DDZ40(10)(20)-YY/HS/HD
www.dallmeier.com
74
GoP-size
(only with H.264)
The H.264 coding is carried out by dividing the video stream into so-called GOPs (Group
of Pictures) of a defined length (defined GOP-size).
A GOP sequence always starts with an Intra-Frame (I-Frame), which contains all image
data and serves as a reference for the subsequent images within a GOP.
The I-Frame is compressed with a low compression rate, similar to the JPEG compres-
sion method.
Depending on the defined GOP-size, an I-Frame is followed by one or more Predicted
Frames (P-Frames), which only contain the motion predictions and difference information
about the preceding images (I-Frame or P-Frames) (Long-term prediction).
The compression rate with P-Frames is much higher than with I-Frames since changes
in relation to reference images need to be coded as motion vectors only. The required bit
rate thus decreases so that, with a given total encoding bit rate, more bits are available
for the I-Frame. This means that the quality of the I-Frame can be increased, for example,
the detail resolution in case of a larger GOP size.
However, if there are scenes with many motion changes, a high number of P-Frames
can have a negative effect on the image quality, because the motion predictions become
increasingly inaccurate.
Additionally, a larger GOP-size always leads to an increase in delays regarding process-
ing or accessing a stream.
The GOP sequence ends before the next I-Frame.
At a later stage, the visible single frames are generated at the decoder, using the indi-
vidual GOPs.
The GOP-size 1 (I-Frames only) indicates a low compression factor and should only be
used with specific applications, because the bandwidth requirements increase signifi
-
cantly.
Generally, a GOP-size of between 6 and 15 will provide a good image quality with suf-
ficiently high compression.
If large GOP-sizes are defined, reverse playback can result in frame drops with some
decoders.
Deinterlacing
The aim of this function is to enhance the image quality by computationally eliminating the
so-called interlace effect, which may occur after merging two fields into a frame, from the
frame.
However, whether a higher quality can be achieved depends on many factors such as the
speed of the motions in the image or the type of output device on the decoder side (digital
or analogue monitor).
It is therefore not possible to provide a general recommendation to activate or deactivate
this function. Therefore, we do recommend that you test the output on the monitor/de-
coder after the function has been configured.
For this purpose, select a playback scene typical for the camera concerned. If the inter-
lace effect can be detected with moving objects i.e. if moving edges are slightly displaced
between two fields, this effect can be reduced by enabling “Deinterlacing”.
However, the playback thus may be jumpy. In this case disabling this function may, in
certain circumstances, result in a smoother playback.
