AutoEDL uses machine learning to recreate Edit Decision Lists (EDLs) automatically.

Zorroa’s AutoEDL is a novel machine learning solution to a major problem in the media and entertainment industry: Studios have a final cut and the raw footage for a TV series or movie—but over the years the Edit Decision Lists (EDLs) that define the editorial process have often been lost or are incomplete.

To get these older shows ready for streaming services like Netflix, Hulu, and Amazon the raw footage has to be recut to match the original edit.

These collections often contain many similar takes of the same shot. This process is normally a tedious and expensive weeks-long project for a human editor.

Simple image matching doesn’t work because the frames were modified using techniques like color correction, crops, pushes, pan-and-scan, and other post-production methods.

Original show scaled down to NTSC format

Using AutoEDL, a show can be reconstructed at high res to run on streaming services

In a few hours, AutoEDL can automatically reconstruct an EDL by locating final clips in large collections of raw footage. Since AutoEDL relies on computer vision and AI, the tasks can be parallelized and run 24 hours a day on many cloud-based computers to speed up the delivery time.

Zorroa is offering AutoEDL as a service to customers who need to reconstruct complete, accurate EDLs. Contact us to learn more about how AutoEDL can get your archive content on high def streaming services.

  • This field is for validation purposes and should be left unchanged.