The Howland Picture Pages: Digiscoping Experiments
Round 1 The Howland Picture Pages: Digiscoping Experiments
Round 1Page by Royce
Last update: $Date: 2004/02/23 02:02:25 $ (UTC)
As mentioned in more detail on the ScopeTronix Gear page, we use a lot of ScopeTronix equipment in our digiscoping setup. When used for viewing by eye, the 90mm Maksutov-Cassegrain spotting scope is great -- much brighter, sharper, and better chromatic quality than our first, entry level Bushnell 20-60X refractor scope. We also really like the results from the Nikon CoolPix 5700 digital camera when it is used on its own. When we have been able to get close enough to the subject in question (this can be challenging when shooting wildlife), we can get some great photos.
When the two devices are paired up for digiscoping, results are less than stellar, mainly in terms of the clarity and sharpness of the resulting photographs. Using the camera at a full native optical zoom (71.2mm) substantially overpowers the resolving capacity of the setup. Pictures come out very fuzzy, with a "soft focus" filter look. More on this topic below. There is also some chromatic distortion evident, possibly caused by the erecting prism which is used to keep the image from being inverted, a condition that would take some getting used to when tracking mobile subjects.
Attempting to zoom the camera back to increase sharpness introduces optical distortions and "doughnut hole" effects in the photograph at a certain point (which seems to vary), due to the central obstruction of the scope's objective lens. Zooming the camera back also begins to introduce vignetting. These effects occur despite the significant flexibility in positioning the camera on the scope by adjusting the STAK50 adapter or the MaxView 40 eye relief. Essentially, there is a very small window of camera zoom where the image is not terribly fuzzy, and where distortions and vignetting are not wrecking the shot.
Vignetting is a fact of life when digiscoping, of course, and it was the main issue I originally was expecting to have to overcome with the combination of the 90mm Mak and CoolPix 5700. But as it turns out, generally the highly adjustable nature of the camera, STAK50 adapter and MaxView 40 eyepiece mean that vignetting is not the most significant problem.
I posted some brief comments on my early experience to the Yahoo group "digiscopingbirds" (an excellent group, by the way), and received a few comments that are worth noting here.
This page attempts to show a collection of comparison images highlighting the impact of altering different variables in the digiscope setup, during my first round fumbling and testing. Unless otherwise noted, all of the digiscope photos on this page were based on the following configuration:
See the Digiscoping Experiments Round 3 page for tests involving different equipment. And see the Digiscoping Conclusions page for my evolving conclusions distilled from on-going rounds of testing and developing experience.
Bandwidth Warning: There are large images linked from this page! Use caution if you have a slow connection.
|
Digiscoped photo:
(click image for larger photo) |
This was my first digiscope test, shooting a Pileated Woodpecker from about 75m away (guesstimate) on January 18, 2003. The day was overcast and we were all under cover of trees, so lighting was poor as evidenced by the very slow shutter speed. This produces some blurring of the subject since the bird was in motion, but this sample picture taken from the test series shows the least motion blur. For a comparison of the same bird in a very similar location and shooting condition a few days earlier, see the link in the first paragraph. Clearly the digiscoped image is much lower in clarity. The digiscoped photo is also more dull, but I put that down to poor light combined with high magnification. The thumbnail image has been scaled down, but otherwise it and the original are unretouched in any way. Photo info:
Larger photo info:
|
||||||||
|
Original digiscoped photo:
(click image for larger photo) Retouched digiscoped thumbnail:
|
My second test of digiscoping came a couple of weeks later, on February 1, 2003. This time around the day was bright and sunny, and my initial subject was a Black Squirrel sitting fairly still way up on a sun-drenched tree, soaking up some rays. While he wouldn't turn to face me, there is enough detail to determine how clear things are (or not). I think the distance here was about 50m. (Hmm, possible need for a range finder occurs to me now.) While I don't have a comparison image for the squirrel showing ideal quality, again it is clear that the digiscoped image substantially lacks clarity. Coloration is better this time, compared to the Woodpecker above. I feel this is attributable to better light. For comparison purposes, I threw in a retouched thumbnail as well. In this image, I used Thumbs Plus to equalize the histogram and auto-balance the colors. I did a "digital camera cleanup" which boosts saturation and contrast. I then ran a few passes of gaussian unsharp mask. Finally I scaled down to thumbnail size. The result is a little more vivid, but nothing can really compensate for a significant lack of sharpness in the source image. As before, the thumbnail of the original image has been scaled down, but it and the original are otherwise unretouched. Photo info:
Larger photo info:
|
||||||||
|
Baseline photo:
(click image for larger photo) Digiscoped photo (8X):
(click image for larger photo) Digiscoped photo (1X):
(click image for larger photo) |
On the same day I did the Black Squirrel above, I tried some shots of a tree. The nice thing about a tree is it didn't move. (Unless you for seen the latest Lord of the Rings movie, you'll agree.) As with the squirrel, sunlight was decent. I was at a distance of probably 100m. The first image shows a baseline shot of the tree with 12X zoom on the camera (8X native plus a 1.5X teleconverter). The red box outlines roughly the area of the next image, which is a digiscoped photo with the camera at full 8X zoom. The third image, centered roughly where the second one was, is also digiscoped but with the camera at 1X zoom. The doughnut effect is triggered by two things. First is vignetting -- I did not adjust the STAK50 adapter to account for reduction in camera zoom moving the lens back from the scope eyepiece. Second is the central black hole, caused by the scope's central obstruction interfering with the camera's view in this particular arrangement. All photos are unretouched, except for scaling the thumbnails.
|
||||||||
|
Digiscoped photo (8X):
(click image for larger photo) Digiscoped photo (5.4X):
(click image for larger photo) Digiscoped photo (1.9X):
(click image for larger photo) |
Here is a different series of photos taken of the same tree as above, showing the effects of varying zoom levels. Each image is centered on the same point, and all other variables besides zoom level are held constant. As before the 8X digiscoped image is fuzzy and dim. The 5.4X image isn't much better for clarity or brightness. You can also see that vignetting is happening (again, I didn't adjust the STAK50), but it isn't actually all that bad yet. The 1.9X image is noticeably clearer and brighter, but you can see the spherical distortion and dark-ish area in the middle indicating that the scope's central obstruction is starting to interfere with the camera. In this case, I did adjust the STAK50 adapter so you can see that vignetting is not even as bad as in the unadjusted 5.4X shot. However, even though this got rid of most of the vignetting, the image is still pretty much useless due to the remaining distortions. All photos are unretouched, except for scaling the thumbnails.
|
||||||||
|
Digiscoped photo #1:
(click image for larger photo) Digiscoped photo #2:
(click image for larger photo) Digiscoped photo #3:
(click image for larger photo) Digiscoped photo #4:
(click image for larger photo) |
My next set of test shots occurred a few days later, on February 4, 2003. While setting up on another tree that was nicely lit, I realized there was a Red Squirrel sitting atop the tree. I was about 50m away and had not noticed him warming up way up there. I managed to snap off a number of shots after focusing in. I did not try altering any of the configuration between shots. The squirrel was very brightly lit as can be inferred from the shutter speeds on the first two shots, which are over 10 times faster than for any other full zoom digiscoped pictures I had yet taken. While color and brightness are pretty decent, clarity is still noticeably lacking. Pretty much beating a dead horse on that point, by now. The images reduce down okay, but they continue to have that 1960's Star Trek "soft focus" effect -- where is Nurse Chapel? Also, these shots are more noisy than previous ones, as they were all shot at ISO 400 equivalent. None of the shots were retouched aside from scaling.
|
||||||||
|
Digiscoped photo (8X):
(click image for larger photo) Digiscoped photo (4.6X):
(click image for larger photo) |
Here is the same Red Squirrel seen above, heading down the tree trunk after the tree-top sequence was finished. These two shots show the effects of different zoom levels on the camera. The second shot shows a bit of vignetting, still not bad although I don't think I had adjusted the STAK50 adapter. The clarity is perhaps marginally improved. The second shot is also a bit darker; I think spot metering would have worked better than center-weighted. The photos are unretouched, aside from scaling the thumbnails.
|
||||||||
|
Baseline photo:
(click image for larger photo) Digiscoped photo (8X):
(click image for larger photo) Digiscoped photo (5.4X):
(click image for larger photo) Digiscoped photo (4.6X):
(click image for larger photo) Digiscoped photo (2.9X):
(click image for larger photo) Digiscoped photo (2.2X):
(click image for larger photo) Digiscoped photo (1.8X):
(click image for larger photo) Digiscoped photo (1X):
(click image for larger photo) |
After finishing with the Red Squirrel shown above, I finally ran a series of shots on one of the stumps on the same tree, which is what I had set out to do in the first place. On this sequence, I simply altered the camera zoom level, keeping everything else constant. Light was still pretty good, so I kept the ISO setting at 400. The stump used for the digiscoping photos is the left-most one of the three in the baseline photo. The photo taken at 2.2X zoom on the camera is the last one in the series that doesn't exhibit the "doughnut hole" in the center of the image, but it does show a pattern of spherical distortion surrounding the center of focus. Both the 2.2X and 5.4X images show abnormally slow shutter speeds compared to the rest of the series, which I think can be attributed to clouds passing overhead temporarily blocking the otherwise bright sunlight on the stump. All of the photos are more or less equally bright, but clarity does improve as the zoom level decreases. Are we getting the message yet with this trend? I did not adjust the STAK50 adapter for any of these hots. Also, all of the images are unretouched aside from scaling down the thumbnails.
