Sorry about this. What kind of footage is it? If you have 4K footage, like you have with a drone, it takes a good deal longer than HD footage. I demo with assets that are even smaller so that my analysis goes much faster. I would recommend transcoding any oversized footage to a reasonable size before attempting the effect. I hope that helps.
If you are, indeed, crashing, and have a report to share, please PM me a link to that report.
Hi my name is Sky Kennedy, and I am using the content aware fill effect for the first time jin AE. I do not have a choice that to leave the footage at 4k because what we are doing with it. Unfortunately, it looks like I will have to do some heavy Rotoscope work here as that really is my only other option.
Is Adobe addressing this issue. Your post was from April. Has anything additional been developed to address the amazingly slow working times for this effect?
Same problem here. My footage is full HD, 14 mins long. I have a decent PC, in which I use Adobe Premiere for editing wo any problem. Content-aware filling running right now for the last 6 hours, and it analysed only 9%. It would take ages for me to clear out 4 unnecessary objects from the still background. Is it normal?
14 minutes is an eternity in AE. Do you need every one of the frames in your 14-minute shot?
The proper use of any compositing and visual effects app is to only process the footage that is going to be in the final edit. More than 90% of my comps are under 7 seconds, some are only a few frames. Content ware fill was never intended to work on a 14-minute shot. We need a lot more detail to help you with your project including screenshots or a link to the video.
I'm simply removing a light switch from a wall on a 37-second clip but it's 4K footage. I created two different reference frames in Photoshop using content-aware fill. In After Effects, the reference function took about 15 minutes. The rendering stage is now about 1 frame every minute, meaning it will take 1120 minutes to render this 37-second clip.
If you are simply removing a light switch from a wall then creating a PSD from a frame, fixing the light switch in the frame, masking it, then adding it to the original shot is going to take about 1/10 the time it will take to use CAF. If the camera is moving you will have to do some motion tracking but the process is really easy.
Oh, I punted on Content-Aware Fill. It was just an experiment that failed. I went with the tried and true method of masking it out, duplicating and repositioning the under-layer and voila. It took five minutes.
Stan Arthur Manager of Video Production STETSON UNIVERSITY COLLEGE OF LAW Office of Media Services Studio: 727.562.7830 Mobile: 727.430.4674
9
Replies
AdChoices