Video Raw Meta Data

This page specifies the meta-data for raw video recordings.

If a stream has the content-type VideoRaw, we recommend that meta-data about the stream adheres to the structure and naming laid out in the following. While most meta-data is optional, at least fields width, height, and color_format under encoding must be present. Colors are not compressed, i.e., one color channel corresponds to a stream channel (e.g., an RGB color pixel takes up 3 channels).

<encoding>           # specification of how each sample (frame) is encoded
  <width>            # width of each sample, in pixels; must be present
  <height>           # height of each sample, in pixels; must be present
  <color_channels>   # number of channels per pixel (1 for grayscale, 3 for RGB or YUV, 4 for RGBA)
  <color_format>     # the color format; typical values are GRAY (grayscale), RGB (red-greed-blue, in that order), RGBA (RGB with additional alpha channel), or YUV (for Y´CbCr); must be present
  <color_space>      # color space that the video is in, e.g. sRGB; assume sRGB for RGB format if missing
  <codec>            # should be present, assume RAW if missing
</encoding>

</display>           # specification of how the images shall be displayed
  <origin>           # coordinate origin (can be top-left,bottom-left, or, rarely, top-right or bottom-right); assume top-left if missing
  <pixel_aspect>     # aspect ratio (x/y) of the pixels; assume 1.0 if missing
  <x_dpi>            # horizontal resolution of the content, in dots per inch
  <y_dpi>            # vertical resolution of the content, in dots per inch
</display>

<acquisition>       # information on how the data was acquired
  <manufacturer>     # manufacturer of the acquisition device
  <model>            # model of the acquisition device used
  <serial_number>    # serial number of the acquisition device
  <view>             # view of the camera (e.g., left-eye for a view of the left eye of the subject in an eye tracker)
  <distortion>       # distortion parameters of the camera used (using Brown's distortion model)
    <cx>             # distortion center in x direction
    <cy>             # distortion center in y direction
    <fx>             # focal length in pixel units
    <fy>             # focal length in pixel units
    <p1>             # tangential distortion coefficient 1
    <p2>             # tangential distortion coefficient 2
    <k1>             # radial distortion coefficient 1
    <k2>             # radial distortion coefficient 2
    <k3>             # radial distortion coefficient 3
  </distortion> 
</acquisition>

<content>            # information about the video content
  <copyright>        # copyright statement 
</content>