How much data does each camera video generate every day?

How much data does each camera video generate every day?

Calculation formula

In general: video file size = (video bit rate + audio bit rate [if any]) x duration / 8

The calculation formula for 24 hours of single-channel video storage for 1 day ∑ (GB) = stream size (Mbps) ÷ 8 × 3600 seconds × 24 hours × 1 day ÷ 1024.

• Full HD 1080P (1920x1080) format: Calculated on the basis of 8Mbps code stream, the total data storage capacity for 1 day is 8Mbps÷8×3600 seconds×24 hours×(1 day)÷1024＝84GB.

• HD 720P (1280×720) format: Calculated based on 4Mbps code stream, the total data storage capacity for 1 day is 4Mbps÷8×3600 seconds×24 hours×(1 day)÷1024＝42GB.

• Standard definition 4CIF (704*576) storage format: Calculated on the basis of 2Mbps code stream, the total data storage capacity for 1 day is 2Mbps÷8×3600 seconds×24 hours×(1 day)÷1024＝21GB.

• CIF (352 288) storage format: Calculated on the basis of 512Kbps stream, the total data storage capacity for 1 day is 512Kbps÷8×3600 seconds×24 hours×(1 day)÷1024＝5.4GB.

We can also deduce in reverse: if a video file has a size of 10GB and a duration of 90 minutes (5400 seconds), then its bit rate (video + audio) is:

10 x 1024 x 8 / 5400 = 15 Mbps。

several prior knowledge concepts

Pixel and display issues probably mainly involve three concepts: pixel, resolution, and density.

1. Code stream

Bit stream size, including video bit rate and audio bit rate; video bit rate is 3Mbps, audio bit rate is 512Kbps

Therefore, for video files, the file size is mainly related to the duration, audio bit rate, video bit rate, and the number of audio streams and video streams.

For 720P and 1080P, there are various bit rates, ranging from a few Mbps to dozens of Mbps, it all depends on what encoding is used inside the video and what equipment is used to record the video.

For example, the maximum bit rate of DJI's Zenmuse H20 series:

Infrared: 2Mbps

Wide Angle: 15Mbps

Zoom: 70Mbps

Current: 32Mbps

2. Pixels

The pixels we often say generally refer to how many pixels an image contains. Let's take 720P and 1080P as examples:

The actual pixels of 1080P are 1920*1280, and the multiplication result is 2073600, that is, there are 2073600 pixels, which is often said that 1080P is 2 million pixels.

The actual pixel of 720P is 1280×720, and the multiplication result is 921600, that is, there are 921600 pixels, which is often said that 720P is 1 million pixels

The formula for calculating the space occupied by an image: image pixels * image bit depth / 8 = occupied bytes

If a picture has a pixel size of 1024*1024 and a bit depth of 32 bits, the space occupied by the picture is:

1024102432/8 = 4194304 byte = 4096 KB

3. Resolution

Resolution often refers to the device output resolution or screen resolution, that is, how many container points are there horizontally and vertically on the screen, and each container point is used to accommodate a pixel. For example, iPhone 5s has a resolution of 1136*640. It looks at 1080P, but in fact, it can only see more than 1136*640 pixels, and cannot take advantage of 1080P, and many details will be ignored.

4. Frame rate

Frame rate: FPS (Frames Per Second) how many frames per second. Affects the smoothness of the picture, which is directly proportional to the smoothness of the picture: the higher the frame rate, the smoother the picture; the lower the frame rate, the more jumpy the picture.

The frame rate generally does not exceed 24. If a bit rate is limited, such as 800kbps, the higher the frame rate, the encoder must increase the compression ratio of a single frame, that is, reduce the image quality to carry enough frames. If the video source comes from a camera, 24FPS is already the limit of the naked eye, so generally 20 FPS can already achieve a good user experience.

Source: https://www.jianshu.com/p/3252f8931dc3

5. Density

Refers to how many pixels per inch, unit PPI (Pixels Per Inch).

Screen PPI calculation method:

For example, the screen resolution of the iPhone 5s is: 1136 640, and the size is 4 inches, then the PPI is (1136 1136+640*640)/4, which is about 325.

If the PPI exceeds 300, it can be regarded as a retinal model screen, and the pixels can no longer be distinguished by the naked eye.

Suppose I have two mobile phones A and B, and the width resolution of both screens is 320*640, but the size of A is 3.5 inches, and the width of B is 7 inches, so it is obvious that the density of mobile phone A is high.

All A's images look sharper, while B's look blurrier and larger.