|
||||||||
|
Digiscoped photo (8X):
(click image for larger photo) Digiscoped photo (5.9X):
(click image for larger photo) Digiscoped photo (4.2X):
(click image for larger photo) Digiscoped photo (2.9X):
(click image for larger photo) Digiscoped photo (1X):
(click image for larger photo) |
Once I finished with the above tree sequence, still out on February 4, 2003, I looked for another subject for trying some different sequences. I found a large, uninhabited wasp nest that was falling into disrepair. I felt the textures and washed out colors might make for an interesting test. My observation point was about 25m out. By this time it was starting to cloud over and the wind was picking up. So in addition to losing the light, I was going to have a moving target to contend with. Nothing like a good challenge! In this series, all taken at ISO 400, I did adjust the STAK50 adapter to try to compensate for vignetting at the various zoom levels. Unfortunately I forgot to take a baseline photo so you'll have to try to imagine what a wasp nest the size of a soccer ball would look like in a clear, bright photo. However, the same general trend described before can be seen -- things get sharper as the zoom level decreases, but there still remains some lack of clarity even towards the end of the series. Notice the different form of optical distortion with the final image. Adjusting the STAK50 eliminated the "doughnut hole" but produced a different side effect that essentially ruins the image anyway. As before, the images have not be retouched except to scale the thumbnails.
|
||||||||
|
Digiscoped photo #1 (ISO 100):
(click image for larger photo) Digiscoped photo #2 (ISO 200):
(click image for larger photo) Digiscoped photo #3 (ISO 400):
(click image for larger photo) Digiscoped photo #4 (ISO 800):
(click image for larger photo) Digiscoped photo #5 (auto):
(click image for larger photo) Digiscoped photo #6 (ISO 100:
(click image for larger photo) |
For this next series on the same wasp nest, I picked a fixed zoom level (3.6X) and went through the range of ISO settings from 100 through 800 and then automatic. The ISO 800 image is noticeably noisy even in the thumbnail resolution, while the 400 is not bad and the others are fine. The final image is thrown in for comparison. It is an ISO 100 shot at 2.4X zoom, taken immediately after this series at 3.6X zoom. The images have not be retouched except to scale the thumbnails.
|
||||||||
|
Digiscoped photo #1:
(click image for larger photo) Digiscoped photo #2:
(click image for larger photo) Digiscoped movie:
(If this link doesn't work for you, you may need to update your version of Quicktime. Or just right-click and save to your local hard drive and try playing it from there.) |
I close off this round of experiments by circling back to the Pileated Woodpecker, possibly the same individual shown in the first photo of this page. February 8, 2003 I encountered a pair of these Woodpeckers, one of which I had a reasonably clear view of (despite some intervening twigs) as it sat high up in a dead tree working on some established and new holes. I snapped a few shots and a couple of minutes of video. The best two still images are shown here. These images were shot at 8X camera zoom, at a distance of probably 100m. Light conditions were a little better this time than the first time I digiscoped the bird, but still the day was partly cloudy. The bird was pretty stationary so there is little motion blur. As always, clarity suffers. I was going to start changing configuration to try other zoom and ISO levels when the Woodpecker was flushed by the other one flying overhead, and they both took off. As mentioned, I also took some Quicktime video which I think turned out acceptably considering the low resolution. Note that this file is very large -- 16MB. Apologies for the sea-sickness-inducing instability at the start, but the Nikon electronic remote doesn't seem to work when the camera is in Quicktime recording mode. So I initiated recording from the camera's trigger which created waves for 2 or 3 seconds. The exposure is also a little on the dark side. Since I don't have any Quicktime editing software, you're stuck for now. The still images are likewise unretouched.
|
Unless stated otherwise, all photos and other content on these pages is copyright © 2002 - 2004 by Royce Howland.
Please don't steal anything for your own purposes, in particular if you have a plan for making fabulous stacks of cash in so doing. If you would like to use a photo or other content you find here, or just want to contact us about something, please send email to royce@3ge.com.
