Saturday, January 18, 2020

Some Reflections on The New Fad – YouTube SDR Comparisons

It looks like everyone wants to be on YouTube these days. Nowadays it’s being used to compare SDRs, and videos are published on SDR related Facebook groups. I’m doubtful if any of what I see really helps the reader/viewer to choose the right SDR. And most often they are using signal levels I would call “armchair listening”.

Worse yet, there is often little or no attempt to describe how they set up the “test”. Are all parameters identical or are they comparing apples and oranges? And very often there are no comments from the authors themselves. One would at least expect a conclusion: Is one better than the other? I suspect the reason is that usually, there is no audible difference.

I have made loads of A-B audio comparisons for my own use, and two things are certain: 
1. Results will vary. There are so many variables at play. SDR A can be better than SDR B on frequency X, but worse on frequency Y. 
2. Audio is subjective. It’s because ears are different and the decoding brains are different. If you ask 10 people which is better, A or B, they will split. And one tends to favour the audio flavour from the SDR software one uses the most. My opinion is that Jaguar for Perseus in most (but not all) cases has better readability than SDRUno, HDSDR and SDR Console. But a long-time SDRUno user may  not agree. And when DX signals are 100 % readable it really doesn’t matter which software you use, because what the DX-er is after is the ID. It may be different for a program listener (aka SWL).

If one really wants to try to measure the difference between SDRs, why not set up a test using digital modes, like DRM, WSPR or FT-8? Then you would at least have objective criteria to judge from.

I am sure they are having great fun when they set up these tests and do so with good intentions. Alas, they bring nothing new into the equation. Let the lab tests rule.

1 comment:

Anonymous said...

Yes have to agree with some of the You Tube videos, it's hard to get real results just based on some dodgy video.
Lab testing is probably the best,at least there is a reference point.
However good software is the most important, otherwise it's like having a car without a steering wheel.