First, thanks to everybody who weighed in with their strategies to make the image in question. Some very good sleuthing on behalf of our readers and as you’ll see the answer is a combination of a lot of your suggestions and keen observations. The answer to how the life-sized condor print was made was based on biological knowledge, technological application at the DSLR level and use of sophisticated software.
Table of Contents
The Biology
Condors can fly up to 200 miles a day in search of a fresh carcass to feed on. When you fly that much, you need to keep every feather in tiptop shape. Wing sunning is a ritual performed every morning. While the actual purpose of the sunning behavior is debated, one theory is that the heat of the sun warms and softens the keratin in the bird’s feathers, allowing the feathers to take on optimal shaping. Whatever the purpose, I knew if I could find the right condor in the right setting I had a chance at my shot. Though condors sometimes rotate a bit while sunning, they generally orient their spread wings perpendicular to the sun and hold relatively still. If I got lucky, one would hold still long enough that I could snap multiple frames of it to stitch later.
Knowing the likely whereabouts of the condors was key. I scouted several possible roosting spots along the South Rim of the Grand Canyon, but never had luck with finding the birds there in the good morning light.
I tried for the shot again in Utah, but as the condors roost in trees there instead of on cliffs, I was shooting upwards at an angle I didn’t like.
Finally I hit pay dirt at Navajo Bridge. Decades back the original bridge crossing the Colorado River was replaced by a new bridge parallel to it and about 140 feet away. The bridges make ideal roosts for condors, much to the delight of onlookers at the walkover bridge.
I showed up before sunrise and Condor 354 was perched on a girder of the vehicle bridge. I could shoot from the pedestrian bridge and have him at eye level for the shot – ideal. Also the morning sun would be coming in at a nice angle. Next came the waiting – would 354 sun his wings while the light was still good? Condor 354 rewarded my patience (I’d been trying for months to get this shot) and spread his massive wings in the warm early morning light.
The camera stuff
To get the resolution I wanted I would not just need a high resolution camera, but would need to stitch multiple high res files together. This would require being close enough to the bird that just a portion of the bird would fill the frame. The South Rim locations I’d scouted seemed ideal as I could shoot a full frame DSLR with a 500mm lens and be close enough to fill the frame with half a wing or so. However the Navajo Bridge location required me to shoot from one bridge across a 140 foot gap to the other bridge. At that distance and angle the entire bird would pretty much fill a D810 frame.
Time to go for all the reach I could muster. I took a Nikon D7200 and put an 800mm on it with a 1.25x teleconverter. This gave me an equivalent focal length of 1500mm. The D7200 has a tighter pixel pitch than the D810, so would give me as much resolving power as any possible combination in the Nikon DSLR lineup (going with a lens mount adapter on a Nikon 1 mirrorless would open up some interesting possibilities). With the camera oriented vertically, the condor from head to tail would fill the frame and 4 frames across would be able to capture the open wings. I was prepared for 5 frames across, but 354 was at an angle so in the end I just needed the four shots and I would have enough overlap to stitch.
I shot on a carefully leveled tripod with a gimbal head and swept back and forth across the scene constantly in hopes of grabbing 4 or 5 frames where the condor didn’t move or the wind ruffle his feathers. I had previously practiced this with a D4s shooting at 11 fps in constant motion and discovered at such a great focal length I needed 1/5000 sec minimum not to suffer too much blur. But I also needed depth of field to get the width of the bird’s body all in focus so if I shot at f/11 and 1/5000, it would necessitate an ISO of 2500 in full direct sun, not a big issue with a D4s, but a noise monster on a D810 or D7200. As my camera to subject distance at Navajo Bridge limited me to the D7200, I shot at a lower ISO 800, then shot at 1/800 sec and f/11, stopping briefly between pans to minimize camera movement. It would then take two to three seconds to pan across the entire body. I panned both left to right, then back right to left, then back again and just kept shooting. The goal was to get one sequence with an attractive head angle, ample wing spread and of course no motion on the bird’s part between frames. On most passes it was obvious while shooting that there was movement or the head turned away. The sunning lasted four minutes in which I shot 100 frames. Time to download all those shots and try stitching a shot together.
The Software Solution
With 354 angled slightly, for a life-sized image the final dimensions would have to be 7.5 feet from wingtip to wingtip. Of all the sequences I shot, there were several with appealing head angles. Of those, I only got one to successfully stitch using the photo merge panorama function in Lightroom 6.
