0

Is it possible to create a secondary layer for a portrait video with AWS MediaConvert, of the same video but blurred, and that fills the frame?

I'm trying to achieve this result.

enter image description here

somonek
  • 373
  • 1
  • 6
  • 13

1 Answers1

1

There is no blur effect within the service at this time, so it would require a separate operation on the source file. This could be automated with AWS Lambda.

One method to generate the desired background image could use the Image magick Lamba layer found here: https://serverlessrepo.aws.amazon.com/applications/us-east-1/145266761615/image-magick-lambda-layer

The Lambda described above would generate a background image which you can then use as an additional input to your MediaConvert job. This process could be automated and triggered via Lambda and S3 File Events (new file arrival event).

aws-robclem
  • 324
  • 2
  • 5
  • Thanks @aws-robclem, this is a good suggestion to get closer to a solution. I'm looking to have the blurred layer being the video itself, so also playing along with the main one, but with the blur effect and in the background. It's a common patter seen in TV news, and some social medial platforms these days. I will look into the lambda layer for now. Thanks – somonek Jun 07 '23 at 12:01
  • You can try the BoxBlur filter in ffmpeg to generate a second video asset from your main asset; then add both assets to your MediaConvert job. There is a free ffmpeg layer for lambda available to automate this step in conjunction with S3 file events. – aws-robclem Jul 05 '23 at 15:27