Talk:Morse code

Bandwidth

@Lklundin: Part of my edit summary was cut off because it was too long. The missing part said "The limit on speed is set by the abilities of the operator, not by the bandwidth of the channel". In comparison to the transmission speeds in use at the time, the bandwidth was essentially unlimited. In fact, it was not even accepted that faster speed required more bandwidth until around the 1920s. Before that it was believed that radio telegraphy occupied a single frequency only. SpinningSpark 11:55, 15 February 2015 (UTC)

@Spinningspark: Thank you for opening this discussion. I think we have a different understanding of bandwidth. What I mean by that is the common meaning from information theory (which is explained at Bandwidth (computing), namely the speed at which a certain amount of information is transferred. So this is _not_ about the speed at which the Morse operator can produce dots and dashes. Instead, given a certain, fixed speed of the Morse operator, the speed of communication is the rate of which the actual text is transferred. In the design of Morse this speed was increased by encoding the alphabet such that as much text (information) as possible is communicated by a given amount of dots and dashes. Although the whole concept of information theory had not been formalized at the time then the Morse alphabet was designed, there was still an understanding at that time, that an encoding of the more frequent characters with shorters codes would increase the speed of the communication. Apart from the hint at this with the link to the Huffman coding, the article is quite weak regarding this aspect, and it would be good to extend it with typical data rates (i.e. actual amount of information transferred, not number of dots and dashes) for typical operator speeds. Btw, a similar and related case is that of cryptography, where also in the 19th century cryptography techniques were used without a formal understanding of how well they were working, like the one-time pad. Lklundin (talk) 12:35, 15 February 2015 (UTC)

I perfectly understand what is meant by bandwidth here, please don't talk down to me, I have more than forty years experience in this field. Data bandwidth is not the same as frequency bandwidth but the two are most certainly related. The fact remains that nobody talks about Morse code in those terms, especially the early history of it, so it is wrong to put that in the article. You can transmit the data as fast as you like, but it doesn't make any difference if you are limited by the speed of a single operator. The difference can be seen at the point at which multiplexing came in. Multiplexing unarguably increases the transmitted data bandwidth. However, it was still not recognised that there was a limit to the channel bandwidth, they naively thought that they could carry on adding more Morse channels ad infinitum. It was only the advent of telephony that changed that outlook.

Going back to a single operator, the transmission is limited by the symbol rate the operator can produce. The symbols in this case are dots and dashes, but we could equally call them 0s and 1s. Changing the coding does not change the symbol rate produced by that operator. It follows that the transmitted data bandwidth has not changed either. Optimizing the code to efficiently encode the English language is not the same as increasing bandwidth. In my opinion "efficient" was exactly the right word here. SpinningSpark 14:29, 15 February 2015 (UTC)

Letters, numbers, punctuation, prosigns and non-English variants

The digraph AR is currently only represented as the plus sign [+]. Its more common usage as an "end of message" prosign is absent. Likewise, the digraph KN is currently only represented as an open parenthesis [(]. Its common usage as an "invitation to transmit, called station only" prosign is absent. Notes to these effects could be added to the pre-existing entries in the sub-category Punctuation. 50.174.97.118 (talk) 02:14, 25 February 2015 (UTC)B. Salvisberg, WA6A

.

Challenge for the quinary view

I want a citation for that; it's not WP:CALC because sources (e.g. Shannon in the celebrated A_Mathematical_Theory_of_Communication!) say that it's at most quaternary (four symbols) because of how symbols may occur; in particular a single unit-space cannot occur arbitrarily so it may subsumed to the dit or dah to its left. 86.127.138.234 (talk) 22:12, 1 March 2015 (UTC)

And if Shannon is wp:primary source for you but your uneducated and uncited WP:OR is somehow acceptable instead, refer to Richard Blahut's Information Theory and Coding as secondary source expanding on Shannon's analysis at a textbook level. 86.127.138.234 (talk) 22:18, 1 March 2015 (UTC)