Working with progressive and interlaced scan types in AWS Elemental MediaConvert
Progressive and interlaced are two types of video display methods. Modern display devices detect whether a video is interlaced or progressive and automatically play back the video correctly. But, progressive video looks much better on modern screens.
To get the best results with using interlacing/deinterlacing and converting to and from telecine, you must consider how your input video was recorded and what transformations have been done to it. For example, when you apply deinterlacing to an input that is not interlaced, your output video quality suffers.
Topics
Basic scan type vocabulary
- Progressive video
-
Progressive video includes all lines in all frames. It looks better on modern screens because it drastically reduces the amount of image flicker the viewer sees on the screen. Devices displaying progressive video will re-draw all horizontal lines in a frame. For example, a device running at 50 Hertz playing a 1080 progressive video re-draws 1080 lines (every line in the frame) 50 times per second.
- Interlaced video
-
Interlaced video uses a technique that doubles the perceived frame rate of a video display without consuming extra bandwidth. On older displays, most people won't notice decreased video quality with interlaced video. Devices that support interlaced video re-draw every other horizontal line in a frame. For example, a device running at 50 Hertz playing a 1080 interlaced video redraws 540 lines (half of the lines in the frame) 50 times per second.
- Field polarity for interlaced frames
-
Interlaced video contains two fields of a video frame, each one made up of every other horizontal line the image. Field polarity in video distinguishes between these two sets of lines. A set's polarity indicates whether the top field comes first or bottom field comes first. In the following illustration, the set with top field polarity is shown in blue and contains the topmost line. The set with bottom field polarity is shown in red and contains the second horizontal line from the top. The complete frame contains both, with each set being refreshed alternately.
When you create interlaced outputs with MediaConvert, you can specify which field polarity comes first with the setting Interlace mode.
Settings for scan type conversion
To convert between interlaced to progressive video, specify the MediaConvert settings covered in this topic. This topic offers conceptual information and guidance for choosing values for the MediaConvert settings related to interlacing and deinterlacing. For directions for specifying them, see the procedures in the topic Configuring scan type conversion.
Valid values for some of these settings depend on what you choose for the other settings. For a table that shows how to specify them together correctly, see Requirements.
- Deinterlacer preprocessor
(Deinterlacer
) -
Use this parent setting to enable and disable deinterlacing. If you simply enable the deinterlacer without specifying any further deinterlacing settings, your job will convert interlaced content to progressive. For the default deinterlacing to work correctly, your input video must be interlaced and the frames of your input video must not have metadata that tags them as progressive.
- Deinterlace Control
(
DeinterlacerControl
) -
This setting is a child of the deinterlacer setting. You can optionally use Deinterlace control to have MediaConvert deinterlace all frames of your input video, including those that are tagged as progressive. Only use this setting when you know that this metadata in your input video is wrong.
- Deinterlace algorithm
(
DeinterlaceAlgorithm)
-
This setting is a child of the deinterlacer setting. You can optionally use Deinterlace algorithm to specify the way that MediaConvert does the deinterlacing to get the best quality for your content. For sharper pictures, choose one of the motion adaptive interpolation options (Interpolate or Interpolate ticker). For smoother motion, choose one of the blend options (Blend or Blend ticker). When your source file includes moving text, such as a scrolling headline at the bottom of the frame, choose the ticker version of the algorithm.
- Deinterlace mode
(
DeinterlacerMode
) -
This setting is a child of the deinterlacer setting. You can optionally use Deinterlace mode to modify how MediaConvert applies deinterlacing.
Keep the default value, Deinterlace, to do regular deinterlacing.
Choose Inverse telecine to convert hard telecine (29.97 fps, interlaced) to progressive video at 23.976 fps. When you use inverse telecine, you must still specify your output frame rate as 23.97. MediaConvert doesn't automatically set this.
Choose Adaptive to have MediaConvert automatically detect interlaced inputs and apply deinterlacing and inverse telecine to them. Adaptive deinterlace mode is useful when you use output presets, job templates, or custom programming to apply the same job settings to transcode an entire library of assets.
Note
When you choose Adaptive for this setting, MediaConvert automatically uses inverse telecine as well.
- Interlace mode (
interlaceMode
) -
When you create interlaced video, from either progressive or interlaced inputs, use this MediaConvert setting. The default value of this setting is Progressive, so you can ignore this setting unless you want an interlaced output.
Note
When you use an interlaced input and you keep the default setting, Progressive, for Interlace mode, you should also enable Deinterlace. Otherwise, your progressive output will have very poor video quality.
When you create interlaced outputs, use Interlace mode to specify the field polarity of your outputs. You can either directly specify the field that comes first, or you can set it to follow the polarity of the source input. For jobs that have multiple inputs, the output might have a mix of top and bottom field first, depending on the polarity of the inputs.
When you set Interlace mode to follow the source and your input is progressive, the output's field polarity depends on which of the follow options you set. Follow, top field results in an output that's top field first. Follow, bottom field results in an output that's bottom field first.
- Scan type (
inputScanType
) -
Use this setting only when your input is progressive segmented frame (PsF). MediaConvert automatically detects progressive and interlaced inputs. But it doesn't detect PsF. When your input is PsF, set Scan type to PsF for better preservation of quality when you do deinterlacing and frame rate conversion.