4
The reference relationship of the Single P (Normal P) mode is simple. Each P-frame uses
one forward reference frame.
The Single P (Normal P) mode can be used in any scenarios.
Smart P:
In Smart P mode, the P-frame uses the IDR frame (long-term reference frame) and forward
reference frame (short-term reference frame) as reference frames. The time domain
correlation of two reference frames is used to improve the encoding compression performance.
This mode is mainly used in the monitoring scenario.
In the monitoring scenario, the camera is fixed at a position, and the human and objects in the
scenario may be static or moving.
In static regions, the time domain correlation of the long-term reference frame and the
current frame significantly reduces the bit rate and respiratory and smearing effects.
In motion regions, motion estimation is performed by using the short-term reference
frame. In Smart P mode, the IDR frame interval is prolonged, and the virtual I-frame is
inserted periodically, which significantly reduces the bit rate by 30% to 50% in the
monitoring scenario and improves the picture quality. However, this mode is not
applicable to scenarios in which the camera can move.
Dual P:
SP indicates a special P-frame (also called an SP frame). It is recommended that the QP
value of the SP frame be less than that of any other P-frame. If u32SpInterval is 0, the
SP frame is not supported.
In Dual P mode, the P-frame uses the nearest two forward reference frames. Using the
time domain correlation of more reference frames enhances the encoding compression
performance.
This mode is mainly used in motion scenarios with low delay requirements.
The compression performance of the Dual P mode is lower than that of the BipredB mode
but higher than that of the Single P mode. The Dual P mode features no encoding and
decoding delay because the two reference frames are forward reference frames.
This mode is suitable for “Sport channel” which is with “fast moving” media content.
