Calculate FX3/CX3 latency in Video Streaming Design - KBA223271
Employee
Apr 02, 2018
04:33 AM
- Subscribe to RSS Feed
- Mark as New
- Mark as Read
- Bookmark
- Subscribe
- Printer Friendly Page
- Report Inappropriate Content
Apr 02, 2018
04:33 AM
Version: **
Translation - Japanese: ビデオストリーミングデザインでのFX3/CX3のレイテンシ計算 - KBA223271 - Community Translated (JA)
Question:
How can we roughly calculate the latency introduced by FX3/CX3 in the video streaming design?
Answer:
The latency can be estimated roughly using the information in the following table.
Device | MIPI Block | FX3 48KB Buffer Filling Time | USB (~ 300 MBps) | Total Latency |
FX3 | NA | 614.4 µs | 48KB/300=160 µs | 774.4 µs |
CX3 | 30us | 614.4 µs | 48KB/300=160 µs | 804.4 µs |
The results in this table are obtained based on the following:
- CX3: input:1080p@30 fps, RAW10 output: USB3.0. The latency time is the time between the input and the output in this case.
- FX3: input:1080p@30 fps, 32-bit parallel output: USB3.0. The latency time is the time between the input and the output in this case.
Note:
- We considered 1080p@30 fps for calculation in this case. Each line will take 33.33ms / 1080=30μs.
- The number of lines for the 48-KB buffer should be (48*1024) / (1920*10/8) = 20.48
- Buffer filling time should be line time* No. of lines in a buffer = 30*20.48=614.4μs
For other resolutions, frame rates, and buffer size configuration, the results are different. However, it can be calculated in the same way as shown above.
Labels
Rate this article: