Ultra Wide Panos – CS2 or earlier technique
Just how wide is wide enough? You can go back nearly to the beginning of photography and find our fascination with wide capture, starting with Joseph Puchberger in 1843. He patented a hand crank driven swing lens panoramic camera that used curved Daguerreotype plates 19 to 24 inches long (that must have been a painful process). Ultra wide lenses such as the Nikon 15mm have been around for a few decades. The 15mm only captures 110°. Ultra wide cameras like the Widelux have been around just as long. The Widelux with its rotating lens captures 140°. Panorama cameras hit their apex in 1994 with the Fuji GX617. While anything greater than 100° (on the horizontal plane) is considered a panorama, it’s when we start hitting 150° that we feel we’re looking at a wide photograph. But what about the vertical plane?
I bought my first Nikon 15mm in 1980 and I never left home without it. A noted photographer commented that I only saw the world through an ultra wide. It wasn’t until I was shooting at Glacier Point in Yosemite Nat’l Park and I traded the 15mm with another photographer shooting with a 13mm that I really got a feeling for wide. But no matter what I shot with, there was always one thing missing when shooting with ultra wides or pano cameras for me. The height!
We have a peripheral vision of 180° and a binocular vision (important for depth perception) narrowing that down to 140°. These measurements are again on the horizontal plane. While this information is easy to obtain, our vertical vision is not widely known or documented. We have a total of 135° of vertical vision. This consists of 60° up and 75° down and when taking binocular vision into account it’s reduced to only 95° of vertical peripheral vision. Shooting with Nikon’s 10.5mm with its 180° horizontal coverage, we come close (but no cigar) to capturing the one element that brings visual depth to a wide vista. Vertical height.
I’ve shot my fair share of panos over the years, using probably every special easel and stitching software available. While some of the panos are “cool,” none truly leave me or others with the feeling that you’d get standing next to me when I took the photo. Panos are typically along the horizontal plane, ignoring the vertical plane, which is why they are lacking the giant impact. Bruce Dale had an excellent article on panos in a recent Nikon Capture magazine and while he is still only working on the horizontal plane, his technique got me to thinking. Not too long after that I was looking at my first ultra wide pano print, coming off the Epson 7600. Here are the mechanics of taking your own ultra wide panos.
The Ultra Wide Tools
Creating ultra wide panos (UWP) can be a really time intensive, high skill adventure. You can minimize both time and skill by having the right tools (which is different from owning all of the possible tools). There are lots of tools you can buy to do panos, some being really cool and some being really hokey. I’m a bare minimum, but first class tool kind of guy. With that in mind, you can accept these as just suggestions, knowing that I’m making just these few tools work for me. You’re certainly welcome to buy others and more while knowing you can be just as successful with this bare minimum.
The most important tool is the lens, that’s true for any photographic enterprise. But it’s especially true with ultra wide panos. We’re going to end up in the digital darkroom, bringing together six individual images. We want to make that process as fast, seamless and painless as possible. This is accomplished by using the unique characteristics of PC lenses. Yeah, you can do it with non-PC lenses, but you’d best have one heck of a PhotoShop knowledge library to draw on because you’ll need it!
We need the very essence of the PC lens, its perspective control. There will be vertical lines in our image, somewhere. If we have any tilt to the camera, these lines will converge no matter how good of a lens we shoot with. This wreaks havoc on us in the digital darkroom when we bring our six images together. We must set up our camera so the film plane is parallel with our subject. By doing so, our vertical lines won’t converge and our work in the digital darkroom is a snap! With our camera set up parallel, we can shift the PC lens to frame the subject as desired. We also depend on the rotating nature of the PC lens for taking the top and bottom panels of our UWP (more on that in a second). Nikon made two wide PC lens focal lengths, 28mm and 35mm. I have and use both. Nikon still makes the 85mm PC lens and I haven’t tried it yet for UWPs, but I’m sure it would work perfectly.
