Dale Kelly wrote: > That is, assuming the combined trellis and R-S FEC > schemes are more cleverly exploited as you suggest. > The potential is clearly there but, do you believe > that this will actually happen in the US system? I > know that I'm a pessimist but it is for good > reasons. Dale, first, to be clear, the numbers I showed for the Gemini chip were the numbers the CRC reported. So there is nothing new in how the two FEC schemes are used there. Those C/N figures show that echo energy is being used as part of the usable signal. The fact that the baseline is still 15 dB of C/N shows that trellis and R-S are used in series. As to your question, if you don't have to change the transmission standard, then the improvement should be easy to introduce. Each manufacturer can choose to make the improvement on his own schedule. In this case, a known improvement today would be to use the trellis code to pinpoint where the error occurred, indicate that to the R-S code, then use R-S only to correct the error bytes. That alone will allow correction of twice as many bytes as what is possible now (20 bytes vs 10 correctable per 208/188-byte frame). Al mentioned this several times, and it's mentioned in one of the ATSC standards now. I think A/53. The effect that has is to edge down the C/N margin before errors become noticeable. Changing the basic FEC scheme to a turbo code would be much more difficult. A-VSB cleverly introduces turbo code, by layering it on top of the existing standard. And it gets very close to the Shannon limit that way. But I'm not dismissing the possibility that other clever things can be done with the existing transmission standard. Unleash enough math PhD candidates on the problem and see what comes out. I'm only advocating making the best with what we have, for the wide stream, for the time being. Bert ---------------------------------------------------------------------- You can UNSUBSCRIBE from the OpenDTV list in two ways: - Using the UNSUBSCRIBE command in your user configuration settings at FreeLists.org - By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word unsubscribe in the subject line.