Skip links

How to Prep Handheld Footage for Software Stabilizing When it Contains Photography Flashes

As an editor working with a wide array of footage, you’re probably bound to find some stuff come in that was handheld or not stabilized via gyro/steadicam/jib or tripod.  The good news is rather than applying the use of a tracker, editors can take advantage of the built-in stabilization filters.  They work tremendously well (provided your footage doesn’t have too much depth variation in the Z axis, ie: active zooming).  But one instance where they could have trouble in analyzing your source footage is if it contains flashes.  That’s because flashes like those from a still camera are a sudden burst of light.  Today’s CMOS cameras make it very obvious because the frame becomes divided with different levels of exposure (for more about that look up rolling shutter or flash banding).

 

Here is one way you might be able to workaround the issue, should you encounter it in your own footage.

 

This particular shot had the potential to look real nice, but it needed some work.

 

[ORIGINAL]

 

Applying some NLE stabilization won’t work as well as anticipated due to the camera flash

 

[STABILIZATION FILTER APPLIED]

 

To get around this take the original media file and cut the frames containing the flashes out.  The result will feature a very minimal jump cut.  With the flashes removed, re-exported the footage, bouncing it back into your project’s bin as a single clip.  Why?  Because, for starters, your NLE’s stabilization filters analyze the entire clip it is applied to–not just the section or subclip you might have marked in/out and placed into your timeline.  Creating a new clip, free from any photography flashes ensures there will be no abrupt lighting changes that those filters could mistake for a bump in framing.  Place the clip in your timeline, apply any color correction settings along with the stabilization filter and check out the resulting clip below.

 

[FINAL VERSION]

Leave a comment


This site uses Akismet to reduce spam. Learn how your comment data is processed.