Finding Nikon PC lenses is a little bit of a trick (not so for Canon). I got my oldies but moldies from eBay. The trick in buying them at eBay is twofold: don’t be impatient and send them in once you’ve bought them. Don’t get sucked up into the auction fever at eBay, set a mental limit to the maximum you’re going to pay and just wait for the right one to come along at that price. You’ll often see the 28PC for example going for $500-900, yet I got mine for less than $300. It took a little while, but it was more than worth the wait. Don’t buy these lenses new if you can find them because that’s around $1200! Keeping your investment small, makes taking UWPs just that much sweeter.
Once you’ve acquired your used PC lens, send it into Nikon to have a CLA, clean, lube and adjust performed. The tracks on the PC lens often take abuse, causing them to be out of alignment. Having them cleaned and aligned makes all the difference in the world in taking sharp images. If your PC lens is so loose that the front element moves on its own from the pull of gravity, send it in! I sent both my 35mm & 28mmPC lenses in and it made a world of difference.
The camera body you use can make your shooting easier. Shooting digital makes the whole process just so much easier and more fun that I can’t imagine shooting UWPs with any other medium. My personal preference is the D2H. The main factor is its 100% viewing. That makes it possible to work within really tight tolerances to maximize the finished image size. This is one time the 8fps just isn’t needed, in fact I shoot on S firing mode when doing panos. If you shoot with a 96% viewfinder camera like the D1X or D70 (the other two bodies I’d create UWPs with), I’ll discuss in a moment how to make them work in your favor with their less than 100% viewfinders.
Next you need a nodal plate. Not a noodle plate, a nodal plate. What the heck is that you ask? It’s an incredibly important tool that in all honesty is no more than a metal bar that holds your camera away from the tripod head. I’ll explain how you use it in the next section, but the nodal plate is what we use to pivot on the lens’ nodal point. We need that so when we pan on the tripod head to take our panels, we’re panning on the lens’ nodal point.
The nodal plate is attached perpendicular to the film plane. In other words, it sticks out in front of the camera body. I use the Really Right Stuff MPR-CL plate, which is an off-the-shelf, custom nodal plate. There are other plates that do the same job, but are either not long enough for every lens or aren’t the cleanest design. The key is finding a plate long enough to give you enough length to make the various adjustments for your lens. If you have just one lens, you can get a plate just long enough for that one lens and you’re all set. The MPR-CL is 155mm in length with an Arca Swiss quick release clamp on one end. This is cool because there will be times when you’ll want to do “regular” photography while shooting a pano. Being able to quickly take the body off and just shoot normally without messing with your tripod setup is a plus.
Why do we need this nodal plate? We must work literally around the lens’ nodal point. A lens’ nodal point is where the light paths cross inside the lens barrel. The nodal plate permits us to set up the camera/lens so the lens’ nodal point is directly over the center pivot point of the tripod head. You could say we want to work on the optical pivot point of the lens. We want to rotate on that point as we pan to make our pano panels because it prevents lines from converging (making image assembly a snap). Does finding that nodal point let alone execution scare the bejiggers out of you? No worries, this very complicated sounding process is very simple to accomplish, with the right tools. It is imperative you do work on that point, so you can’t skip on this integral piece of equipment.
As a quick review, we use the PC lens and nodal plate to keep our lines all nice and neat. The PC lenses keep lines from getting thicker or thinner and the nodal plate keeps those same lines from bowing in or out. If we have a tree in the frame, without these two essential tools, the tree will be much wider at the bottom of the frame than the top. And depending on which panel it’s found, it will bow left or right. Neither of these attributes are ones we want, which is why we have these two tools.
The next piece of equipment is real simple, levels. You should have a built-in level on the top of your tripod and/or tripod leg. In addition, I highly recommend a level in the hot shoe of the camera. I use the Hama Camera Spirit Level. It’s a two way level for pitch and yaw. I also have a carpenter’s level resting on top of the nodal plate. I also refer to the level built into the ballhead, the last and most important tool of all.
The tripod head you use will either make your life a breeze or a nightmare (or some shade in between) when it comes to making UWPs. I’m speaking from experience. I tried a number of heads and while they worked, I lost some of my image coverage in the assembly because of the imprecision in the head (not to mention some challenges lining up certain elements in the photo). I have found a head though that is so precise there are no wasted images.
