Early-Decodable Frame Transitions for Faster Media Source Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing media delivery systems experience significant delays during channel or source changes due to buffering and decoding delays, leading to a poor user experience, especially with advanced video compression standards and flexible picture structures.
Innovation Solution
Implementing a client-side solution that manages video and audio decoding independently, utilizing early-decodable frames to generate visual effects and minimize perceived delay by displaying relevant content during transitions, such as fade-out, fade-in, or superimposition, until a normal play state is achieved.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If traditional buffering and decoding processes are used for media source changes, then video decoding completeness is ensured, but channel change time increases significantly (2-3 seconds or more)
Solution Approach 1:
The system performs preliminary identification of early-decodable frames (such as I-frames or key frames) before the complete decoding process finishes. By preparing these frames in advance during the decoding process, the system can immediately display them when a channel change is requested, eliminating the need to wait for the entire 2-3 second buffering and decoding process to complete.
Solution Approach 2:
Instead of waiting for complete video decoding (100% action), the system accepts and displays a partial decoded state by utilizing early-decodable frames that are available sooner. This partial action approach reduces channel change time from 2-3 seconds to a much shorter duration, trading off some decoding completeness for speed while maintaining acceptable visual quality.
2Manufacturing precision
If complete video buffering is performed before playback, then video quality is maintained, but initial display delay increases (about 1500 ms for initial video buffering)
Solution Approach 1:
The system performs preliminary decoding of key frames (I-frames) and stores them in a buffer before they are needed for display. When playback or channel change occurs, these pre-decoded frames are immediately available for display, eliminating the 1500 ms initial buffering delay while maintaining video quality through proper use of compression standards.
Solution Approach 2:
The video stream is segmented into independently decodable frames (I-frames, P-frames, B-frames). The system identifies and prioritizes decoding of I-frames which are self-contained and can be displayed independently, rather than waiting for complete sequence decoding. This segmentation allows immediate display of usable frames while reducing initial display delay.
3Reliability
If random access point acquisition is performed during channel changes, then decoding synchronization is achieved, but channel change time increases (about 250 ms for RAP acquisition)
Solution Approach 1:
The system performs preliminary identification and buffering of random access points (RAPs) and key frames before channel changes are requested. By having these synchronization points pre-identified and ready in the buffer, the system eliminates the 250 ms RAP acquisition delay during channel changes while maintaining decoding synchronization through proper use of these pre-positioned reference frames.
Data Source
AI summary
Methods and systems are described for generation of a smooth, engaging transition from one media source to another or from one portion of a media source to another portion of the media source. In response to a request to change display of content from a first source to a second source, a first decodable frame of the second source is determined before completion of decoding a first segment of the second source. During the decoding of the second source, a transition based at least in part on the first decodable frame of the second source is generated for display. In response to completion of the decoding of the first segment of the second source, the second source is displayed without the transition. Artificial intelligence systems, including neural networks, and models are utilized to improve the transition. Related apparatuses, devices, techniques, and articles are also described.