Menu
Amazon Elastic Transcoder
Developer Guide (API Version 2012-09-25)

Clip Stitching

You can configure Elastic Transcoder to stitch together parts, or clips, from multiple input files to create a single output. This allows you to create a single output from different inputs, such as separate recording sessions or highlights from a game.

To combine clips from multiple inputs, specify multiple inputs when you create a job. Elastic Transcoder stitches clips together in the order that you specify in the job request. If you want to transcode multiple inputs without stitching them together, create a separate job for each input.

For each input that you specify, you can specify a start time and a duration, which allows you stitch together only the parts of each input that you want to include in the output. If you want to clip the beginning and the end of the output, specify clip settings for the first input and clip settings for the last input. For example, if you want to remove a trailer at the beginning of the output and credits at the end of the output, specify a start time and duration for the first input that removes the trailer, and a start time and duration for the last input that removes the credits.

Your files must share consistent audio and video settings so that Elastic Transcoder can stitch them together. For example, if your inputs have different video aspect ratios, your preset should use either a padding policy or a sizing policy that allows Elastic Transcoder to produce a consistent aspect ratio for the output. For more information about how these policies affect aspect ratios, see Sizing Policy and Aspect Ratios. Also, either all input must have video or no inputs should have video, and either all inputs must have audio, or no inputs should have audio. Likewise, caption settings must be consistent. Captions of the same language across inputs must specify the same value for language or they are not considered to be part of the same caption track.

The placement of watermarks is determined by whether your preset has a content target or a frame target. A content target means that your watermark is placed according to the content of your original input, ignoring any padding added by the padding policy. For example, if one clip doesn’t need padding and the watermark is in the top left, then the watermark for that clip is placed a specified distance away from the edge of the video. If a clip does need padding, then the watermark is placed a specified distance away from the edge of the padding. A frame target means that your watermark is placed according to the content of your output. It is always placed a specified distance away from the edge of the video. For more information about watermarks, see Watermark Settings.

Elastic Transcoder creates thumbnails after it has stitched together all of the inputs.