The Really Right Stuff new BH-55 head is the head for creating UWPs (and every other use as well). The key ingredient is the fact that the quick release clamp platform and the base of the BH-55 are parallel when leveled. This is essential because as you rotate the camera to take your series of panels, if there is any slop in this alignment, even the teeny-tiniest, your final images will not perfectly line up and you’ll lose image content. There is no way I can emphasize this enough and not until you start to assemble your first UWP will you know exactly what I’m talking about.
With all the correct tools, you can create your own UWPs with real ease. Our goal we’re working towards is matching the final composite image with a view that is similar to all one would see if standing at the place where we took the image. This means replicating our peripheral view. If shooting with a 28mmPC, the final 6 image UWP has the approximate view of 200° x 85° and with the 35mmPC, 165° x 75°. This more than matches our field of vision and brings a power to your images that goes beyond words. That’s our goal, isn’t it?
With these few and simple, yet the best pieces of equipment, the fun can begin!
The Ultra Wide Technique
Ultra wide panos are one of the grandest ways I’ve found to communicate in one image the grandeur I’m seeing. But just because you have a wide vista before you doesn’t qualify the view for the making of an UWP. All the rules of ultra wide photography are still in force when framing up an UWP, the main one being, to exclude all the junk and include all the elements that tell the story. How do you know if the vista in front of you will make a great ultra wide pano?
My general rule of thumb is that 3 of the 6 images are strong photographs without being part of the ultra wide pano. If I were to be just shooting normally, 3 of the 6 views in the UWP would be good individual images (why I like the RRS MPR-CL plate as I shoot normally a lot while creating UWP). The placement of these 3 good images in the overall UWP is not important, just that the strength of these 3 images be such that they carry the viewer’s eye through the whole image.
While it’s part of what makes every great image, LIGHT is incredibly important in the UWP. You’ve got a lot of area to cover, which raises the potential for lighting nightmares. You’ll notice that all of my UWP are shot in a maximum of three stops of light. This makes metering and assembly in the digital darkroom a whole lot easier. More importantly though it means we have details in highlights and shadows. If the highlights or shadows are not an important part of what you are communicating, like say deep shadowy forms in a sunset, then you can go beyond the three stop range and still be successful. I just can’t stress enough the importance of great light in making the UWP work.
Let’s say you’ve found the perfect locale and subject for your UWP, what next? While it’s obvious, it’s really not and that’s setting up the tripod. The height you set up your tripod makes all the difference in your final image. When shooting normally with a tripod, you simple turn a knob and reposition the camera for the framing desired. We can’t do that when shooting UWPs! We’ve got to keep that camera back plumb to the earth. Probably the hardest thing for folks to learn when first starting out is to previsualize the final image and set up the tripod accordingly. Keep in mind that the camera back is going to remain basically plumb, as in vertically perfectly plumb. We just have the tripod height and the rise of the PC lens to actually compose our image.
I typically start with the PC lens on the body unshifted and look at various heights through the lens handheld. The more UWPs you do, the easier and quicker you can do this but as a general rule of thumb, the tripod normally is set up with no leg extension. I shoot pretty close to the ground. This is because in my style of photography, I like foreground anchor points for the viewer to be able to relate to. This gives a scale and visual depth to a wide angle image. This is especially true for UWPs, so shooting low is pretty standard for me.
Once you have your tripod height selected, you want to make darn sure it’s on incredibly solid ground. There is nothing worse than to have gone through all the steps to make your UWP only to have your tripod slide on some rock that it was precariously perched on and everything going out of whack. Once I have the tripod at the height I want, I literally lean on the top of the tripod head and push down and wiggle the tripod. Once I’m convinced it is in place I go about leveling the top of the tripod. Shooting with the Gitzo 1348 and 1127G, there is a bubble level built into the top of the tripod platform. I use this for my first, but not last leveling. I often will place the carpenter’s level on the tripod as well to double check that it is as level as humanly possible. It’s now time to set up the camera.
