Documentation

The best way to set up adaptive quantization is to keep the default value, Auto (AUTO), for the setting Adaptive quantization (adaptiveQuantization). When you do so, MediaConvert automatically applies the best types of quantization for your video content. Include this setting in your JSON job specification only when you choose to change the default value for Adaptive quantization. For this setting, keep the default value, Enabled (ENABLED), to adjust quantization within each frame based on temporal variation of content complexity. When you enable this feature, the encoder uses fewer bits on areas of the frame that aren't moving and uses more bits on complex objects with sharp edges that move a lot. For example, this feature improves the readability of text tickers on newscasts and scoreboards on sports matches. Enabling this feature will almost always improve your video quality. Note, though, that this feature doesn't take into account where the viewer's attention is likely to be. If viewers are likely to be focusing their attention on a part of the screen that doesn't have moving objects with sharp edges, such as sports athletes' faces, you might choose to disable this feature. Related setting: When you enable temporal adaptive quantization, adjust the strength of the filter with the setting Adaptive quantization (adaptiveQuantization).