter, for most subjects.

 

Here's one with the a7R III, the G Master 100-400mm and the FE 1.4x teleconverter that tests DR, resolution and image quality of the rig:

 

Magpie Struts by David Stephens, on Flickr

 

I think it passes with flying colors. Anyone that's tried shooting a magpie in full sun knows what a difficult subject that they can be.

As promised, I tried again today, using a slightly different approach. Rather than taking all the images at the same distances and then cropping the shorter focal lengths to equal the subject size of the 2.0x teleconverter image, I equalized the distance in relation to focal length. Specifically, I took the bare 400mm lens image at 4', the 1.4x teleconverter image at 5' 7" and the 2.0x teleconverter image at 8", using a construction tape measure. The subject sizes in the images were still a bit different, but, at least, the depth of field was more comparable shot to shot vs. keeping the distance constant. I used the 42.4mp Sony a7R III for all three shots. These are handheld, so I took several safety shots of each and made sure to use the sharpest of the groups. Of course, stabilization was engaged. All are at ISO 100, f/11, 1/200-sec.

 

Bare Lens

Bare Lens by David Stephens, on Flickr

 

1.4x Teleconverter

1.4x Teleconverter by David Stephens, on Flickr

 

2x Teleconverter

2.0x Teleconverter by David Stephens, on Flickr

 

I'm very happy with the bare lens and find the 1.4x teleconverter not far behind. I have no reservations about using the 1.4x for critical subjects like birds and furry mammals.

 

The 2.0x teleconverter has serious problems in my testing. First, I think that the varying images sizes has to do with the focal lengths being "nominal" readings. For instance, I've read that actual focal lengths for some 100-400mm lens may actually be closer to 105-385mm. I've never read the same things about teleconverters, but wouldn't be tremendously surprised if the Sony isn't actually closer to 2.1x. That said, there are sharpness issues with the 2.0x. Contrast and color are acceptable. The lens seemed to be back-focusing, being in focus behind the plane where the focus was taken. Despite taking the 800mm image from twice the distance of the 400mm image, the depth of field seems quite shortened. Look at the very fine bristles at the top of the bloom to see sharpness that makes me think that the lens is back focusing.

Edited December 12, 2017 by dcstep

The purpose of a tele-extender is to have a pocketable device which allows you to crop in the camera rather than in processing. IMO, image quality in the samples you provide is adequate for this intended use, and comparable to or superior to results I get with my Nikon gear. I don't use the 2.0x as often as the 1.4x, but due to the largesize and speed loss. However, I can fit both into one lens pocket in my bag, which would be impossible if replaced with a 600 or 800 prime, and not cost-prohibitive.

 

IS notwithstanding, I would suggest using a tripod when testing for sharpness. Otherwise you are lumping the effectiveness of IS and vagaries of focusing with performance of the optics.

 

The DOF should not change with focal length, using the same relative aperture while adjusting the distance for the same field of view. The same image size at the focal plane yields the same DOF relative to the subject. This is obscured by the fact that the background is magnified relative to the subject, due to the narrower field of view. On the other hand, if you shoot from the same distance and crop to the same field of view, the DOF decreases proportionately. Cropping to demonstrate a 2.0x extender, you also lose half the resolution, reducing the 42 MP sensor to 10 MP.

 

In conclusion, your test does not dissuade me from using tele-extenders on a zoom lens, including my current Sony 70-200/4 nor the 100-400 I hope to acquire in the near future. I don't see a fast 400 prime in my future, unless I buy a mule to carry it and have an extra $12,000 to spare. Sometimes good enough is good enough.

Lots of people are in Ed's camp, that is, "I don't care about a decline in IQ, I just want more reach."

 

I'm in the camp, "Image quality trumps focal length." I will add, within limits. From testing, I know that the IQ of my Sony FE 1.4x teleconverter is very slightly below my bare lenses, but getting more pixels on the subject, without cropping, is worth the very small trade off. Somewhere between my 1.4x and my 2.0x teleconverter, the image degradation is too much for me to accept. I'm not going to carry it in my bag, when I know that I can simply crop to a better image than I would with 2.0x, even with a 50% crop or greater.

 

Wildlife photographers, shoot lots of feathers and fur, which tend to display any weakness in optics. When you get used to the very best IQ, such as with Sony's G-Master lenses or Canon big EF super-telephotos, poor IQ just jumps off the page.

 

By showing examples, I was attempting to demonstrate to others and they can decide for themselves what's "good enough."

 

By the way, I hand hold these test shots, with IS and high shutter speeds, to test the same way as I shoot. I rarely use a tripod.

100-400mm plus 2.0x teleconverter, handheld:

 

Broad-tailed Hummingbird by David Stephens, on Flickr

I should comment further about using the FE 2.0x teleconverter on both the a9 and a7RIII, in combination with the FE 100-400mm f/4.5-5.6 GM OSS. The autofocus is excellent on either body, at the required f/11, if you're going to shoot at 800mm. The handheld hummingbird shot was in excellent light. In this condition, the AF is fast and accurate, with no noticeable slowdown from the bare lens. It's important to select the right AF mode. For the shot above, I'm in Lock-on Wide mode, which uses all AF points. For a hummingbird in a bush, I'd use Lock-on Flexible-Spot or Lock-on Center. For birds in flight, it's Lock-on Wide, or Lock-on Center, depending on the background. The shot above was at ISO 1600, to get the SS up to 1/3200-sec, in case there's flight.