I should step back for a moment because before we’re in the field actually shooting, we should determine the nodal point for the lens we’re going to use. This is really simple, no high degree of math or rulers required. We start by attaching the nodal plate to the camera. Shooting with the D2H, 28mmPC and using RRS MPR-CL plate, I sight through the slot on the MPR-CL plate to first line up the nodal plate on the body. With the D2H turned upside down, I slide the MPR-CL onto the body plate (RRS BD2) and slide it so I can see evenly spaced in the slot the number 8 at the back of the lens barrel and f/22 on the front of the lens barrel. With this accomplished, we’re going to place the camera setup in the tripod head.
To find the nodal point, we’re going to need a stick, branch or some straight line about 3-5 feet in front of the camera. Then we need something in the background, anything about 30-50 feet away from the camera. What we’re going to do is look at the relationship of the stick in the foreground with the object in the background. Then we’re going to slide the nodal plate back and forth in the clamp of the tripod head so that when you pan the camera with the ballhead, the stick and object stay lined up no matter where the camera is directly pointed.
Let’s say you have a car antenna five feet in front of the camera. In the background off in the distance you see a patio umbrella in your neighbor’s backyard. The car antenna is directly centered in front of the umbrella when you look through the lens. It needs to stay in this exact same position as you rotate/pan the camera back and forth. So if the antenna/umbrella are in the dead center of the viewfinder and lined up with each other, when you pan so they are either on the extreme left or right side of the viewfinder, they are still perfectly lined up with each other. If they are not, then you simply move the nodal plate either forward or backwards until they remained lined up when you pan. It’s really easy and takes only a few moments. If you’re using the RRS MPR-CL, D2H and 28mmPC on the BH-55 head, all you have to do is line up the grove on the end MPR-CL with the leading edge on the BH-55 clamp and you’re there. Whether by accident or design, it works every time. With your nodal point determined, you’re ready to go.
With the tripod set up solid and leveled, time to attach the camera with its nodal plate. Attach the camera to your predetermined nodal point. With that accomplished, you want to then level out the camera. This is accomplished with the bubble level residing in the camera’s hot shoe. Using the BH-55 ballhead, there is a bubble level that resides in the top of the clamp. This replaces my using a carpenter’s level on the nodal plate. I make sure that the level on the clamp/nodal plate and the hot shoe match up. Once that’s done, I rotate the camera so I pan the distance that is required to make the pano. If the bubbles change, I relevel, otherwise I move onto the next step.
The next step is metering. There are lots of ways to accomplish this and to be honest with you, each to his own way as long as the results are correct. There are two ways I do it. I either meter with a second D2H with the 12-24DX attached or meter with the D2H with the 28mmPC. Metering with a PC lens requires that first, you realize your metering mode has automatically switched to center-weighted; there is no matrix mode. Because the lens has no CPU (it does break in half), the D2H as well as most Nikons automatically default to center-weighted. Because of this, I have CS#b6 set to Average so that’s the metering pattern that’s active when the camera puts me in center-weighted mode.
Since the PC lens has no CPU and because light changes when you shift, you have to meter before doing and shifting.
Additionally, the PC lens because it shifts has no automatic diaphragm so there is no automatic aperture. This means it has a manual aperture that you must operate manually with a preset aperture set up. There are two rings, one actually opens and closes the aperture. The other permits you to create a preset stop so you can work with the lens wide open (so you can see through the viewfinder) and then close down to your predetermined aperture by simple rotating the aperture ring until it stops at the preset. Once you have the lens in hand, it’s all very clear.
Personally, I generally shoot at f/16-22. It’s very important that when you’re actually shooting you don’t accidentally change your f/stop. This will change your focus point and totally ruin your efforts. With the f/stop selected, I meter and manually set the shutter speed. I’ll take a test exposure and look for any blinkies on the monitor. If there are none, I’m good to go. But what if there are, say in the sky, some clouds that are blown out?
The normal cure for this would be to pull out the Split Graduated Neutral Density filter. If, across your three top images of your UWP, you have the general same exposure range and problem, then this is a good option. But if 1 or 2 of the 3 upper panels don’t share the same exposure problem, I won’t use the SGND filter. Why? Because of the variance in exposure and its affect on color. It makes it really a pain in the %$#@ to match up the panels later in the digital darkroom! I’m speaking from experience. What I do instead is take two images for the one panel with blown out highlights. I take one image normally and then I take a second one to just record the highlight detail. I then bring those two images together in the digital darkroom to preserve both highlight and shadow detail and then insert the final 1 image into the UWP. (If you don’t know how to do this, you need to attend a DLWS J.) With metering accomplished, we’re getting to the fun part, making the images!
