Weather stories have never been my favorites. I get that significant weather events are necessary to report on, but I always dreaded being handed the non-specific photo assignment “get me a weather feature.” Shooting video, I’ve been able to do some cool stories, riding along on a city plow and a Coast Guard ice breaker, for example. In the absence of a specific story to pursue, I like to try to do something more creative.
During the recent blizzard, I knew I wasn’t going to be able to get very far away from home, so I set up two GoPro cameras to record the storm. GoPros are great because you can put them places other cameras can’t go, and they have a built-in interval function that will shoot a photo once every x number of seconds.
I actually prefer using one of our Sony camcorders, like the PMW-200, to shoot time lapses. The Sony captures the single frames into a video file in-camera, saving you the trouble of assembling all the .jpg files afterward. You just import the video clip off the card into your NLE and you’ve got a time lapse video. The problem is that these cameras are much larger and more expensive than GoPros. Plus it means tying up my primary camera for several days. I’ve also shot some time lapse projects on an old Nikon D2HS body, but the built-in interval timer has a maximum of 999 frames, so someone has to reset the timer before it runs out. This works fine for a very long term project, like this five-month time lapse of the construction of a post and beam barn. It also works for a shorter term project where the camera is accessible, like this multi-camera time lapse of a single day in the Charles W. Morgan restoration.
The GoPro is not without its drawbacks. The camera’s battery only lasts about two hours, so external power is necessary (as it is with all longer time lapse projects). The GoPro comes with a nice waterproof case, which makes it ideal for being out in the weather, but the case makes the USB/power port inaccessible. So in the past I’ve hooked the GoPro up to power and taped it inside a plastic bag or an old piece of Tupperware with a hole cut for the lens.
Location always ends up being the first thing I have to worry about with a time lapse. I need a location with a nearby power outlet, and somewhere the camera is not likely to be stolen. I was lucky to find two good vantage points that were indoors and near power outlets for this storm. The first, seen in the video above, was inside a window in the staircase of a parking garage in downtown New London. The cinefoil behind the camera, used to block reflections in the glass, also served to hide it. Also, it was pretty high above the landing in the staircase. Has someone seen the camera and wanted it badly enough, they could have climbed up to get it. But the garage is right net to our office, and the stairs are locked at night. I figured the amount of foot traffic through there in the middle of a blizzard would probably be minimal. Just to be safe, I used our old HD Hero, so if a camera was stolen, I wouldn’t be the brand new Hero 4.
The last thing I had to figure out was the interval. I try to keep it as short as possible. This means more files to process, but you can always speed up the video to make it shorter. You can’t go the other way without making the playback look choppy. After some quick math I figured a photo every 30 seconds over a two day period would be close to filling up the 16 GB SD card. I set the camera up at about 11 a.m. on Monday and headed out to my second location. I picked the overhead view of New London because it was a recognizable place that would have a good amount of foot and vehicle traffic so you would see something more than just snow piling up. The problem with a high vantage point is it’s hard to tell the difference between six inches and two feet of snow from that far away.
For the second location, I wanted a very low angle so you could really see the snow pile up. I used the door to my back deck – about as safe a location as I could choose. When the field of view is going to filled with snow, it’s helpful to have something in the foreground to lend perspective. I set a stuffed snowman toy on the deck (tied down to avoid any wind issues) and set the Hero 4 to the same 30 second interval. I lucked out that the snow drifted up against the door so you get the cool effect of accumulation and then settling, which was a nice surprise.
I retrieved both cameras on Wednesday morning and downloaded the cards. One frame every 30 seconds equals 120 photos per hour. Multiply that by 48 hours and you have 5,760 pictures. I edited the videos in Final Cut Pro 6, which takes some time to process all the files. The first step is to change the still/freeze duration under user preferences to 1 frame. I then import all the folders that I copied from the GoPro’s memory card and drag them onto the timeline of Sequence 1. This yielded about a three-minute video. I create a second sequence, and drag Sequence 1 onto the timeline of Sequence 2. This turns all of the individual still files into a single video clip that can be manipulated using the tools in Final Cut. I used the motion tool to zoom in, eliminating the black bars on the side and filling the 16×9 video frame. If necessary, you could color correct the whole clip. I also change the speed of the clip, cutting the duration down to 60 seconds. I added a piece of royalty-free music from Kevin McLeod, and then render and export. This is a fairly simple process, but the time to copy files, import, render, export, compress and upload ends up being over an hour for each video. Every transformation needs to be rendered before you can play it back, so making a lot of changes will chew up a lot of time.
This method has worked pretty well for me, minus the processing time. There may be better software solutions out there. If you’ve got a suggestion, feel free to leave it in the comments.