A Tale of Two Teleconverters

W.H. Majoros

1. INTRODUCTION

As the 2006-2007 winter came to a close, I began to plan for the arrival of spring, and the associated invasion of warblers into the North Carolina piedmont.  My primary concern was one of firepower----i.e., whether I had enough "big glass" to effectively capture the tiny, feathered gems that would soon be flitting about frantically in the upper stories of the local forest canopy.

In terms of "big glass" my current arsenal consisted of only two "big guns": the Canon 400mm f/4.0 DO, and my Orion 6-inch Maksutov-Cassegrain telescope. The former lens I previously reviewed here; this is the lens I affectionately call my "DODO" lens (DODO = Diffractive Optics for Digital Ornithology). The latter "lens" (actually an astronomical telescope co-opted for use as a terrestrial photographic lens) is reviewed here; this is the lens I previously referred to as the "Big Freakin Lens", and which I now refer to simply as "Hubble" (for obvious reasons). 

While "Hubble" has produced a number of decent bird photos (and very cost-effectively so), I've been thinking about replacing it with that beast known popularly as the Sigmonster----an 800mm f/5.6 lens manufactured by Sigma. The advantage of the Sigmonster would be its ability to autofocus when used at 800mm (or even at 1120mm if I were to upgrade to a Canon 1D Mark II professional body, which retains AF capability at apertures as small as f/8).

Since Hubble has a focal length of 1800mm and the DODO lens has a focal length of only 400mm, I was unsure of whether the 800mm of firepower provided by the Sigmonster would be enough. So I decided to spend a couple hours shooting wintering passerines with the DODO lens affixed with a 2.0x teleconverter, to see just what 800mm would buy me in terms of my ability to capture small birds at a 1.6x CCD crop factor. Some sample bird shots at 800mm (i.e., 400mm prime plus a 2.0x teleconverter) are shown later in this article.

For the sake of comparison, however, I decided to first conduct a "controlled" experiment involving a U.S. $20 bill and various combinations of teleconverters applied to the 400mm "DODO" lens, to first get a feel for the capabilities of my existing hardware. The next section of this article deals with these experiments.

2. TELECONVERTERS OUT THE WAZOO

For the first set of experiments I decided to compare the resolution and contrast of the Canon 400mm f/4 DO lens with and without either of the two Canon Extender EF II teleconverters. Since birds tend to move about a lot (thereby making controlled imaging experiments difficult), I chose to limit this first experiment to images of a $20 bill affixed to a rusty iron fence mesh.  In the next section I show some images of wintering NC birds with several combinations of teleconverters.


Figure 1: The Canon EF 400mm DO lens mounted on the Canon EOS 30D camera, with both Canon EF Extender II teleconverters (1.4x and 2.0x) attached. The tripod is an older model manufactured by Sanford & Davis for use with light telescopes.

For all experiments I mounted the Canon 30D digital SLR camera with lens attached on a heavy-duty Davis & Sanford (now Tiffin) tripod equipped with a simple fixed-position head. I have used this tripod for many years to support a 1250mm-focal-length astronomical telescope-----the Celestron C-5. It has proved to be a very reliable tripod for large-focal-length applications, and should therefore prove fully sufficient for these experiments. All experiments were conducted on the same day, at roughly the same time.

Figure 2 shows the results of the first experiments, which involved both a 1.4x and a 2.0x teleconverter. The bottom image is the baseline-----the Canon 400mm f/4 DO lens with no teleconverters attached. For the middle image I attached a Canon Extender EF II 1.4x teleconverter, whereas for the top image I used Canon's EF II 2.0 extender. Higher-resolution versions of these images can be viewed by clicking on the figure.

Figure 2: (top) Canon 400mm DO plus Canon Extender II EF 2.0. (middle) Canon 400mm DO plus Canon Extender II EF 1.4. (bottom) Canon 400mm DO.

