I will reply to your comments, one-by-one.
73 Dave KB4Y
>After my first attempts in receiving and transmitting HD SSTV –
>pictures (including some excellent help and teaching of others who
>contribute to this reflector) I wonder if there could not be
>something else in between the all or nothing philosophy in this mode,
>especially if the channel is quite noisy.
Have you tried increasing the redundancy ? This is very useful during
of static crashes that can make getting a transmission through very
>Question to the experts: Did anyone try to cut the file to be
>transmitted into several blocks of a size which would be determined
>by the quality of the communication link and have the transmitter
>request a confirmation signal from the receiver (and if not received
>an OK to repeat the block again)? This would be very similar to
>Pactor (may be with varying block lengths). It would also require a
>different protocol compared to today but the encoding / decoding
>scheme and the type of modulation could be the same.
Barry Sanderson KB9VAK does this often. He says that many shorter files
will decode faster than one long one. There is also the added benefit
one short file is not decoded successfully, only that one short file
needs to be
sent again. The drawback is having to concatinate the files back to a single
file again. The Linux version of HDSSTV keeps the decoded results of each
block as a separate file. That makes for the posibility for a
have it's good blocks merged with the good blocks from a previous
>The requirements for the PC's would be quite high in order to avoid
>very long transmission times needed for the block wise decoding and
>the confirmations by the receiver but I would assume it would be a
>faster way to get a picture received through an unstable HF channel.
>Perhaps it could be worthwhile to think about the requirements those
>pc's or dsp-interfaces would need to fullfill. If we look at the
>spec's of the new PC's coming to the market we all see the huge
>improvement in computing power, but I don't know at the moment what
>would be required.
I would think that digital SSTV should be a broadcast type mode that
require acknowledgements from a receiver in order to get the communication
through. Error detection and correction is already in the program.
if necessary, is possible where it would be identical to the original