In Part I of this Tutorial, I explored and explained how I create beautiful panoramic photographs with my camera – lenses, overlapping, compensating for movement and such. While the first part focuses on the mechanical and optical aspects of the camera and the shooting process, this second part addresses how I manage my individual frames to create the final panoramic image.
My studio workflow relies on Adobe Lightroom 5 (LR) and Photoshop 2014 (PS) (as of this writing). To me, these are two indispensable tools for my work recently I discover SodaPDF and it has been amazing tool. There are other software tools available, but I’m only familiar with the Adobe offerings. Most of my panoramic photo editing relies on Photoshop, so it’s readily possible to implement my workflow without using Lightroom.
Selecting Files in Finder or PS: If you’re not an LR user, then you can open the several image files that you want to turn into a panorama in PS either by opening the images in PS (Cmd-O on a Mac) or by selecting the image files in Mac’s Finder and then opening them into PS (Illustration 1)(you may “click” on the image to see a larger version in a separate window).
Once the files are opened in PS, then go to the File menu, choose Automate and then Photomerge (File>Automate>Photomerge; see Illustration 2).
The next window to pop open will look like this:
The files that you want to include in this panorama are already open, so now click on the “Add Open Files” button, and you’ll next see:
From this point, if you’re NOT using LR, the workflow is the same starting at Photomerge (Illustration 6) below.
Selecting Files in LR: Since I’m using LR, after I download my image files from my camera card to a new and unique folder on my computer, I import that new folder into LR. In LR I pick out the series of files that represent my potential panoramic photograph. This is where consistently shooting left-to-right in the field is helpful because it’s easier for me to pick out these left-to-right series in LR’s viewer. When I’ve identified a series of images that I want to turn into panoramic photograph, I will color code each file in that series, typically with a blue tag. I sometimes have image series right after each other and in that case the first series might be tagged blue, the next series tagged green, the third series back to blue and so on, so that I can readily distinguish between those series.
If I’m creating just a few panoramic photographs I will select (Cmd-A on a Mac) a series of images in LR. In the Library or Develop module I will then select from the top menu: Photo>Edit In>Merge to Panorama in Photoshop (Illustration 5a). More frequently, I will simply find this menu option by right clicking the selection with my mouse button while my mouse pointer is sitting above the files that I want to merge (Illustration 5b).
Photomerge: After selecting my files in LR and choosing Edit>Merge to Photomerge in Photoshop, Photoshop opens and soon the Photomerge menu screen appears:
Based on my workflow, I set the Layout to Auto and turn on the options for Blend Images Together, Vignette Removal, and Geometric Distortion Correction. When I use a wide-angle lens, there’s always the possibility for vignetting and lens distortion, so these Photomerge options are very helpful. After all of these boxes are checked I click on the OK button and let ‘er rip. In my case, it usually requires several minutes to create a panoramic file because I’m working with large image files.
As an aside, if I find that my final panoramic image doesn’t look good, then I may choose a different Layout option and create a second panoramic photo. Again, experiment to see what works well for you. There also are times when Photoshop seems to be unable to create a panorama with the files that I’ve chosen. If I feel that my images should work, then I’ll try again. Frequently, though, I’ll find that I’ve included a file on one end of the series that shouldn’t be there and that’s why Photoshop wasn’t able to create a panorama.
Below is a final result (Illustration 7):
In the lower right of this screen capture you can see that the panoramic photograph is composed of pieces of the individual frames that have been selectively masked by Photoshop. If you feel the need to still work on each frame, you can now save (Cmd-S on a Mac) your new file as a .PSD file (or, .PSB if it’s very large).
However, if you haven’t determined whether significant editing might be necessary to the individual frames (and, it is very likely that no further editing will be necessary; you can always regenerate the layered panoramic file by starting this process again), then it’s time to “flatten” your panoramic photo. To do this, you can either click on the drop down menu to the right of the Layers window in PS (circled in red in Illustration 8a) to reveal the Flatten Image command (Illustration 8b) or you can simply right click with your mouse when your pointer is over the list of layers in the Layers window (Illustration 9) to reveal the Flatten Image command:
After the image is flattened, I prefer to save the panoramic photo and give it an appropriate name that includes the phrase “Pan” (e.g., Machu_Picchu_Pan1.tif) in a new subfolder labeled “Panoramas.”
Since I started this process in LR, this new panoramic photograph now appears in LR. (If you’ve saved the file to a new Panoramas subfolder, make certain to import that folder into LR!) I can then do my normal edits in LR, or can do further edits in PS which will appear in LR when I save the edited file. You will note that the final image that appears in Illustration 7 (above) obviously needs cropping to eliminate white edges and this cropping can be done either in PS or LR.