All three images have been scaled by intregral factors to produce comparable image sizes. For the purpose of bird photography (next section), this is a very reasonable experimental protocol, since for bird photographs the size of the subject relative to the frame size is the most natural frame of reference for the viewer. Thus, the question which we wish to answer with these experiments is: when resized to match the natural resolution of the 2.0x image, how do the 1.0x and 1.4x images compare in terms of contrast and level of detail?

Although the bottom panel of the above (thumbnail) image appears in this scaled version to be superior to the other two, examination of the full-frame image (click on the figure to see the full-frame image) reveals that at this magnification the 2.0x image is supierior to the others. The 1.4x image appears especially fuzzy in comparison to the other two.

For the next experiment I decided to push the system even further, by stacking teleconverters. The Canon Extender EF II's can be stacked in only one ordering, due to their physical asymmetries----the 1.4x can be applied to the output of the 2.0x, but not vice-versa. I therefore decided to compare the resolution and contrast of the 2.0x setup to the "stacked" setup involving the simultaneous use of the 2.0x and 1.4x extenders, for an effective focal length of 400 x 2 x 1.4 = 1120mm. Figure 3 shows the results.

Figure 3: (top) Canon 400mm DO plus Canon Extender II EF 2.0, for an effective focal length of 800mm. (bottom) Canon 400mm DO plus both extenders, 2.0x and 1.4x, for an effective focal length of 1120mm.

From Figure 3 (click to see enlarged images) it can be seen that the use of stacked teleconverters (bottom panel) has not resulted in greater resolution or contrast. Thus, at this sizing of the raw images, it seems the 2.0x teleconverter is all that is needed. In Figure 4 (below) I have provided both images from Figure 2, but after 2x digital zoom. The top image (produced using only the 2.0x teleconverter) is again clearly superior.

Figure 4: (top) Canon 400mm DO plus Canon Extender II EF 2.0, for an effective focal length of 800mm. (bottom) Canon 400mm DO plus both extenders, 2.0x and 1.4x, for an effective focal length of 1120mm.

In Figure 5 (below) I have again provided both images from Figure 3, but after 3x digital zoom.

Figure 5: (top) Canon 400mm DO plus Canon Extender II EF 2.0, for an effective focal length of 800mm. (bottom) Canon 400mm DO plus both extenders, 2.0x and 1.4x, for an effective focal length of 1120mm.

In Figure 6 (below) I have applied a 4x digital zoom. Now it can be seen that the image taken at 800mm has "maxed out" in terms of its ability to be digitally zoomed, as signs of pixelation are now readily apparent. Thus, while the top image (using only the 2.0x TC) retains more contrast than the bottom image (using stacked 1.4x + 2.0x TC's), pixelation in the top image renders the photo less useful at this magnification. For a smooth image at this magnification we are forced to accept the image from the stacked teleconverters as superior to the image from the 2.0x TC only.

Figure 6: (top) Canon 400mm DO plus Canon Extender II EF 2.0, for an effective focal length of 800mm. (bottom) Canon 400mm DO plus both extenders, 2.0x and 1.4x, for an effective focal length of 1120mm.

Right about now you should be getting pretty sick of looking at $20 bills and are wondering what any of this really means for bird photography. Let's therefore move on to a less rigorous but more practical exercise.

3. A MORE PRACTICAL EXERCISE

After shooting my fill of Andrew Jackon's immortalized face, I appealed to the local avifauna for help with obtaining some test images. My first willing volunteer was a Slate-colored Junco (Junco hyemalis) lounging about in the lower branches of a tree roughly 70 feet away (Figure 7, below):
Figure 7: Slate-colored Junco (Junco hyemalis). Effective focal length: 1120mm (400mm + 2.0x + 1.4x). Distance was about 70 feet.


