beautypg.com

Grass Valley 8995DNC v.1.3.0 User Manual

Page 64

background image

64

8995UPC/DNC/UDX — Instruction Manual

Configuration

Edge Expansion

– select the

Enable

control to turn on the Edge Expansion

function. This can be used to enhance the visual appearance of the
edges of the image.

Note

Edge expansion is not functional with 1080p/sf 23.98.

Extended Color Space

– enabling this control doubles the dynamic range

for chroma samples while halving chroma resolution. This change
removes the least significant bit of each chroma sample and adds one
more most significant bit.

Vertical Edge Filter

– enabling this control applies horizontal low-pass fil-

tering to the luma and chroma samples at the video image vertical
edges to lessen the black-to-video matte-to-video distortion that occurs
in the scaler function.

Matte Edge Border

– enabling this control applies a 3 pixel wide mask to

vertical image edges matching the brightness and hue of the vertical
bars. These masks are applied to lessen the transition distortion
between matte and video that occurs in the scaler function.

Cadence Controls

The 8995 module has automatic cadence detection on the input signal. A

Cadence

control is provided to set one of two modes for signals:

Auto

Film/Video (3:2 or 2:2 pull down)

, the default,

or

Fixed-Video (2:2 pull down)

In

Auto Film/Video (3:2 or 2:2 pull down)

mode, the scaler in the 8995 module

detects the pattern of video field updating in the input video and performs
motion-adaptive de-interlacing based upon that pattern. Normal video has
a video field updating pattern of 2:2, meaning one frame of video consists
always of two fields captured at 1/59.94 Hz intervals.

Video created from film that has a 1/23.98 Hz capture interval must use a
process called 3:2 pull down to produce video fields at 1/59.94 Hz inter-
vals. The first frame of film may be repeated twice, and the second frame
may be repeated three times to give a repeating sequence of 5 video fields
for every 2 frames of film.

Typically, the sequence is defined as 2:3:2:3 over a span of 5 video frames,
or four film frames. Sometimes a portion of a video program with 2:2
cadence is mixed with other video data having a 3:2 cadence. The 8995
module scaler may detect one cadence type and de-interlace assuming that
cadence type for the whole image, or it may oscillate between the two
cadence types. This may produce disturbing artifacts in the video.

If artifacts occur in

Auto Film/Video (3:2 or 2:2 pull down

mode, the user may

disable automatic cadence detection by selecting the

Fixed Video (2:2 pull

down)

mode, forces the de-interlacer to perform motion-adaptive de-inter-

lacing assuming only a 2:2 cadence in the video, reducing the chance of dis-
tortion related to cadence detection. The result may slightly reduce
de-interlacing quality for 3:2 cadence video.

This manual is related to the following products: