MediaPackage CMAF vs MediaConvert CMAF

Question

Hello, 
im researching vod solution with AWS. From what I understood to support multiple devises I can default to cmaf. So I did setup MediaConvert job to produce cmaf and (alternatively I used System-Ott_Cmaf_Cmfc_Avc_Aac_Sdr_Cbr job template). Produced outputs are ok, I was able to play them on some hls player. But Im struggling with next step - using Media Package. I set up Packaging group with CMAF and Ingessting asset (source is result of cmaf MediaConvert ) is failing. When I produce with MediaConvert hls and in MediaPackage use CMAF packaging group it works fine. Im bit lost why. Any ideas?

Mike-ME · Answer

AWS MediaPackage supports HLS and SMIL/MP4 inputs for VOD applications.
Please see the details here.
https://docs.aws.amazon.com/mediapackage/latest/ug/supported-inputs-vod.html
CMAF outputs from AWS MediaConvert are fragmented MP4 files. SMIL/MP4 inputs to MediaPackage should not be fragmented.
Perhaps consider storing the transcoded assets as HLS segments and using MediaPackage to create CMAF outputs.

For Live streams, MediaPackage supports HLS inputs.
https://docs.aws.amazon.com/mediapackage/latest/ug/supported-inputs-live.html

robc_amzn · Answer

^^ what Mike said.  
Shortest workflow is to make simple HLS from MediaConvert to s3 bucket.  You can automate this via S3 file events (new file arrival)  and AWS Lambda.

Step two is to repackage that new HLS file group as CMAF, MSS or DASH using MediaPackage-VOD.  Encryption is optional.

MediaPackage CMAF vs MediaConvert CMAF

回答を追加する

関連するコンテンツ