D’oh, that didn’t work.
Ah, that’s better. Subtracting the overlaps, this stitched image ended up as 10000 pixels wide. My goal was to print at 250 ppi so people could put their noses right up to the print and count the vanes in each feather. Sadly, 10000 pixels at 250 dpi would only yield a 40-inch wide print – why heck, even some puny Red-Tailed Hawk has a wingspan wider than that.
Enter Perfect Resize. Formerly know as Genuine Fractals, Perfect Resize is an upsampling program. It analyzes every detail in your image, then using cleverly written algorithms creates new pixels to bridge the gaps between the original pixels and create a much larger file. You can also upsample in Lightroom or Photoshop, but in my experience, image quality deteriorates more quickly with those methods than with Perfect Resize. Perfect Resize has multiple upsampling methods to choose from. I chose Genuine Fractals, entered 96 inches wide for my destination size, then experimented with the sharpness and smoothness sliders until I got the result I wanted. This is very time-consuming because with such huge files it takes several minutes for Perfect Resize to crunch the data every time you push a slider or enter a new parameter.
Because I started with a very large, very sharp file, I ended up with a great result. Note however, that if you attempt to use Perfect Resize with soft files, you’ll just end up magnifying all the defects and end up with mush. My final file ended up being over 2GB, but when zipped got down to 1.9GB which just made the cut to send online to the printing company.
Summing Up
What a crazy fun and challenging assignment. A lot of things had to come together to get this shot. First I had to find the birds (there are only 70 total in all of Arizona and Utah). Next I needed one to bust a good pose in good light close enough for me to get the frames I needed. I maxed out the capabilities of the camera gear I had available. Finally, I pushed the software as far as I felt it would go and still produce the result desired. I’m pretty amazed that over 75% of the pixels in the final file were made up by software, not captured by the camera. I planned this all out in advance, felt it would be possible, practiced the techniques I’d need, then when the time came, Condor 354 was a champ.
The Plight of the Condor will be showing at the High Country Conference Center in Flagstaff, Arizona from now until Jan 5th, 2016. The public is welcome so please drop by if you’re in the area. To learn more about California Condors and how you can support them please visit peregrinefund.org/condor.
All Content ©John Sherman
Wow! What a great shot! Congrats.
I love your answer and really loved that you used knowledge of the species to get the shot. Perfect Resize is great, been using it for a long time (when Genuine Fractals) with great results. Perfect Resize is reason enough to own Perfect Photo Suite.
So this wasn’t a 24k pixel wide image.
It was a 10k image that you upsampled, something you explicitly said you didn’t do early on in the comments section of when setting out this problem.
smh fam.
John, I can’t thank you enough for starting my Saturday morning with a wonderful story. Your ‘final’ picture of #354 screams out for the backstory. The awesome detail of the final result and your entire plan and workflow illustrate your commitment to both your profession and documentation of a species that may or may not be seen by future generations. Thank you.
I would love to see a fully grown and extended Condor superimposed over a typical human(with arms extended) to fully appreciate the majesty of this bird.
Thank you so much for this wonderful and detailed blog. I love this ‘how it’s made’ project. I am just starting out with bird photography and I found your explanation to be to informative and fascinating. Most of all, you have helped me to realize just how much work goes into a photo like this! Thanks for taking the time to share your process!
If you are interested in image interpolation and going further than the Lanczos and spline algorithms, there are a few content aware — though I use the term very loosely — resizing algorithms that were specifically developed to recognize lines, avoid stair effects, preserving local contrast and so on… For more information you can go here: en.wikipedia.org/wiki/…ge_scaling
Anyway, at some point an executable including most of those algorithms was written, it’s available here: code.google.com/archi…magefilter
It’s a no longer maintained project and the was still some work to be done UI-wise, but it’s functional and if you want to experiment with different resampling algorithm, you’re golden.
Good shooting but I would surely have edited out those godawful wing tags.
Yes, I know the condor is endangered and tagging is necessary to keep track of individual birds, but they do so detract from the resulting image.
I occasionally have the same problem with collared wild dogs in Africa – I dislike tags and collars so much I simply don’t take the shot as I am stickler for not altering the truth of the image and anyway it’s just too much trouble when I have other equally good or better shots.
