Barco TransForm A, Руководство пользователя

Страница: 85 / 202

4. Operating
Barco – TransForm A - Workstation for Windows – DOC-3266-2 – Revision 16 – August 2007
_____________________________________________________________
4-5
1) SECAM only with Q
UAD
A
NALOG
V
IDEO
C
ARD
, D
UAL
DVI I
NPUT
C
ARD
and S
TREAMING
V
IDEO
C
ARD
2) Overall configuration has to be considered
3) For Q
UAD
A
NALOG
V
IDEO
C
ARD
or S
TREAMING
V
IDEO
C
ARD
and D
UAL
DVI I
NPUT
C
ARD
respectively
To calculate the bandwidth for windows of different sizes or frame rates the following formula can be used:
B = res
x
* res
y
* fr * cd
The abbreviations have the following meaning:
•
res
x
, res
y
– resolution of the source in x- and y-direction in pixels
It is important to consider all scaling factors used to display the window. In general up-scaling is performed
in the O
MNI
S
CALER
therefore data is transferred 1:1 and the original resolution must be inserted into the for-
mula. If the source is displayed smaller than 1:1 the source is down-scaled in the input card. Only this re-
duced number of pixels is transmitted to the O
MNI
S
CALER
and must be considered. The original size of the dif-
ferent video types can be read from the table above.
•
fr – frame rate
For video a frame rate of 25fps (PAL/SECAM) and 30fps (NTSC) is standard. For dynamic RGB input a frame
rate of 20 fps in normal cases gives an acceptable performance. If the RGB-source is mostly static even lower
frame rates can be selected.
•
cd – color depth
For all video windows with the default Pixel format YUV 4:2:2 the factor is:
cd = 2
For RGB windows two different Pixel formats may be useful:
Pixel format = RGB 5:6:5 cd = 2
Pixel format = RGB 8:8:8 cd = 3
With the average bandwidth of the O
MNI
B
US
A12 of 200 MBps per input card the sum of the bandwidth demands
of the windows provided by one input card should not be bigger than this value. Anyhow also higher bandwidth
usage might be possible but the overall configuration should be considered (In case of doubt contact the Barco
support; please refer to section
8.3 Contact .) :
200 MBps = B
1
+ B
2
+ … + B
n
With the effective bandwidth of the O
MNI
B
US
A18 of 400 MBps the sum of the bandwidth of the individual win-
dows may not be bigger than 400 MBps to provide optimal performance:
400 MBps = B
1
+ B
2
+ … + B
n
In a Processor configuration with 100 MBps the following correlation must be considered:
100 MBps = B
1
+ B
2
+ … + B
n
With a single Processor
AGS-3390-1/-2
with 200 MBps the following correlation must be considered:
200 MBps = B
1
+ B
2
+ … + B
n