---Original Message----- From: William Chops Westfield To: PICLIST@MITVMA.MIT.EDU Date: Wednesday, 23 July 1997 09:59 Subject: Re: PIC RF link software.. > To implement compression for GENERAL data in a PIC sounds complicated, but > would be possible (but it would take long to write the software). If you > are going to transfer text for example, it would be better to write a very > simple routine that specifically transmits text more efficiently. Or if you > will transfer simple images you can use the method used in fax and > IFF-ILBM. The big question is if you really need to go through all this > work just to shrink the data size down to not much less than half? The only > efficient compression methods are MPEG for video and audio, and JPEG for > photo images, but they are of course very complicated to implement. > >MPEG and JPEG are both lossy compression schemes unsuitable for data (I'm >pretty sure.) The sort of compression implemented by WINZIP and related It would depend on data. If it's code, its certainly unsuitable, but datalogging ? >tools (including gzip, which has available source) (this includes the >compressions like V.42bis used in modems) can easilly achieve 2:1 >compression on typical text and binary files, they are highly array-based >and sometimes recursive , which would make them a pain on PICs. If you know your data's maximum rate of change, say you're logging level of water in a tank, then RLL or a delta system may be ok. >We're on our third generation of compression for object files here at cisco, >having done unix compress, lharc, and gzip styles... > I sure wouldn't want to do a dictionary based compression system on a PIC. An AVR - maybe. MikeS