Practising large scale panoramic photography is a step up from creating a panorama with an iPhone or even a 360° image with a GoPro Max, Insta360 or Ricoh Theta. We’re going to be looking at panoramic landscape photography in particular where the scene has left, right, top and bottom edges.
Many of the principles will carry across to the spherical panoramas used in 360° photography, but we need another post to do that topic justice – the ultimate 360 Photography guide..
What is Panoramic Photography?
Simply put, Panoramic photographs are images created from a row or more of overlapping photographs. The finished panoramic image is “stitched” together using software, and presents as one photograph.
It should be obvious that each image needs to be identical in terns of its exposure, so that when they are stitched together, the effect is uniform without having to visit photoshop to iron out the differences.
So we’re going to look at the equipment you will need, the problems you will encounter and consider ways to resolve them. Taking a three shot panorama is one thing, taking a gigapixel panorama of 50 images or more is quite another.
Panoramic Photography Equipment
Use the best camera you can get your hands on. The more definition you have in individual images, the better the final result will be.
You need a steady tripod, preferably without a centre column. The reason is that centre columns introduce movement in almost every tripod I’ve owned. The one exception is my Feisol CT3442 which doesn’t have a centre column. I’ve found that any movement outside of the axis of rotation encourages stitching errors.
If your camera is not level then as you rotate it to shoot rows of images it will force you to crop a part of the highest image in the panorama because each successive shot in the row steps down (or up) so you’ll crop to the nearest rectangular shape for the whole frame. This can render a whole shoot useless. Check out my review of three levelling bases.
A panoramic head introduces predictability into the equation. If you shoot three rows of ten overlapping images, you can work out the degree of rotation and tilt, and repeat it exactly for all rows. In panoramic photography we need to save time so that we can capture the light. For this reason alone, a panoramic head is a good investment.
Many landscape panoramas are shot with middle ground and background only. So parallax errors are easy to get rid of in stitching. However, if you do have a foreground and many panoramas shot in portrait mode do, then you will almost certainly run into parallax problems. .A Nodal Rail helps to find the nodal point for your camera/lens combination and eliminate the problem at source. Why is this important?
Close one Eye. Hold one finger up and align it with a vertical object in the background. Now turn your head. See how the finger appears to move in relation to the background?
Turning your head in this example, physically moves your eye. So the effect is exaggerated a little. Imagine the camera on a tripod. It rotates around the axis of the centre column.
Moving the camera along a nodal rail helps you to align the nodal point with the axis of rotation. Once that is done, the parallax errors will be minimised and your pictures will stitch without error.
If you’re doing landscape photography, then you’ll want to invest in an L Bracket so that you can shoot in portrait mode. There are three reasons you might want to do this.
- So that you can include sky and land in one image. Easier to stitch if there are edges in the photograph.
- To minimise the number of rows you shoot. Why make life more difficult than it already is?
- To keep the centre of gravity stable on the tripod. Most ball heads have the ability to swing over to 90° or more. But this pulls the centre of gravity away from the centre of the tripod and you risk losing the ‘level’ as you rotate the camera to shoot successive frames.
Shutter Release Cable
This helps to avoid unnecessary vibration.
This is a level of equipment required to create panoramic images to a professional level. Its possible to do it with less equipment but it will be slower and way more error prone.
In terms of learning curves, I’d advise against diving in the deep end. Start simple and move on up once you’re confident you’ve mastered the camera/tripod combination.
Camera Settings for Panoramic Photography
If like me you’re used to shooting RAW, leaving white balance on Auto and adjusting white in post processing, then you’re going to have to change your ways for panoramic photography. This goes back to. the fact that you’re shooting multiple frames and stitching them together afterwards. It’s imperative that every frame has the same white balance.
Technical advances in digital cameras make this less of an issue than it used to be. You can shoot at ISO 1000 perfectly comfortably with most modern cameras. The reason you might want to would be to do with moving objects in the frame. Even clouds can ruin a panorama if they move too much between frames.
So shoot at an ISO level that enables you to work fast once you begin shooting.
A lot of people shoot aperture priority in landscape photography. This isn’t going to work with panoramic photos. For the same reason I cited with white balance, you want every frame to be shot with identical settings, so that the whole panoramic image looks like a single shot.
Software for Panoramas
Keep it simple is the key here too. If Adobe Lightroom works for you, that’s fine. Move on up on the basis of what you need to take your panoramic photography to the next level.
DxO Pure RAW (Noise Reduction Software)
Noise reduction is still a thing. The Canon cameras I use benefit from shooting to the right, by which I mean adjust the exposure so that the brightest bits nudge the right hand side of the histogram. Adjusting in post processing to get a more aesthetically pleasing image. Retrieving detail from an exposure that is too bright (not blown out) is easier than retrieving detail from dark shadows.
More often than not, it is pulling detail from dark areas that give you visible noise. One way you can deal with this is to use noise reduction software such as DxO Pure RAW to remove the noise before you start processing the image.
Note that with astrophotography, noise reduction is achieved by stacking multiple frames. This is more effective than other methods as it doesn’t remove the stars!
I tend to pass my photos through DxO Pure RAW to remove noise before bringing them into a RAW processor. Whatever you use, Lightroom, DxO PhotoLab or ON1 Photo RAW, you need to be able to process one image then copy the processing to all the images you have shot.
Lightroom also has the ability to merge your photos into a single panoramic image.
Photoshop has the ability to merge multiple photos into a panoramic image. I use photoshop to produce my final image..
Dedicated stitching software, better at stitching panoramas than anything Adobe have created.
