Cloud Transcode Open source transcoding stack.

• Home
• Get Started
• What now?
• Activities
• Basics
• Specs
• output_asset
• input_asset

sponsored by

Sport Archive Inc.

Here we describe the type of output you can generate with CT. We describe the JSON data you must send to the stack in the output_asset section of the JSON payload you send to the TranscodeActivity.

Navigation Menu

1. Common attributes

   All transcoding requires common attributes regardless of their types.

2. Video output

   Use this syntax to transcode a video to another video. We use ffmpeg command for this.

3. Presets

   Presets allow you to easily describe transcoding prameters in a file and ask the Transcoding Activity to read it and pass it to FFMpeg.
   

4. Watermark

   If you want to overlay a watermark image on top of your video. Use those options
   

5. Thumbnails output

   Use these options to output thumbnails from a video. We use ffmpeg command for this.

6. Image output

   Use these options to output an image from another image. image to image transcoding. We use convert command for this.

Common attributes

Those options are common to all output types.

{
...
    "output_asset": {
            "type": "VIDEO",
            "bucket": "ClientA-bucket-out",
            "file": "/output1/test1_sd.mp4",
            "s3_rrs": true,
            "s3_encrypt": true,
...

These are the common JSON fields describing any output files of any types.

The output_type field is dependent on the input_type field. You cannot convert a VIDEO into a DOC for example.

Video Output

You can use presets:

{
...
    "output_asset": {
            "type": "VIDEO",
            "bucket": "ClientA-bucket-out",
            "file": "/output1/test1_sd.mp4",
            "s3_rrs": true,
            "s3_encrypt": true,
            "preset": "720p-generic.json",
            "watermark": {
                  "bucket": "ClientA-bucket-in",
                  "file": "/watermark/logo.png",
                  "size": "75.2:28.4",
                  "opacity": 0.2,
                  "x": -20,
                  "y": -20
            }
     }
 }

Or you can execute arbitrary FFMpeg commands:

{
...
    "output_asset": {
            "type": "VIDEO",
	    "bucket": "ClientA-bucket-out",
	    "file": "/output1/video1.mp4",
	    "s3_rrs": true,
	    "s3_encrypt": true,
	    "custom_cmd": "ffmpeg -i ${input_file} -c:v libx264 -preset slow -crf 22 -c:a copy ${watermark_options} ${output_file}",
	    "watermark": {
	    	 "bucket": "ClientA-bucket-in",
	    	 "file": "/no-text-96px.png",
	    	 "size": "96:96",
	    	 "opacity": 0.2,
	    	 "x": -20,
	    	"y": -20
	    }
     }
 }

There are the JSON fields describing VIDEO output files when using Presets. You can override presets values within the JSON.

Presets

{
    "name": "System preset: Generic Apple TV 3G",
    "description": "Generic transcoding preset for: Apple TV 3G, Roku HD/2 XD",
    "size": "1920x1080",
    "frame_rate": 30,
    "video_bitrate": "5000k",
    "audio_bitrate": "160k",
    "video_codec": "libx264",
    "audio_codec": "libfdk_aac",
    "video_codec_options": "MaxReferenceFrames:3,Profile:high,Level:4"
}

The presets can contain the following entries but can be overriden by providing the same entry in your JSON payload:

A list of pre-made presets can be found in the CT project and here: https://github.com/sportarchive/CloudTranscode-FFMpeg-presets

Watermark

...
    "watermark": {
        "bucket": "ClientA-bucket-in",
        "file": "/watermark/logo.png",
        "size": "75.2:28.4",
        "opacity": 0.2,
        "x": -20,
        "y": -20
    }
...

Use this object within a VIDEO output description to overlay an image on top of the video.

You can control many aspects of that image using the options below.

Thumbnail Output

Generates a thumbnail from a precise location in the video (in seconds)

{
...
    "output": {
            "type": "THUMB",
            "mode": "snapshot",
            "bucket": "ClientA-bucket-out",
            "file": "/output1/thumbs/test1_thumb.jpg",
            "s3_rrs": true,
            "s3_encrypt": true,
            "size": "160:120",
            "snapshot_sec": 5
    }
}

Generates N thumbnails at each intervals (in seconds)

{
...
    "output": {
            "type": "THUMB",
            "mode": "intervals",
            "bucket": "ClientA-bucket-out",
            "file": "/output1/thumbs_interval/test2_thumb.jpg",
            "s3_rrs": true,
            "s3_encrypt": true,
            "size": "-1:159",
            "intervals": 5
    }
}

These are the output attributes you must provide to generate thumbnails from a video. Thumbnails are an output_type on its own.

Image Output

Generates an image from another image.

{
...
    "output": {
            "type": "IMAGE",
            "bucket": "ClientA-bucket-out",
            "file": "/output1/images/image.*",
            "s3_rrs": true,
            "s3_encrypt": true,
            "resize": "180x180",
            "quality": 70,
            "crop": "600x600+0+0"
    }
}

These are the output attributes you must provide to generate a new image from another image.

Custom command

Both VIDEO and IMAGE transcoding can accept a custom command so you can run the command you want even if it’s not support by the default JSON format.

You can craft your own custom command and CT will execute it for you. In order to reference the input file, watermark options and the wanted output file, three keywords are available to use in your command line. Those keywords will be automatically replaced by CT at runtime by the proper value.

• ${input_file} : Will be replaced by the location on the local filesystem of your input file specified in the input_asset section. The file referenced in the input_asset section is downloaded on the local disk in /tmp.
• ${watermark_options} : Only usable for VIDEO output. This will be replaced by the proper options to overlay a watermark on your video. Just specify a watermark section in the output_asset section (see below for details) and CT will add the proper options and will download the image for you.
• ${output_file} : This will be replaced by the location and name of the wanted file. It will be located in /tmp. The resulting file will be uploaded for you to the desired S3 location.

For examples, see the client_example/input_samples folder in the CPE project. There are example JSON files there that illustrate the different options you have.