For sure, there is not a lot of activity on that group, and I must have joined it for any useful hints on optimising JTDX for WSPR use, of which I found none.
 |
| Add caption |
But as is inevitably the case, anything that looks like criticism brings out the worms from the wood.
Immediately, a US-UK pair of members asserted, without any evidence to back their claim, that "we all know the truth" about JTDX's [superior] performance. My test was, the UK operator claimed, "bogus" - again without saying what aspect of the test I had knowingly misrepresented.
The admin for the JTDX FB page - presenting himself as DL5NAM - wanted to know what the arrangement for the test was, and how I had split the signal for testing.
My simple approach was merely to record WSPR time slots (21 of them), and decode each recording using both software in turn. I even made sure to record half the audio with JTDX, and half with WSJT-X (there was no difference in spot number or SNR, no matter how the audio was recorded).
So there was no need for complex signal splitting or multiple software incidences.
If anyone can reasonably argue that audio recordings compared through the two softwares are not a valid means of presenting signals for decoding, then I would be pleased to receive the observations.
If, instead, people just want to assert, shout and insult their way to 'correctness', then I have no interest, thanks.
A check was made of 'live' decodes against what the same software then decoded when the audio recording was run through it. There was no difference in the number of decodes, or their reported SNR for either of the softwares.
So, for the moment, until I hear persuasive arguments to the contrary, my method, whilst not the only way of comparing WSPR performance between softwares, is probably the easiest by far - and entirely valid and reliable in its findings.
No comments:
Post a Comment