home *** CD-ROM | disk | FTP | other *** search
- SB ALL@AMSAT < DB2OS $GGJA-4173-86
- WO-18: Picture protocol (corrected version)
- R:900302/0615z 43586@W3IWI.MD.USA [Balto/Wash MD/DC/VA/DE] Z:21029
- R:900302/0545z 33653@WA3ZNW.MD.USA.NA [Wheaton,Md] Z:20906 N:MCBBS
- R:900302/0531z 12386@K4NGC.VA.USA.NA [Woodbridge, Va] Z:22193
- R:900301/0017z 50847@N4QQ [Silver Spring,Md] Z:20901
- R:900301/1953z 16180@N4QQ-1 [McLean, VA USA] Z:22101
- R:900301/1545z 14544@GB7LDI [JO02ON]
- R:900225/1024z 41768@4X1RU
- R:900225/0758z @DK8AT []
- R:900224/2337z @DK0MAV [NORD><LINK PEINE, JO52CI, TheBox 1.6 OP: DF3AV]
- de DB2OS @ DK0MAV
-
- Posted: Wed, Feb 21, 1990 1:49 AM GMT
- From: WEBER
- To: amsat
- Subj: correct protocol
-
- I have been told there is an erroneous description of the picture
- protocol published on Compuserve. To avoid the possibility of people
- wasting their time with it, below is the correct, complete description.
-
- The format of picture data for the camera color composite signal is
- as follows:
-
- Callsign (for this transmission scheme) is PHOTO (for the camera).
- SSID is picture number (modulo 16). We are currently often overwriting
- a given picture number with different pictures perhaps twice a day. This
- is done when a picture is clearly black and there is no point in getting
- all of it.
-
- The first two bytes of data portion of the UI frame (packet)
- contain the 10-bit X position of the first sample in that packet, the
- two most significant bits of the X position occupy the two least
- significant bits of the first byte, the eight least significant bits of
- X are in the next byte. Legal X values are from 0 to 644. Y values are
- in third byte and range from 0 to 241. If the two least significant
- bits of the first byte are both set (forming an X 768..1023), the data
- beginning with the second data byte is clear text information about the
- picture (which we call the "picture header") occupying 252 bytes. The
- six most significant bits of the first byte are reserved for future
- protocol definitions, only one of which is presently defined.
-
- The data then consists of a sample for every third X value. This
- has little to do with RGB, but is relevant to color as the NTSC waveform
- is sampled at precisely 3 times the color burst frequency. A zero byte
- value indicates a simple compression sequence as follows: 00, the count
- of repeats, then the value repeated. Actual samples with a value of 0
- (hor. sync) are changed to 1 to allow indicating a following compression
- sequence. When the X value is greater than 644, increment the Y value
- by two and subtract 645 from the X value. At the end of one of these
- compression passes, encoding returns to the top of the pic, moves one
- byte to the right and proceeds. This takes place three times to
- complete one half pic (every other line). Alternating half pics (even
- vs. odd) lines are transmitted every 101 minutes (every other
- visibility pass at these latitudes). There is not vertical sync transmitted.
- The picture begins at the end of the header.
-
- The 645th byte of a reconstructed line is the count of zeroes which
- was in the original digitized sample between the 644th byte of that
- line and the beginning (the first non-zero) of the next line. If you
- are reconstructing the original waveform replace that count byte with
- that many zeroes.
-
- I must again emphasize that this protocol is very temporary and can
- be expected to change with little advance notice. The compression scheme
- is poor and was selected only for its simplicity. It also will change.
- These changes will be handled with Weberware version updates. Also,
- please note that we have the ability to download individual R, G or B
- signals from the camera and their protocol may be different. Ditto for
- the 1.2 GHz experiment. For those, however, you can expect a bit more
- advance notice of variation.
-
- As of today, only one picture has been arguably of the earth. We
- have been taking pictures (15) for only 8 (12 of our 15 on the first
- day) days now and it will require time to polish our procedures. We are
- concentrating on the equator now because pre-launch projections
- suggested potential problems at high latitudes. The first 12 pics were
- over the U.S. and none were of the planet. We have taken one pic at
- night (in the dark) to characterize performance. Two interesting
- tidbits. Shots that turn out to be of the sun don't move the impact
- sensor because, presumably, the iris is stopped down and has little
- distance to spring closed. Deep space photos increment the impact
- sensor considerably. The second tidbit has to do with the night photo. It
- has an average light level much lower than even the deep space photos taken
- during the day. Perhaps a little atmospheric glow around the spacecraft
- even at that altitude?
-
- Thank you all for your interest.
-
- 73's,
- Chris
-
-
-
-
- /EX
-