The timelapse photography is a technique that allows you to add another dimension to your works. Instead of a single image, the result is a movie created of frames which were taken at much slower speed than the movie is actually to be played. The technique is easy, just requires more time to properly capture the images, as well as some extra work, processing power and disk space.
In many situations an extra challenge appears - a gradual and significant change of lighting in the photographed scene, like during sunset. This is where it is especially worth to consider HDR processing techniques for the frames. Unfortunately in some situations it may be not enough, if for example lighting changes fast and randomly, like on a partly cloudy day, when the sun is from time to time blocked by the clouds.
The timelapse movies can be made of frames taken from a static tripod, or from an automated tracker, that can be used to add extra motion. The tracker could follow movement of stars, or slowly roll in one direction, etc. You'll get best results if your movie is played at fast frame rate, like 20-30 fps, or even more. That means a lot of input frames are required.
The example presented below is a result of merging a series of 130 photos taken from a static tripod, for the period of 39 minutes. Each photo was taken at 15 seconds of exposure and there were 3 seconds of delay between each shot. At first I had to decide on the exposure settings that should be used for all photos. It was after the sunset, so the sky was getting darker, therefore I decided to overexpose the first photos. At the same time the almost full Moon was rising on the opposite side of the sky, so it ensured that both the sky and ground will be bright enough later. Below are 3 sample, not processed photos - the first one in series, the last one and one somewhere in the middle.
As you can see, the unprocessed photos do not look good. The first one is overexposed, while the other two are certainly underexposed. If the exposure time was not that long (15 sec per image), HDR sequences could be considered. Unfortunately here single photo per frame had to be enough. A bit more can be squeezed from RAWs, so I have used them instead of JPEGs. The camera was a very old Canon 650D; modern cameras characterize with much better dynamic range, so more detail could be extracted.
Another step is to decide on how to process each of the photos. In easyHDR just load one (or an individual HDR sequence) and use the normal processing mode. Choose a setting from presets, or create an own preset. It's better to try the same settings on several photos from the timelapse series, like in the example below.
To process all the photos, just use easyHDR's batch processing tool. First create the tasks - one task is a single movie frame (use the "Auto create tasks" dialog). Then make sure all tasks are marked (use mouse, or press Ctrl+A to select all) and apply the settings. It is possible to set multiple tone mapping settings for all tasks. It is advised to save the task list to file before running the batch.
Apart from RAW import settings, tone mapping settings, result format and destination, is also possible (only for RAWs) to apply dark frame as well as flat frame correction. To create such frames please use the tool available through easyHDR's main menu: under: File->Make master calibration frame. There is one extra option for the tone mapping: "Force disable hostogram clipping". You may check this checkbox for all tasks to disable automated histogram clipping normally done by easyHDR as one of the tone mapping algorithm steps. In some circumstances it may reduce possible flickering of the movie.
Once the preparations are done, you may start the batch processing. The operation can be stopped any time and continued later on (that's one of the reasons it is good to store the task list before).
The final step is to assemble the movie from the processed frames. I used ffmpeg. Below is an example command line that performs the operation. It tells ffmpeg to create the movie in MP4 format, with very good quality, at 20 frames per second, using photos named 'Img_%%d_easyHDR-enhance.jpg' where the '%%d' is an index starting at 8078. For more information on use of ffmpeg, please refer to its documentation.
ffmpeg -framerate 20 -start_number 8078 -i Img_%%d_easyHDR-enhance.jpg -c:v libx264 -pix_fmt yuv420p -crf 5 result.mp4