 

When it's overcast, things slow down a bit with the TC, but no more than my Canon 5D4 at f/8 in the same conditions. Things only become really problematic at conditions where I'm at ISO 20,000, or above, which is also a problem for the Canons. Remember, I'm at f/11, not f/8. The best choice for really low light is to put my 500/f4 on the Sony and shoot at f/4 or f'/5.6, when the 1.4x TC is attached.

I've got a little more to add, after a few thousand more shots. The a9, with the FE 2.0x teleconverter on the FE 100-400mm has a much higher keeper rate than the a7RIII. Even in good light, I had times where the AF wasn't spot on when using the a7RIII, even in a long series of shots, with chances for the AF to self-correct. The a9, OHOH, will self-correct and continue to improve focus, if it's not exactly right at the start. Also, the speed of acquisitions seems much faster with the a9. Sooo, if I'm shooting BIF, even slow birds, it's the a9. If I'm shooting perched birds, I prefer the a7RIII, for its higher resolution and DR.

Using the a9, this morning, with the 100-400mm and the FE 2.0x teleconverter, I had several shots to choose from. Shooting at f/11, the camera/lens quickly grabbed focus and kept focus, yielding me several good shot. 20-fps gave me plenty of wing positions to choose. I liked this best:

 

Red-tail Hawk Flies By by David Stephens, on Flickr

Good example of how 20 fps is useful. The 100-400 moved higher on my wish list too.

Another example of the usefulness of 20-fps:

 

Broad-tail Hummingbird Takes Off by David Stephens, on Flickr

 

Shooting at f/11 (because of 2.0x teleconverter on an f/5.6 lens) shows limitations of the rig. I'd rather be at 1/4000-sec. and like ISO 800, instead I'm at ISO 3200 and 1/2500-sec. The slight wing blur is pleasing. Interesting, the a9's performance at high-ISO makes noise a very minor issue. (I did apply DxO PhotoLab's PRIME noise reduction, which is excellent). Still, this is not ideally sharp, due to SS.

My current rig includes the Sony a7R III, the FE 100-400mm f/4.5-5.6 GM OSS and the FE 1.4x and 2.0x teleconverters. I've had the bare lens and the 1.4x since September and I've been pleased with both image quality and autofocus performance, with no noticeable slowdown in AF.

 

I received the FE 2.0x teleconverter last week and was very happy to report that it autofocuses on both the a7R III and a9 bodies, with very little, if any flow down. I thought, holy cow, if I can shoot this thing in good light, I've got 800mm of reach in a very small package.

 

In early testing, I was very happy to run across a kestrel, in great light, at a reasonably close range, but I was very disappointed that none of my shots were as sharp as I'd come to expect with my bare 100-400m and even with my 1.4x teleconverter attached. (My Canon equipment was also sharper). See example:

 

Untitled by David Stephens, on Flickr

 

I decided to test, in good light, using a static subject. I took shots with the bare lens, all from the same distance, and then cropped the bare lens shot and the 1.4x teleconverter shot to match the 2.0x shot. So, the 2.0x is uncropped, the 1.4x is cropped a bit and the bare lens shot is cropped 50%. Here are the results:

 

Bare Lens

Bare Lens - 3892x2596p to approximate 2.0x teleconverter by David Stephens, on Flickr

14x Teleconverter

1.4x teleconverter - 5481x3656p Crop by David Stephens, on Flickr

2.0x Teleconverter

2.0x Teleconverter - Uncropped 7952x5304p by David Stephens, on Flickr

 

I cropped further so that you can study detail. Keep in mind that I took these all at the same distance, so the DOF shrinks as the focal length goes up. To get the DOF equal, I would have needed to back up from the bare lens to get the perspective and magnification the same. So compare in-focus to in-focus. I want to use real, in the wild, elements, to see realistic detail. BTW, resolution on the camera is 42.4mp. Magnified results:

 

Bare Lens

Bare Lens - Cropped to 1955x1304p by David Stephens, on Flickr

1.4x Teleconverter

1.4x teleconverter - 2737x1825p Crop by David Stephens, on Flickr

2.0x Teleconverter

2.0x Teleconverter - Cropped 3912x2610p by David Stephens, on Flickr

 

Sorry for the flaw in my methodology. My focus was on the front of the bunch on top.

As expected, the bare lens is best, even with the severe cropping needed to get the subject the same size. The 1.4x is close and acceptable. I can't accept the 2.0x's performance and will send it back.

Sorry, but those images are not sharp.

Sorry, but those images are not sharp.

 

See first post: "In early testing, I was very happy to run across a kestrel, in great light, at a reasonably close range, but I was very disappointed that none of my shots were as sharp as I'd come to expect with my bare 100-400m and even with my 1.4x teleconverter attached. (My Canon equipment was also sharper)."

 

Sorry, but it's a long story, requiring reading the thread, noting the change to a second FE 2.0x and moving from a7RIII to a9.

Not moving (much). No problem with a9/100-400mm/2.0x combination:

 

Bee On Thistle by David Stephens, on Flickr