I knew it was just a matter of time before I had to start learning to use Photoshop. For years, the free and open-source GNU Image Manipulation Program (GIMP) served my limited graphics-producing needs well. In fact I still remember some tricks to make that program perform at a much higher level than it was originally designed for (thank you plug-ins!). As Adobe started adding more and more features into their Lightroom/Photoshop suite–in addition to the slow but inevitable march towards a dastardly subscription model–I shifted back to see what I was missing. Over the past year, I slowly introduced myself to this powerful program, and am still taking baby steps to learn the ins and outs.
Now that I’ve gotten comfortable with some of the basic, but still extremely useful, functions of Photoshop, I figured now was as good a time as ever to start sharing step-by-step tutorials in case you find yourself stuck in the same boat I was one year ago! Today’s guide will focus on a very specific kind of image stacking and blending, something I like to call image splicing. You can follow this process whenever you have two or more images made from the same position, but want to selectively use subjects from any number of the images.
In this example during a late-Fall photography walk through Sankeien Garden in Yokohama (more photographs to come in a later update!), many couples were busy with wedding photographs. Though I make sure to never intrude on another’s photoshoot, I got many casual snaps as I passed by the pairs (and I was one of many that stopped to look and see!). As a result of not being able to control the shooting environment or the amount of time I could stand still to get a shot, I ended up with two similar photographs that separately are “okay”, yet both have better elements I wanted to combine. Seen above in the title image, there’s more empty space to the left of the man for a better composition, and the look to his bride is one of genuine happiness. The image on the right completes the left well, as she is also looking directly at him (and the bottom half of the shot isn’t blocked by a passerby!).
So the question, then, what’s a good way to combine the best elements of each? Though this is by no means the only way, here’s the process I followed after getting the basic tonal edits in Lightroom (for full-size, click on images as needed).
Difficulty Level: Easy
Margin of Error: Medium
Estimated Time of Completion (First Time): 15 Minutes
1. Open as Layers in Photoshop, then Auto-Align
After opening two (or more) images in the same workspace, ensure they are all highlighted in the “Layers” panel, then go to Edit–>Auto-Align Layers.
For convenience, leave all settings to “Auto” (assuming you were careful to shoot the images from a relatively fixed point) and let the program do its thing.
As you can see above, even though I was in the same place with only a split-second between shots, there are still a couple areas that don’t overlap perfectly (the checkerboard white pattern around the edges). We’ll come back to those later.
2. Determine Which Objects to Delete
Since I wanted to keep the parts of the images where both subjects are looking at each other, I need the image with the woman looking down and the man’s face blocking the shot as the image on top of the stack. That already happened when I Auto-Aligned, so I moved on to select the lasso tool. For this step simply draw a rough line around the objects you want deleted. I recommend setting the “Feather” option to anywhere between 3-5 pixels to help fade in the edges in case the image isn’t perfectly aligned.
Once you have the animated, dotted line completed (and you have the top layer correctly highlighted), just press delete on your keyboard and…poof!
The majority of the processing is finished at this point, but there is often clean-up work to do.
3. Inspect for Aligning Artifacts, Content-Aware to Fix
At first glance, it looks like I could just crop the image and be done with it. However, looking a little closer there are often errors and artifacts that need to be cloned over. In this example, it’s easy to see zoomed in a bit more that there is a visible “break” in the rope next to the man’s hand as well as several doubled artifacts from the two images not perfectly overlapping on the ground (again circled with the dotted lines).
In this case, since the errors are not important to the photograph and occur near patterns of natural background, I chose to use Content-Aware Fill to clone out these instances. If you want to make multiple selections when using the lasso tool, simply press and hold Shift while making the additional selection. Before going through with Content-Aware Fill, go ahead and merge the layers so Photoshop treats the fill as one image.
Now, simply right-click within one of the selections and select Fill.
There are several options to help the Content-Aware algorithm function in different scenarios. For the most part though, Color Adaptation, Normal, and 100% opacity will get you the result you’re looking for.
Once again, after a few moments…poof! The errors should have been cloned over. If they are not, try reselecting and trying again, or attempt one of the other modes (changing the size of the feathering on your selection can sometimes help too). Now, it’s just a matter of maintaining continuity in the final image.
4. Edit Continuity Errors
So now that the aligning errors are taken care of, I need to check if the cloning affected continuity with the rest of the image. In this instance, cloning out the rope in-between the couple makes the rope on either side of them not make sense. Since the rope doesn’t necessarily contribute to the image, it’s fine to clone it out entirely now. As before, select as needed.
Right-click within one of the selections and select Fill again. We’re almost there!
5. Crop as Needed
As mentioned earlier, we come back now to removing the excess transparency areas in the image, as well as cropping to help improve the composition as needed. Simply select the crop tool and move the borders as you see fit in your image.
If you are working between Lightroom and Photoshop, you can simply click “Save”, and the finished file will appear in the same catalog the source images came out of! From there, you can double-check all your tonal edits and crops, and export the final image. For my shot, I ended up cropping out a bit of the sky to help put more focus on the couple (they also more closely approach a half-and-half composition this way).
If you were wondering what the random red and blue dots are in the sample screenshots of Photoshop, they’re exposure clipping indicators. Lightroom uses a similar feature that I leave on in Photoshop for continuity’s sake.
That’s all for this tutorial, guys and gals, I hope you learned something today! As I use new features more and more, I’m sure there will be more tutorials to come in due time. I already have some ideas for basic panorama images, as well as enhancing shallow depth-of-field in images that lack it. Once I come up with a great sample image to work with for these concepts, the tutorial will be next on the list! As always, thanks for dropping by, and have a great day! 🙂
P.S. I may not have time to schedule my end-of-year update post before I go on Christmas Leave. If I miss that window, you can expect me to be back updating the site no later than mid-January!
Thanks…u should also try using layer masks….that makes much easier to correct images like these.
Thanks Aditya! I am still learning the ropes on layer masks, but I agree, those can help too. I have used them on photo collections before to good effect.