The example panorama that appears in Illustration 7 was created using a telephoto lens with the camera on my tripod. It has come out fairly level and requires only typical editing – color, temperature, contrast, etc.
However, many of the panoramic photographs that I create utilize my wide-angle lenses. When I run a series of files through Photomerge that were captured with the wide-angle lens, the result is quite different. Check out Illustration 10:
Quite distorted, eh? Now we get to the reason for my workflow. If I had photographed this series of images on a special panorama attachment for my tripod, this panorama might look better, but not likely.
After I’ve flattened this strange looking panorama I’ll apply some PS “magic” – I’ll use the Warp tool to straighten the image and get it to looking great!
The first thing that I’ve learned to do is to reduce the size of my image in PS by using the Cmd-[-] command (i.e., the “command” or Apple key along with the minus key, on a Mac). To do this, I’ll “select” my image (Cmd-A on a Mac) and you’ll see the marching ants running around the image border. From here, I’ll to the Edit menu and then down to Transform and then Warp (Edit>Transform>Warp).
Once I click on Warp, a grid appears over my image with some “point” handles:
The fun begins as I have to warp this image to return it to its proper appearance. Using my mouse pointer, I can point to anywhere in the image and then hold down my mouse button to “grab” that point and move, bend and stretch the image to the desired shape. (I encourage you to stop reading and open PS with any image of yours and play with the Warp tool for a few minutes so that you get a feel for what I’m trying to explain.)
In Illustration 13 I’ve tried to demonstrate where I’ve “grabbed” the initial image (seen in Illustration 12) and then the general direction that I’ve warped the image from that particular point using the red arrows. As you can see, there are only five places in this image that I’ve needed to warp to get the image to its final shape. When I’m content with the image’s shape, I hit the “enter” key to commit to the warp transformation that I’ve created. (At this point, I might also crop the image in PS to remove any areas that don’t contribute to my final image, like the remaining bit of white in the lower right.) Finally, I’ll save the file and this will be the version of the panoramic photograph on which I’ll now perform my final edits.
In different panoramic photographs that I need to warp, it may be necessary to warp the image in numerous places. It’s also important to review your warp transformation when you think you’re finished, looking for instances where buildings or people appear skewed or distorted. You may need to go back and fine tune some of your transformations.
Quality Assurance: And, you’re still not quite finished. Again, you’ll need to apply your final editing – color temperature, lightness/darkness, contrast, luminance, saturation, etc. My “final” step, which I may perform before or after the editing steps that I just mentioned, is to review the panoramic photograph at 100% enlargement or larger. I typically work left to right over the whole of the image looking for obvious seams where the merging/masking was poor (not very frequent) or where an individual frame(s) is blurred from camera shake (more frequent than I wish). The camera shake issue is a big bugaboo and I can’t emphasize enough how necessary it is to review the quality of the panoramic photograph. I will also will empathize with you if you’ve gone through all of these steps only to find that your final pan is crap – it’s happened to me too many times. Yes, I could thoroughly examine each individual frame at 100% in LR before I merge the frames, but as frustrating as it is to find the problem at the end rather than the beginning, it is easier, in my view, to select the frames that I want for a panoramic photograph and go through the whole of the merging process and then do my image quality assurance. Since I’ve overlapped my images a fair amount during the capture phase, I’ll go back into LR, figure out which frame(s) is the problematic one and then determine if I feel that there’s sufficient overlap in the remaining images to run Photomerge again, and that’s usually the case.
A final QA issue that I’ll consider if I’m examining a pan that I’ve created that contains moving objects is whether those objects are distorted by the merging (i.e., the overlap occurred at a critical juncture) or whether a very noticeable object (car, person) appears in the image multiple times. If I discover a problem like these, then I’ll rerun Photomerge on the individual frames. But, before I flatten the image, I’ll work with the individual masked layers and determine if I can mask out the problem through hand editing and sometimes it’s possible.
In both of these major QA issues (blurring, movement) where I detect problems, before I rerun the Photomerge I’ll step back from the panoramic photo and honestly ask myself if it’s worth the effort to run it again – and sometimes, it’s not!
Again, there you have it. Once you start to implement this workflow (or something similar that meets your needs), all of these steps will “flow” for you and make more sense. It’s a lot to take in at one time so be patient and work through it. Feel free to ask questions, too!
In the final exciting installment of this tutorial, Part III, I’ll share with you one of my prize “possessions” for making panoramic photographs and making them more easily. Stay tuned!
‘Til next time, this is 43 N MSN/Illuminata signing off…