I would say that the reason is that without injecting known information to the decoding process the computer program appears not to be much better at copying weak signals than a good CW operator.
JT65 SOFTWARE SOFTWARE
It took me a while after JT65 was released until I realised what was possibly going on. As I stated above, in my opinion t his piece of software lack some properties that a secure communication protocol should have. To understand what I am talking about on this page one has to be familiar with the WSJT software package developed by K1JT. Therefore digital EME QSO's are no longer treated as the big achievements moonbounce contacts once were, and some people judge these contacts as being of less value, instead they treat them as Internet or Cluster QSO's. This has lead to the fact that random operation, with many unknown parameters, is not a common way of operating any more. As the program is not "random friendly" i e, people have a hard time finding each other on the bands, the use of real time liaison or continous spotting of ones presence is heavily relied on. K1JT has in this way decided to set his own standards, and thereby in my eyes unfortunately invented a program that heavily relies on known information. In my opinion this is not a good protocol, and it lacks some of the basic functions to enable an amateur radio QSO where all information is transferred via the airwaves at all times. In effect, the program creates (K1JT uses the terminology "reconstructs") data that it has not received over the airwaves by making use if the available known data on the computer. The decoding is heavily relying on known data that is injected via a database or input boxes in the program user interface. In my opinion WSJT/JT65 is not in all ways complying with these standards, despite what the operator can see printed on the screen. Well, to make a long story short, when a new communications protocol is to be launched there are certainly some requirements that need to be fulfilled to make it compatible with long established amateur radio standards. In my eyes this invalidates JT65 contacts at S/N levels around –25 db or lower when comparing them to long established standards in the radio amateur world.įirst, why am I, SM2CEW Peter Sundberg doing this? This again tells us that this is not a protocol designed to dig in the noise for weak radio signals, instead it looks for fragments that can be matched to known information present on the computer only.
For those of you who have visited my webpage before, the added new information is presented with a " New! " marker. Furthermore there is more information on Short Hand messages presented in the text below. I also added some new statements from K1JT explaining how the Deep Search mode uses known information to print EME messages on the screen.
JT65 SOFTWARE FULL
"Deep Search can not communicate callsigns" by Klaus von der Heide, DJ5HGĪdded information about the recent Dubus article by DJ5HG showing how JT65 is using very little information (14 bits) to produce full calls, locator and report when the Deep Search (DS) mode is used. "A Comment on Joe’s Paper How many bits are copied in a JT65 transmission" by Klaus von der Heide, DJ5HG Please read the article, it reveals the true state of JT65, and clears up the smoke screens that K1JT has provided us with. JT65 "Deep Search" is probably the biggest con (bluff) in the history of ham radio. We all know what we call doping in sport arenas. Take away the known info from the computer (calls + locator) and a decode is totally impossible. Actually, doping is done to the extent that the message received via the airwaves is a fraction of the printed message on the computer screen. JT65 Deep Search "copies" two calls, locator and report in 48 seconds…!! This is only possible by "doping" the decoding process with the missing bits, which is exactly what K1JT is doing.
Klaus makes it clear that at –27db only one call is possible to transfer in 60 seconds. Well, this is because mathematics can be used to determine the possible data transfer at a given S/N ratio. How can we say this after reading the articles? It will show that the so called Deep Search mode is not a deep search, it is solely a reconstruction module, designed to recreate a fully known message. Please read this interesting articles by Klaus DJ5HG, presented in Dubus Magazine. Analyzis of Deep Search, an article written for the Swedish ham magazine QTC in November 2007