PD Modes
The PD modes were developed jointly by Don Rotier and Paul Turner. 
Although
they're a little complicated, they represent a fair trade-off between 
color accuracy
and transmission time.
VIS CODES
PD50 93d (decimal)
PD90 99d
PD120 95d
PD160 98d
PD180 96d
PD240 97d
PD290 94d
COLOR MODE Y, R-Y, B-Y (See Appendix B)
SCAN SEQUENCE Y, R-Y, B-Y
NORMAL DISPLAY RESOLUTION
PD50 320x256 (all include 16-line header)
PD90 320x256
PD120 640x496
PD160 512x400
PD180 640x496
PD240 640x496
PD290 800x616
TRANSMISSION TIMES
PD50 49.7 seconds
PD90 90.0 seconds
PD120 126.1 seconds
PD160 160.9 seconds
PD180 187.1 seconds
PD240 248.0 seconds
PD290 288.7 seconds
COLOR SCAN TIMES (Y, R-Y, B-Y)
PD50 91.520ms
PD90 170.240ms
PD120 121.600ms
PD160 195.584ms
PD180 183.040ms
PD240 244.480ms
PD290 228.800ms
TIMING SEQUENCE
Note: two complete lines are shown.
(1) Sync pulse 20.000 1200hz
(2) Porch 2.080ms 1500hz
(3) Y scan (from odd line)
(4) R-Y scan averaged for two lines
(5) B-Y averaged for two lines
(6) Y scan (from even line)
Repeat until correct number of lines are transmitted for sub-mode

Appendix B Y, R-Y, B-Y Color Encoding
(Y, R-Y, B-Y) encoding is used in Robot and PD modes, as well as the 
seldomused
Wrasse SC2-120 mode. This method of encoding is also referred to as
(Y/C), (Y'U'V'), (Y'CBCR), and other names. Although all these refer 
to the same
overall method of encoding, the exact scalings used vary somewhat by
implementation and developer.
The encoding used in ChromaPIX and W95SSTV is compliant with
ITU BT. Rec. 601. (the 1953 NTSC standard) Because there has been no
clearly-defined standard, the scalings described below may vary 
slightly from
those used by other developers.
To convert non-linear RGB to [Y, R-Y, B-Y] (scaled 0-255) :
Y = 16.0 + (.003906 * ((65.738 * R) + (129.057 * G) + (25.064 * B)))
RY = 128.0 + (.003906 * ((112.439 * R) + (-94.154 * G) + (-18.285 * 
B)))
BY = 128.0 + (.003906 * ((-37.945 * R) + (-74.494 * G) + (112.439 * 
B)))
Again, as with RGB encoding, these values can be converted to 
frequency, for
SSTV transmission :
Frequency = 1500 + (v * 3.1372549)
(where `v' is the Y, R-Y, or B-Y value)
Conversely, Y, R-Y, B-Y values scaled 0-255 may be converted back to 
nonlinear
RGB:
R = 0.003906 * ((298.082 * (Y – 16.0)) + (408.583 * (RY – 128.0)))
G = 0.003906 * ((298.082 * (Y – 16.0)) + (-100.291 * (BY – 128.0)) + 
(-208.12 * (RY – 128.0)))
B = 0.003906 * ((298.082 * (Y – 16.0)) + (516.411 * (BY – 128.0)))

From:
http://www.barberdsp.com/files/Dayton%20Paper.pdf


73 de Roland PY4ZBZ


--- In digitalradio@yahoogroups.com, "Simon Brown" <[EMAIL PROTECTED]> 
wrote:
>
> Does anyone have a full description of the SSTV PD modes developed 
by Don Rotier (SK I belive) and Paul Turner?
> 
> Simon Brown, HB9DRV
>


Reply via email to