How do you embed a huge megapixel panorama in a website without sacrificing performance? Pano2VR Breaks your finished panorama up into ’tiles’ and provides the intelligence to reassemble in a browser. Essential if you want people to be able to really see the detail in your panoramas and load them fast.
We’re getting into virtual tour territory a little. You can also add hotspots so that in a huge landscape for example you can make a village ‘clickable’ and display information about that village.
Hosting Gigapixel Panoramas
The problem with breaking panoramas up into tiles is that a lot of hosting is predicated on inodes – the number of files served to make up a web page. I use Amazon S3 to host panoramas and virtual tours for this reason.
Now that we’ve listed the software and hardware, let’s step through the process of creating a panoramic image.
Landscape photographers know that shooting panoramas can involve hiking long distances to get the ideal location. So pack only what you need. Every kg takes a little more energy to carry.
Like many landscape photographers I’ll scout my location either personally or on Google Maps. I’ll use the Photographers Ephemeris to see where the sunrise for sunset is coming from in relation to the planned photograph. And I’ll pack exactly what I need to get the multiple images I need for the finished panorama.
Don’t forget to pack water and food. Dehydration in a hot climate can kill.
Choice of Lens
Panoramic images are often shot on telephoto lenses rather than the wide angle lens favoured by many landscape photographers. Whatever works best for you is the right lens, but bear in mind that the distortion introduced by wide angle lenses does not lend itself to stitching.
Despite the weight, I often use a 70-200mm lens. Getting the detail is what works for large panoramic images and you won’t get that with a wide angle.
Setting Up the Camera
Ok, I’ve driven to a point close to my location. I’ve hiked from the car to the position I’ve chosen. I’ll take a test shot, handheld to give me an idea of what the potential is real time. from the test shot, depending on the lens I’l be able to work out how many rows and columns I need to shoot for my panorama.
First I’ll set the tripod and level it. This is not absolute, because once I start adding bits on top of the base, chances are the weight will pull the tripod off level, but it will be close. I shouldn’t have to adjust the legs of the tripod after this.
Next I’ll add the levelling base and on top of that the panoramic head, I’ll adjust the levelling base so that my panoramic head is level.
Add the camera and lens to the panoramic head (the nodal rail comes below the camera if you are using one).
Adjust the levelling base again so that you know your entire rig is level.
Put the camera in manual mode. Take some test shots, handheld. You’re looking for a reasonable exposure that will retain detail in the darks and the lights. This may be a compromise in camera, but you’ll be able to rescue detail in the edit. What you can’t rescue is blown out highlights or black areas where there is no detail to rescue.
Use a narrow aperture so that you can focus front to back of scene.
Adjust the ISO upwards so that you can shoot minimally at 1/100 second. This is to minimise cloud movement and prevent it from causing stitching errors later.
Shoot the first row of your panorama overlapping each shot by around 30% so that the stitching software has plenty to work with. This row should be dead level. If it isn’t you’ll be forced to crop later. For the same reason, shoot wider than you actually need so that you can crop to size.
Adjust the tilt on the lens up and shoot the top row if needed.
Adjust the tilt on the lens down and shoot the bottom row if needed.
In architectural photography I’ll sometimes shoot a panoramic image so that I can get the whole scene, in detail into one frame. In this case, I use the shift rather than the tilt so that I don’t have to worry about visible distortion with converging verticals.
Check the output in camera for any obvious errors. If need be, shoot again until you’re satisfied with the result.
Don’t forget at this stage, sit and take a few minutes to enjoy the scene. I don’t say this to be facetious. You’ll notice things that may help in the editing later. My experience is that it really helps in some indeterminable way if I am ‘connected’ to the scene I am shooting. Perhaps its just that I’ll put the extra mile in if I feel a strong connection.
First I import the set of images into Lightroom so that I can tweak the exposure and carry out basic raw processing across the set while I still have the detail captured in the RAW file.
If necessary, I’ll apply noise reduction at this point.
Then it’s time to create the panorama.
In Lightroom it’s the “merge” command under the File menu. Choose “Panorama’ and let it work its magic.
I prefer to use PTGui because as a specialist tool it does stitching very well indeed. In Lightroom and photoshop, I can’t get at the layers after it is stitched. In PTGui I can, so this gives me the ability to adjust the stitch line to get rid of time based anomalies. We’ve all seen the unlikely stretched dogs in Mobile phone panoramas, any moving object along the stitch line will cause problems. These can be resolved easily in PTGui.
Once you are happy with the stitching, output the result as a Tiff format file.
In photoshop, or your favourite alternative, you can process the tiff just as you would a normal photo. Draw the viewers attention to the areas you want them to focus on, emphasise leading lines and boost contrast, adjust colour etc.
Converting to Website Friendly Format
Export the finished image and import it into Pan2VR (krpano is a reasonably priced alternative). The software will basically create a mini website, chopping your image up into small tiles that are quick to load in a browser and providing the ability to render it as a complete single image that the viewer can zoom into, move around and enjoy a properly immersive experience.
Configuring S3 for hosting is not straightforward, I’d write a tutorial but Tony Redhead has already done it so take a look at Configure S3 for 360° Panoramas for a really well constructed guide to setting up an account to host the output form Pano2VR and embedding it in your website.
Although this article is written from the point of view of a landscape photographer, panoramic photography enjoys a number of uses. The same techniques can be used in architectural photography, cityscapes, astrophotography and the practice although exacting will make you a better more considered photographer.
I’ll keep you in the loop with regular monthly updates on Workshops, Courses, Guides & Reviews.
Sign up here or..