While this image seems fairly sharp (considering that I was using stacked teleconverters), it should be noted that in this case I've "cherry picked" the best photo out of a group of about 15. The others weren't as sharp as this one. Furthermore, this photo is just about maxed out in terms of its sharpness----any digital magnification beyond this would produce a noticeably less sharp image. Nevertheless, for web-based bird imagery, this configuration of stacked TC's and the resulting 1120mm of focal length are able to produce useful images at low frequency. If I were to use this combination very frequently I'd certainly find myself taking many more shots than normal, in order to be able to cull the 10% or so of useful images captured. Also note that the bird in the image was stationary; a moving subject would likely reduce the "keeper ratio" to less than 10%.

The Juncos lost my attention as soon as a White-throated Sparrow (Zonotrichia albicollis) dropped out of the tree and began to forage on the ground about 50 feet from me. As can be seen from Figure 8, the best image I was able to acquire from this subject was certainly acceptable, though in no way extraordinary. I again culled this image from approximately 15 photos, resulting in a "keeper ratio" of less than 10%.


Figure 8: White-throated Sparrow (Zonotrichia albicollis). Effective focal length: 800mm (400mm + 2.0x). Distance was about 50 feet.

For the purposes of comparison, Figure 9 (below) shows a White-throated Sparrow that I shot later in the day using only the 400mm DO lens----i.e., with no teleconverters. This bird was less than half the distance from me as the sparrow in Figure 8 (above), making the use of TC's unnecessary. I personally like the image in Figure 9 better than the one in Figure 8, though this is due to a combination of the greater sharpness and the more pleasing color of the scenery in the background.

Figure 9: White-throated Sparrow (Zonotrichia albicollis).  Focal length: 400mm.  Distance was about 20 feet.


Additional images taken later in the day using just the 400 DO lens, without any teleconverters, are shown below. The first is a Pine Warbler (Dendroica pinus), shown in Figure 10, which I was able to capture on my property:

Figure 10: Pine warbler (Dendroica pinus).  Focal length: 400mm.  Distance was about 18 feet.


Another image of the Pine Warbler, this time perched in a tree just putting forth its first blooms, is given in Figure 11 below:


Figure 11: Pine warbler (Dendroica pinus).  Focal length: 400mm.  Distance was about 25 feet.


The House Finch (Carpodacus mexicanus) shown below was also caught near my house with only the 400 DO lens (i.e., no teleconverters):


Figure 12: House Finch (Carpodacus mexicanus).  Focal length: 400mm.  Distance was about 15 feet.


Finally, a Chipping Sparrow (Spizella passerina)  again captured just by luck on my property:
Figure 13: Chipping sparrow (Spizella passerina).  Focal length: 400mm.  Distance was about 12 feet.


From Figures 9-13 it should be clear that the Canon EF 400 DO telephoto lens can perform superbly without teleconverters when the subject is sufficiently close that little digital zooming is later required. For birds at greater distances, however, one should probably resort to the use of teleconverters (possibly stacked) or to a larger-focal-length prime lens in order to capture images of acceptable quality.

4. CONCLUSION

The dollar bill experiments have now convinced me of two things: (1) that the aggressive use of teleconverters----even stacked teleconverters---in favorable conditions and for small subjects at a distance can be beneficial when larger focal lengths are not otherwise available, and (2) that my Canon EF 1.4x teleconverter is very probably defective. I have read reviews of this TC stating that image quality was only very slightly degraded by its use, whereas the images from Part I of this report show that the use of the 1.4.x TC substantially degraded image quality. Variability in the collimation accuracy of this teleconverter has been observed by others. I am now even wondering if the Canon EF 300L f/2.8 lens which I had been using exclusively with this teleconverter might not have been as deficient in sharpness as I had originally thought.

That the stacking of teleconverters can result in acceptable image quality, even at low "keeper" ratios, comes to me as somewhat of a surprise. Unfortunately, the necessity of manually focusing this configuration likely renders this solution impractical for warbler photography, since warblers seem always to be moving about frantically in search of insects hiding in the shady places just under the forest canopy.

The good news is that I received word today from UPS that the Sigmonster I recently ordered will be arriving at my front doorstep tomorrow. For my next review I hope to report on the virtues of this monstrous 800mm prime lens.
back