| Figure 1 | Screenshot of the candidate test sequence |
| Figure 2 | Video shooting configuration |
| Figure 3 | Audio capture configuration |
| Figure 4 | Gaslamp360_8192x4096_30fps_300frames_8bits.yuv |
| Figure 5 | Harbor360_8192x4096_30fps_300frames_8bits.yuv |
| Figure 6 | Kiteflite360_8192x4096_30fps_300frames_8bits.yuv |
| Figure 7 | Trolley360_8192x4096_30fps_300frames_8bits.yuv |
| Figure 8 | Community_7680x3840_29.97fps_150mbps_5mins.mp4 |
| Figure 9 | Intersection_7680x3840_30fps_150mbps.mp4 |
| Figure 10 | GoPro Video capture configuration |
| Figure 11 | Insta360™ Pro Video Capture |
| Table 1 | MD5 checksum for video sequences |
| Figure 12 | A chart showing variation in FS bit rate values for different viewports: 3 vertically and 8 horizontally adjacent positions. The FS bit rate is calculated using the advertised bit rate for the highest quality for viewport tiles and lowest quality for non-viewport tiles |
| Table 2 | Test Sequences |
| Figure 14 | An illustration of the head motion; each point represents the center of viewport at a given time, the colour lightens with the passage of time changing in the order red-orange-yellow-white. The first viewport is marked with a black diamond. Note that the grid on the sphere is shown for clarity |
| Figure 15 | The 6x4 tile grid was able to not only achieve a higher quality viewport, but was able to maintain the quality during head motion better than the 4x3 grid for most cases. Results for human and no head motion (none) are shown |
| Figure 16 | Stacked graph showing ratio of the viewport tiles at different quality levels (1-5) |
| Figure 17 | Average throughput for the different sequences and network bandwidth shown here with error bars |
| Table 3 | Total Stall Duration |
| Figure 18 | A timeline showing the adaptation of quality levels |
| Figure 19 | Symmetric (left) and HMS (right) margins |
| Figure 20 | Average MTHQD results with tile configurations of 6x4 (left), 8x6 (center) and 12x8 (right) |
| Figure 21 | Average MTHQD results for human tests |