I have heard many times that MSHV decodes better than WSJT-X. Some people say that JTDX is even more better. But what is the truth? I decided to find out.
To do that I took about 2 minute samples on different bands and different times. I entered the results into Excel so I can analyze them and make some graphs. All three softwares were running simultaneously and using same audio from same radio and interface so only difference comes from software itself. This of course takes much processing power but my computer is powerful enough to do that.
I wanted to take samples on different bands and times so I have different kind of situations. If I would have taken only one band then results would have presented only that situation. Every band had little bit different situation on those moments. I was lucky to get also some aurora signals to decode later in the evening. I was going to do that test later but now I have it already included end of this article.
One thing is important to say. This experiment is only counting how many decodes there were. It will not take account their quality. So even one software might get more decodes it does not necessarily mean that they were all good decodes. In practice only valid decodes matter.
Station setup used in this experiment was Yaesu FT-1000MP with Microham Micro Keyer II interface. Antennas were following:
160m Inverted L
80m Vertical
40m Vertical
30m Inverted V
20m/17m 3+3-el dualband yagi
15m/10m 3+3el dualband yagi
12m dipole.
Softwares used were:
WSJT-X 2.7.0
MSHV 2.76.1
JTDX Improved 2.2.159
The reason for using JTDX Improved version was that there was no any public updates lately for original one so Improved version was most up-to-date that I could find for this test. When I am speaking about JTDX in this article it means that improved version.
At first I took some time to find settings for every program that were close enough and optimal to make comparison. In this article I have included two sets of results with one difference. In the first set WSJT-X is NOT using AP. In the second set AP is enabled. That is because there is difference between WSJT-X and competitors that is needed to understand. To get more decodes it usually means that there is also partial decodes and possible false decodes. In WSJT-X when AP is not enabled you won’t see that much those decodes. So decoding quality is actually better. But you can of course include those decodes if you want to by enabling AP from decode settings. We will take a look what is the effect of this.
So lets take a look to first set of results. These samples were taken around 09:45 – 10:15 UTC. Yagi antennas were pointing to east. WSJT-X has AP disabled.
We can see that there is some variation but not same software is best all the time. Anyway MSHV and JTDX got more decodes. But remember that WSJT-X was AP disabled.
Now lets take a look what happens when AP is enabled. At this point it was also grayline time so first I took sample on 80m to see is there any difference on grayline decoding.
Ok not much to see this time on grayline but at least WSJT-X is decoding everything. I also made interesting find that there were AP2HA calling CQ with weak signal and WSJT-X decoded that when MSHV/JTDX did not. Another finding was that when there was two stations top of each other JTDX actually decoded that weak station under that stronger station. MSHV did not.
But lets now take samples from all bands with AP enabled. This time yagis are pointing to west.
So now we can see that all three softwares are quite same. Small variations in different situations. What if we sum all those together which one is the winner?
The winner is WSJT-X with 1743 decodes. Second was MSHV with 1741 decodes and last JTDX Improved with 1727. But they all did a good job! Not that much difference between them. Things might have been different in past but with these software versions they are actually all quite same.
Finally lets take a look how they are performing when there is aurora on bands. I turned my yagis to USA and took samples from several bands. In practice every band had different affect by aurora. There are taken after 20 UTC when aurora started to build up.
Not much difference this way BUT on 10m there was strong effect and JTDX decoded clearly more than others. Also after these samples I listened some time VK9CU on 80m and JTDX decoded best. So it seems that when there is aurora or just polar flutter then JTDX might be good option.
These were only my results and you might get different results. I am not saying any absolute truth. These are only interesting notes. But what I can say in my opinion is that you can make QSOs all of these softwares and all of them can decode very well. If you see big differences then you might have bad settings or good imagination 😉 There are small differences and sometimes one is better and another time the another one might be better. If you are chasing rare DX and signal is poor you might actually want use more than one decoder simultaniously. Of course you need a computer that is powerful enough to run all of them same time.
Even I could not find any big differences in decoding there is of course some other things that might have some value to you. For example in JTDX Improved you can adjust time inside the software without touching to system clock. That is sometimes handy feature. But what software is for you is mostly a matter of taste. Just keep DXing with your favourite software!
I might do some tests later with different settings and propagation. I might also try how these softwares perform with FT4 or maybe even with meteorscatter QSOs. If you have any thoughts and what should I try please let me know. I am interested to test.
73!