What happens when you string about 4,000 images together and view them in three minutes? That creates a whole lot of motion. “Valley of the Sun in Motion” is a compilation of 32 scenes from around the Phoenix, Ariz. area taken between June 2011 and February 2015. While I lost track of the grand total of pictures I took in order to get down to the final product, it was well over 10,000.
I think a project of this size requires some explanation and background information, both to help me put closure on the multi-year undertaking and also in case others (like my photography students) are hungry for details and how it might relate to their own work. So here goes:
I started the project in June 2011 after putting together a timelapse assignment for one of my high school photography classes. I always try to find new, relevant examples of successful work to share with students as they start projects. That’s how I came across this compilation called “The Arctic Light” by TSO Photography. It is absolutely inspiring on a number of levels–-one of which was to make me want to give something like it a try.
When my students set out to create their timelapse projects, I headed to a nearby park at sunset with my camera and came back with this. As I combined the pictures and watched the clouds gently roll, the water ripple, the sun set and the sky change colors until the clouds finally glowed red, I was hooked and decided to tackle my own compilation from around the Phoenix area. The Valley is beautiful, although I know it’s no arctic. I also don’t have the equipment TSO Photography uses to create smooth, clean panning (side to side movements). Still, the desert-meets-urban Southwest provides plenty to work with.
I started scouting potential locations soon after the park sunset timelapse. I often headed to a trailhead in the Phoenix Mountain Preserve or looked for elevated vantage points and explored until I came across something that would work. It seemed like every time I found one good spot, I would come up with two or three more “must have” locations. I did a lot of shooting during the summer (yes, even when it was 117. I brought a lot of water, wore a lot of sunscreen and kept outings efficient). The school year slowed me down for a few months, but I picked it back up in November and December 2011. I could have easily spent another six months going to different locations, but decided I had to call it done at some point. In January 2012 I published my first draft. It included a little more than 2,700 photos from 18 locations, and clocked in just shy of 3:30. It was cool to watch and I was happy with it. For a bit.
Then I saw this timelapse from Yosemite and came to the realization that my own video looked like amateur hour in comparison. I know I didn’t have Yosemite in my backyard, and I still didn’t have the expensive and automated sliders needed for the great panning. But I did have a desire to revisit my project and improve it. In December 2012 I headed back out on the trail. I had a few new spots in mind, and a few locations to revisit and improve. Between December and February I made it to 11 new locations. Then another draft. Then another round of videos way better than mine. Another round of new scenes and better editing in the winter of 2015, which is when I finished the third and final version of the video.
The Specifics and Nerdy Camera Talk
I used a Canon 30D, 7D and 5D mark iii for all of these photos. When taking photos I obviously used a tripod, and also a wired remote. For some early scenes I would manually fire the shutter using the remote and just count between shots. I then used a TI-83 graphing calculator as a makeshift intervalometer to automatically fire the shutter at predetermined increments (check out this video for how to do it), then finally bought an actual intervalometer.
Depending on the location and what kind of movement I was after, I varied the time between photos anywhere from one to 20 seconds. Shorter increments captured more details, like people and water moving. Longer increments captured more gradual environmental changes, like the sun setting or shadows moving. For my camera settings, I tried to use a small aperture (the opening in the lens, and small meaning a high number like f/22) to maximize my depth of field. The smaller the aperture, the deeper the focus will be in your frame. In addition to deep focus, the other benefit of using a small aperture is it requires a slower shutter speed to get a good exposure. Normally a fast shutter speed allows you to stop more action. As someone who photographs a lot of sports, I usually fight for the fastest setting possible. But in the case of timelapse photos, “dragging” the shutter, or using slower than usual speeds, can actually help since anything moving will be a little blurry and allow the motion from frame to frame to blend together. My shutter speed usually was somewhere around 1/30th, 1/40th or 1/50th of a second.
For most scenes I put the camera in manual mode and dialed in the settings I needed. In cases where the sun was setting, like this scene, I used the aperture priority mode. This allowed me to keep the aperture the same and the camera automatically adjusted the shutter speed as the sky got darker. The downside to this method is if the shutter speed bounces up and down as clouds roll by or light changes, the images become slightly darker or lighter and the video can flicker (like what happened in this scene if you watch closely).
Making the Video
For post processing, I occasionally had to remove some spots on photos that were caused by dust on the image sensor of my 30D (a downside of an older camera without a self-cleaning sensor. Cleaning swabs may clear existing dust, but new stuff always seems to move in.). I made those edits, and other minor level and color adjustments, in Photoshop using “actions.” This allowed me to record the adjustments on one photo and then apply the same changes automatically to every other photo in the batch whether there were 88 or 592 (my actual low and high counts from the first round of sequences). Eventually I moved on to Lightroom, which made bulk edits far easier.
I combined images in Quicktime Pro, most frequently at a rate between 15 and 30 frames per second. For a few scenes I used 12 frames a second when 15 just made it look too quick. Some of these I exported and posted to my Flickr account as individual videos. Click here to see the individual posts. I then combined all the individual scenes together in iMovie on my MacBook Pro for the first and second draft, and Adobe Premiere for the third videos.
In my original draft of the video I opted for longer scenes, probably because of my own attachment to the individual locations and the time I put into each one. I think I felt like I had to show just how much I captured. That ultimately is what made it not so great. In the revised versions I shortened scenes a lot in favor of a punchier flow and more variety.
The music was a process in and of itself. I didn't want to use a song without the proper permissions, so I spent a lot of time exploring online for music, including Creative Commons, Vimeo’s Music Store and www.mobygratis.com. But after hours of digging through online options I didn’t have much luck with the sounds I was after. Perhaps there were just too many options to filter through. Plus I felt limited by licensing options on some sites. Many would either require me to license my video under a ShareAlike license, or prohibited me from ever using the video commercially. I’m not opposed to ShareAlike licenses for some work in theory, but I wanted to keep my licensing options open and I enjoy making music, so I opted just to create the audio from scratch using loops in GarageBand to include in the 2012 version. It certainly was not Moby quality, but it worked at the time.
Again, enter Yosemite HD. The M83 song they licensed for the video added so much to it, and it made my GarageBand loops sound like a midi beatbox created on a cheap Casio keyboard. Back to the Internet. I still did not have the capital to license a song, but there had to be a better option out there somewhere. You can find anything on the Internet, right?
Right. I found ccmixter.org and was able to come up with several options, even options that allowed me to potentially license the video commercially. I ended up liking this song. It was way too long though, so I trimmed it down using Adobe Audition.
For the final version of the video I went a different direction (again) after a co-worker referred me to Dexter Britain.
Then in the End
I could probably tally up things like hours spent processing images, time hiking on mountain trails, curious looks received from passersby, miles driven, puffs on my inhaler, tracks of GarageBand loops, songs I almost used, etc., but those numbers make me nervous. And more importantly, I think that would make it more about bean counting and less about just doing something creative with photography, getting outside and appreciating perspectives of my hometown I think many people often take for granted. Thanks for watching, and for reading if you’ve made it this far.
Enjoy the video, and the motion.
-Mica Thomas Mulloy