Random Access Point Synchronization in V-PCC Content Files
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for random access in ISOBMFF content files, particularly in V-PCC content, result in significant overhead due to the need for identifying random access points (RAPs) in each component stream, leading to delayed start of reproduction and poor user experience.
Innovation Solution
The proposed solution involves a file generating apparatus and client apparatus that align RAPs across component streams by ensuring decoding timestamps (DTS) and composition timestamps (CTS) are synchronized, allowing for the identification of RAPs in one stream to specify aligned RAPs in others, thereby reducing overhead and enabling prompt random access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If random access point identification is performed on each component stream separately, then random access functionality is achieved, but overhead increases significantly and reproduction start is delayed
Solution Approach 1:
The patent merges the random access point identification process across multiple component streams by introducing a synchronization mechanism. Instead of identifying RAPs independently in each stream (attribute, geometry, occupancy, metadata), the system synchronizes their RAP identification through shared synchronization information, allowing simultaneous access points to be identified across all streams at once, thereby reducing overall overhead and delay.
Solution Approach 2:
The patent applies preliminary action by pre-establishing synchronization information and RAP markers during the encoding phase. The file generating apparatus pre-aligns RAPs across component streams using synchronization timestamps and markers, so that during playback, the reproducing apparatus can directly access synchronized RAPs without performing separate identification processes for each stream, thus reducing start delay.
2Ease of operation
If random access point identification is performed on each component stream, then random access is enabled, but processing overhead increases
Solution Approach 1:
The patent implements universality by creating a shared RAP identification mechanism that serves all component streams simultaneously. The synchronization information structure acts as a universal reference that can be applied across attribute, geometry, occupancy, and metadata streams, allowing a single RAP identification process to enable random access for all streams rather than requiring separate dedicated processes for each.
Solution Approach 2:
The patent combines multiple RAP identification operations into a unified process by using synchronization timestamps and markers that are common across all component streams. This merging approach allows the system to identify RAPs for multiple streams in a single processing pass, significantly reducing the cumulative overhead that would result from independent identification processes.
3Reliability
If decoding timestamps and composition timestamps are synchronized across streams, then RAP alignment is achieved, but encoding complexity increases
Solution Approach 1:
The patent applies parameter changes by adjusting timestamp parameters (decoding timestamps and composition timestamps) to be synchronized across component streams. The encoding apparatus modifies these time parameters to maintain consistency, ensuring that RAPs occur at the same logical time points across all streams. This parameter alignment achieves reliable RAP synchronization while the structured approach to timestamp management keeps encoding complexity manageable.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided are an information processing apparatus, a reproduction processing apparatus, an information processing method, and a reproduction processing method that can easily provide a viewing experience that meets the needs of a viewer. A data acquiring unit (101) acquires a content including a plurality of component streams. A file generating unit (105) generates content random access point (RAP) information indicating random access points of the content on the basis of random access points of the component streams of the content that has been acquired by the data acquiring unit (101).