Home |
Search |
Today's Posts |
#1
|
|||
|
|||
Computer-Read Morse (Was Morse Code Binary?)
In article et, robert casey
writes: Besides, if something is digital, why would you have to try so hard to make it computer compatible? Telling "S" from "O" is hard if you don't already know from looking at other characters what the speed must be. Something easy for the brain but hard for computers to do when the sender varies his speed. It is not "hard" for a semi-modern computer to do. All it means is that the programmer has to write more source code to perform a task called "rate adaption." Lay folk take the very easy approach to visualizing what the computer does (via its programming) as simple decoding of groups of binary state combinations. That's common observation. What isn't so obvious is that, with a clock rate of 20 MHz or better, the computer [program] can test a time sequence for relative occurances of On versus Off by periodic sampling. Segregation of a single O from a single S in morse code can be done by comparing the time length of the On. That is different in morse code for an approximately equal character bit group time. [dashes are supposed to be longer in duration than dots] At a dot rate of 1 KHz (faster than 100 WPM equivalent), the repetitive period is 1000 microseconds. The computer hardware (under program control) can sample that at a, say, 10 microsecond period. That allows 100 samples of each dot or dash which can be stored in temporary computer memory. [that's more memory than is actually needed but this is just an example] Already knowing the sample rate (from the hardware and the time of execution of the iterative routine), another routine can test each sample in turn and accumulate the relative length of each On versus a character group's total time length. That would differentiate dashes from dots. The difficulty for the programmer is setting up a set of rules to determine the relative lengths. Such does NOT have to be precisely 3:1 in ratio of dashes to dots. It can be as coarse as 2:1, which would be called "sloppy" manual sending. :-) Once there has been a differentiation of dashes versus dots, a variation of the rate-adaptive routine can be used to differentiate the longer time between character spaces. Again, the routine's testing can use a wide range of time intervals. When spaces between characters are determined, the much-simpler task of decoding character combinations into ASCII (or equivalent) character combinations can be done (probably by a look-up table stored in memory). With that completed, the computer- handleable characters can be displayed on a screen or sent to a paper print device. I've not done the above but am describing what a programmer acquaintence had already done some years ago using a desktop computer having a 20 MHz clock rate CPU and very moderate RAM for temporary program memory. My (then-new) Icom R-70 was lent to him for some on-air real-time testing on amateur as well as commercial morse code signals (there were very few non-amateur morse signals heard then). With a simple audio- to-digital interface (peak detector with quick fall time), it would "read" every morse signal with a 2:1 dash-dot ratio or better, even in moderate noise. The first few characters of a string of characters might be lost until the rate-adaptive routine adjusted itself, but otherwise was quite acceptible. You might ask "why" do that at all? Answer: It was someone else's personal intellectual exercise to accomplish a seemingly difficult task. Just a fun task for him and of some interest to others who liked to program. It was a hobby project and it proved what was set out to prove to the programmer. That morse decoder wasn't carried to any perfection. It could have been developed further to be quicker in rate-adaption, to add more noise-elimination and so forth. There was NO DEMAND for that commercially, nor was it done for prospective program sales in the future. [hardware already existed to do all the rate-adpation and decoding of repetitive teleprinter codings in a single IC a decade or so ago] Program notes and source code printouts filled a large loose-leaf notebook and were freely shown. Of particular note was the flow-diagrams of the various iterative subroutines, much more readable to me than the C source code (maybe it was Pascal?) of the executive program. The same executable program would work fine with modern desktops or portables having 2 GHz clock rates and 100 MHz rate memory fetch cycles...even to using already- established routines to determine real-time hardware execution times necessary to determine word-per-minute equivalent rates. [portions of this example description were already in this newsgroup a few years ago, principally in reply to Ed Hare and some others] It is less a matter of "adapting morse code to computers" than it is making the computer adapt to decoding morse. A big question is WHY BOTHER? Morsemen, hoping to show they are "superior" to mere machines as well as modern technology, will insist and insist that their human abilities (honed to a fine edge) are oh-so-much-better than any (hack, ptui) computer thingy. They can evidently work miracles using morse code mode communications, far beyond the best that can be done by all other radio services and commercial communications. :-) "Morse code gets through when all other modes do." - Brian Burke "Today is the tomorrow you worried about yesterday..." - anon. |
#2
|
|||
|
|||
Len Anderson wrote: Morsemen, hoping to show they are "superior" to mere machines as well as modern technology, will insist and insist that their human abilities (honed to a fine edge) are oh-so-much-better than any (hack, ptui) computer thingy. So far, it's true. Do you have a problem with the truth, Lennie? Oh...wait...That was a rhetorical question...never mind. They can evidently work miracles using morse code mode communications, far beyond the best that can be done by all other radio services and commercial communications. So far, Lennie, you're the only one saying that. "Today is the tomorrow you worried about yesterday..." - anon. "Today is the tomorrow that Lennie's surprised to have made it to considering how many people he's lied to and ticked off in his life" ~~ Steve Robeson Steve, K4YZ |
Reply |
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
FCC Assigns RM Numbers To Three New Restructuring Petitions | Policy | |||
My response to Jim Wiley, KL7CC | Policy | |||
ATTN: Tech Licensee USA Morse Code Freedom Day is August 1st | Policy | |||
ATTN: Tech Licensee USA Morse Code Freedom Day is August 1st | CB |