Quick review, we’ve set up the tripod height based on the final image composition that we want. We’ve found and set up the camera based on the lens’ nodal point. Everything is leveled and we’ve got our meter setting. It’s time to make magic!
When we get into the digital darkroom to assemble our UWP, we want to have some overlap between the six images so we can merge them cleanly. This amount of overlap varies with each pano shooter. Personally, I’m looking at no more than a 5% overlap between frames. That’s real tight for most folks but easily doable with the fore described setup. Just how much you shift the PC lens determines your overlap between the top three and bottom three images. Your panning with the ballhead determines your overlap on the vertical image seams. The max shift that I go with the PC lens is eight degrees. More than that and you run the risk of having no overlap at all. I will sometimes shoot at 7.5 degrees if I’m not really confident my leveling is perfect. Nikon PC lenses have a scale on them so you know you’re at these settings.
Once you’ve shifted, you’re ready to start shooting. (Be sure you’ve closed your lens down before you actually take your first image.) I set up my UWP by previsualizing it in the viewfinder before I ever take my first shot. I look through the viewfinder and line up the left and right side of my UWP. I look at what elements will be present in each of the six images, especially the seams and sometimes make adjustments so critical visual elements are not at those seams. This is something you learn as you go; there is no clear cut advice to pass along for this part of the process. It’s where being a photographer comes into play.
Personally, I start shooting with the left side of the UWP and pan to the right to complete it. This is in part because when working with cameras with less than 100% viewing, the part of the image you’re not seeing in the viewfinder is the far right. By starting from the left and shooting to the right, you’re naturally leaving enough image overlap for easy stitching later on. Once this is all mentally set, I turn the camera to the left to make the first two images and LOCK down the ballhead. You do not want it turning or twisting one iota!
With the lens shifted, I take my first image. This will be the top, left image. The lens is shifted to 8° and is positioned so the front element when shifted is above the lens axis and the shift knob is at the bottom of the lens barrel. It is so important, that you understand where the shifted lens is positioned. Because wherever that is, top or bottom, we want the next photo to be taken so the lens is rotated 180° from the first image. We use the shift of the lens to make the upper and lower panel of our UWP. By merely rotating the PC lens in this fashion, we’re able to take the upper and lower images of our UWP. This very simple technique is how we accomplish capturing the vertical aspect of our UWP, the vertical peripheral vision that is missing in every other pano technique!
So let me repeat that. We’re going to take our pano in three sections, left, middle and right (you can take more if you like, if your scene can support the coverage). For each section of the UWP, we’re going to take two photos, the top and bottom one. To take the top and bottom photo, you simply rotate the shifted PC lens 180°. You can see, though dimly, in the viewfinder exactly what you’re doing so there is no mystery. It’s so bloody simple I wonder why no one has shared it with us all before!
I continue on with the process until I’ve taken all six images. If the light permits, I’ll check for blinkies on each exposure but when shooting sunsets or blowing clouds, working quickly saves you lots of time in the digital darkroom. If I’m going to be shooting more than one UWP, I take a shot or two that are out of focus to use as markers so I know later where the first and last images in a series are. This is all that is required to capture a horizontal ultra wide pano. What about vertical ultra wide panos? They can be done as well with the same precision, but that’s for another day.
Finishing your UWP just requires some digital darkroom time. There are some tricks that can speed up this process that I’ll have to share with you at some future time. Assembly can be done anytime but you’ve got to have the images to assemble to get to that point. Once you have taken and assembled your first ultra wide pano, a regular pano simply never looks wide enough ever again. While some of this seems scary and very unfamiliar, you’ll have to trust me that it is really simple once you go at it. I can take the six images now in less than a minute and assemble them in less than 30 minutes. It’s time so well spent once you see that beautiful print coming out of the printer. You’ll get hooked, I know. I’m sure that once you start someone will come up to you and say, “You only see the world through an ultra wide pano!”
Part II – Downloadable PDF document.