Introduction
I have already written in
former blog
posts why there is a need to automatically align time lapse members when
shooting either during windy conditions or with a tele-lens and how my app Futura
Time Lapse can help.
A few days ago, the version 3.0 of Futura Time Lapse has
been released, and it comes with several important new features. Let’s discuss
what is inside this new version !
Choosing the right method for aligning images
Why are images not aligned ? There could be different reasons for that. Wind shakes on tripod ? Hand held camera ? Moving subjects that you want to align ?
Eventually, one should remember what it means in terms of geometrical transformation. For instance, gust winds if not too strong will only generate translations upon 2 ayes, X and Y. Hand held cameras will mean you need to add one rotation to the 2 translations. If you shoot a growing crop, you may need to fix the parallax. This means you need 2 translations and 3 rotations…
If you are not exactly at the same distance to the subject, you can add a scaling effect.
To summarize, it means you can align thanks to :
- 2 translations (X, Y)
- 1 rotation (Z)
- 2 more rotations (X, Y)
- 1 Scaling effect
- A mix of these…
If you think you can apply a transformation that fits all root causes, you are right in theory.
Put scale to 1, rotations upon X and Y to 0, rotation Z to 0 if you want and that’s it.
But it is better to “force” the value to 0 if they correspond to a physical phenomenon that demands the value to be equal to 0. Take again the example of wind gusts. If they don’t generate any rotation, it is better to make them equal to 0 rather than asking the app to calculate the best values without knowing these rotations should be equal to 0.
In the previous version, Futura Time Lapse was always chosing the same method (2 translations X and Y, 1 rotation Z). Now, Futura Time Lapse defines the best method that should be used but the user can always override the decision and choose between different methods.
Choosing the parts of the images that should be used as a reference for the alignment
Some parts of the images are “static”, they never move. Mountains, buildings, … Other can’t be rely on for the alignment (branches, clouds, cars, people, any moving subject). It is more efficient, by far, to rely on the parts that don’t move.
In the previous version, Futura Time Lapse was defining these regions automatically. Now, it displays the result, and the user can override / optimize the choices.
In the previous version, Futura Time Lapse was defining these regions automatically. Now, it displays the result, and the user can override / optimize the choices.
Videos
Futura Time Lapse main purpose is to align the images to create a time lapse. But it is useful to immediately check the result or/and the user may just want to create a video. Futura Time Lapse can now generate on the fly videos from the aligned images. It also proposes you to create a video from the images cropped, without any black margins (created by the transformation of the images to align them), and finally a video of a different format, 16:9 as that’s a classic format for videos.
Conclusion
Futura Time Lapse now proposes all the same features than Futura Photo Time Lapse rule but in a much more user-friendly way, and always propose a choice, that the user can override. This is making the tool very useful for both beginners and advanced users.
The 5 min demo of Futura Time Lapse version 3.0 explains how it can easily align hundreds of members of time lapses