Introduction

When dealing with situations such as ffmpeg audio modulation or pause/resume actions, an ffmpeg-like backend might not handle it effectively due to the lack of breakpoint timestamp recording. This algorithm is designed to address these issues. Additionally, it is adaptable for media that can be viewed as a sequence of frames, such as videos.

Explanation

Fundamental Concepts

1. We use the built-in CPU clock which monitors the time elapsed since the music starts playing. This clock giving the current time is called clockCurrent, hereinafter referred to as clockCR, while the clock at the moment the music starts playing is called clockStartpoint, referred to as clockSP.

2. clockStartpoint := clockCurrent is always assigned at the very moment the music starts or restarts playing. This is true for both initial play and mid-play resumption (breakpoint).

3. ps-rs is an abbreviation for pause-resume. It is used to indicate the time sequence relationship between the interface signal and the media service. The explaniation of PS-RS model is provided below.

PS-RS Model

• There are three variables in total: start_time, clockCR, clockSP. start_time is used by ffmpeg-like backends to play the audio from a specific timestamp.

• Sequence Diagram
  

• x: start_time += clockCR - clockSP

• y: clockSP := clockCR

• We then use timer diagram for further explanation.
  

Case Check

Case 1

1. pause at 5s.
2. wait for 10s.
3. resume, then pause again after another 5s.
4. wait for 10s.
5. resume.

Case 2 (Special Case)

1. pause at 5s.
2. resume immediately after pausing. (indicating a 0-second ps-rs interval)
3. pause again after 5s.
4. resume immediately after pausing again.