1 Attachment(s)
500mm+TC Image Comparisons
Note: I Reshot the graphics (now it's version 3) and have replaced the old one. I've also changed the related text. The remake was taken at 9 meters distance from the target.
To learn more about the issue of usable TC combos, I made a systematic comparison of the various combinations of TC's with the Canon EOS 500mm f/4 and a Canon EOS 1D Mark III body. The rig was mounted on a Wimberley II head/ Gitzo 3541LS and rebalanced for each lens combo and locked down. I shot from a sitting position with the tripod legs retracted except for the largest segment, and set the tripod legs to a wide stance to increase rigidity of the mount. I set the camera to a 2-second shutter delay with mirror lockup to avoid variations in my holding technique and reduce mechanical vibration. The IS was at setting 2 and was allowed to stabilize before pressing the shutter release.
My intention was to get an idea of the optimum combination for maximum reach and image quality. I varied the ISO settings in order to keep the shutter speed settings at 1/1000. Effective aperture varied depending on the combination used. All images were made at a distance of 9m from the target. I used a Speedlight 580 II in sync mode at +2/3 comp with flash extender.
My in-camera sharpness setting was level 3, and I used the "shade" WB setting. Each image was collected in RAW format and transferred to CS4, where exposures were equalized approximately, and images cropped to the URC of the $2 bill. The crops were pasted to a new canvas with identifying labels (upper row in the graphic). I enlarged copies of the 500mm, 500mm+1.4x and 500+2.0x images to dimensions equal to those of the 500mm+stacked converter image for comparison (lower row of graphic). Finally, for BPN posting, the graphic had to be reduced to 60% of its size.
Observations:
1. At their respective 100% crops (bottom row), the bare 500mm lens, not surprisingly, delivers the sharpest, but smallest image.
2. However, for prints enlarged to equal size (upper row), the 500mm+2.0+1.4 TC combination produces the greater resolution because it was formed by a greater number of pixels on the sensor. To increase the image from the 500mm the same dimensions as the 100% crop image of the stacked TC combo, the 500mm image had to be enlarged to 2x1.4x100 = 280%, considerably exceeding its pixel limitations, and resulting in the fuzziness you see here.
3. In the stacked TC combo, IQ does suffer from the increased number of optical elements. A close look at the right edges of the "2" also reveals an increasing amount of chromatic aberration -- evidenced by the red/green halo. This effect is progessively more noticeable as TC elements are added.
4. Adding TC's requires strong light to keep high shutter speeds for the 800, 1000, or 1600mm focal lengths because of the decreased effective apertures. Longer exposures combined with long focal length increases the risk of blurring from equipment shake and subject movement. Solid steadiness techniques are essential.
5. For an active subject (like an Acorn Woodpecker), take into consideration the difficulties of quickly relocating and refocusing with stacked TC's when the bird suddenly flies to a different tree.
6. On the 1DM3, all sensors will work with the 500mm and 500mm+1.4x combos, but only the central sensor will operate with the 2.0x or stacked TC combos. The AF response is much slower as well -- a disadvantage when working with active subjects.
If you're not yet ready to invest your life's savings in a 600 or 800, the best IQ is achieved by getting close enough to your subject to achieve the crop/composition you want in-camera (wouldn't that be nice!). When that's not possible, perhaps my $2 bill experiment will help you make a more informed decision about the compromises you'll need to make. I've learned a lot by just going through this exercise -- particularly that I have plenty of room for improvement in my sharpness techniques.