The wing tags don’t bother me. Each condor is an individual and with so very few of them in the world, each individual has a lot to do with the survival of this species. They each deserve recognition and credit and if it weren’t for the name tags we wouldn’t be able to relate to them as individuals. Hopefully the recovery effort will prove successful at which time tags will no longer be necessary. At that point is the time to celebrate by showing pictures of untagged condors. It bothers me when other photographers clone out the tags as this not only alters the truth, but takes away from the celebration that hopefully will come in the future when condors achieve a self-sustaining population.
Without tags we probably wouldn’t know this is Condor 354. His story is pretty uplifting. When Condor 722, born two years ago in the wild, lost both her parents, Condor 354 took over the feeding and parenting of the orphan.
Because the tags make it possible to ID all the birds it creates the possibility of photographing an entire wild population of a species – a mission I’m currently on. The Arizona-Utah condor flock, at 72 members, is the largest of the wild populations and I have so far photographed 58 of them. Wish me luck and I hope number 72 will not be my Moby Dick.
Lastly, I have had the thrill of photographing a wild-born condor (Condor 766) yet to be tagged. S/he (sex is still not determined) has since been tagged. In a month or two there will hopefully be two more wild fledglings and another chance to get a legit untagged condor photo.
I think we are in agreement.
Tags on endangered species are a necessary (aesthetic) evil.
In this case the tag, as you say, contributes to the story but from a purely wildlife photography perspective it’s as photogenic as a wart.
….with apologies in advance to any particularly sensitive warts out there.
I think the tag actually tells a story. While I agree that tags and collars are unsightly in this case it contributes to the meaning and purpose of the picture. Just personal preferences, I suppose.
John, I’ve used Perfect Resize with good results, but I removed it from my computer when I learned it installed third-party anti-piracy software known as Nalpeiron, which has been known to cause system-level problems on both Windows and Mac. As a replacement, I now use the new “Preserve Details” option in Photoshop’s Image Size dialog. While it’s prone to oversharpen a bit, it does a generally fine job with this kind of extreme enlargement.
While I sympathize with onOne’s desire to protect their software, I can’t agree that their goal is worth compromising the stability of my computer. Since they showed no concern over that issue when I contacted them, I won’t ever use their software again.
Thanks for alerting us to that. I’ve had some slow down on my computer and I wonder if Nalpeiron is causing it.
John, I believe the main problem is that operating systems get updated, but there’s no automatic updating for Nalpeiron. So, even if it causes no problem when installed, you eventually get an incompatibility, with no idea what’s causing it.
If you just updated your Mac to El Capitan, it’s entirely possible that Nalpeiron might be running afoul of Apple’s new security measures. You could try Googling it.
On a Mac, the Nalpeiron file sits in a ProTec folder in /Library/StartupItems. Unfortunately, if you remove it, Perfect Resize won’t start. (And uninstalling Perfect Resize does not remove Nalpeiron. Once it’s installed, onOne washes its hands of the matter.)
It’s basically a virus, however well-intentioned. I’m amazed that any respectable company would use it.
Very nice…..Incidentally, I find that merging in Photoshop works better than Lightroom.
Jay, Photoshop surely stitches better, but it creates a TIFF file, while Lightroom creates a DNG. DNG is far better for post-stitch editing than TIFF, because it is still a RAW file. Wish Photoshop could produce a DNG out of stitched images, that would be awesome!
Thanks for a fascinating article and congratulations on accomplishing a phenomenal image. I doubt any of the other 24,999 wildlife and amateur photographers based out of Flagstaff could have done this.
2 questions:
1. Please explain the pixel pitch issue. Are there more pixels per square inch on a D7100 sensor than a D 810 sensor? Is that what counts for resolution?
2. How large were the image files? What kind of computing power did you need for this? Could your everyday computer handle it? If so, what have you got?
Thanks again for another stimulating article!
Yes, pixel pitch is how many pixels per inch – the D7200 has smaller photosites hence crams more ppi in for greater resolution than a D810. For comparison the D810 in DX mode creates 15 mp files, while the D7200 for the same format has 24 mp.
My final file was just over 2 GB. I processed it on a iMac 5K with 32 GB ram and it took several minutes to reprocess each change in Perfect Resize. This might tax some photographers patience, but when you wait a couple months for the bird to be in the right place at the right time, what’s another few hours messing around on the computer.