Saturday, July 4, 2015

A, B, then A again

In a typical sighted listening comparison with more than one person, there are 2 typical outcomes:

The most common is to listen to A, then B, then declare that B is so much incredibly better than A that there is no need to go back and listen to A again.  Your hearing abilities may be criticized if you fail to hear the huge improvement.

In the other typical outcome, for some reason, a decision is made to go back and listen to A after listening to B (perhaps because the audience has an "A" partisan).  Sure enough, the change from A to B was overstated, now A is showing that it does the same things that B did.  In fact it may be a better compromise overall than B.

I believe most such "testing" is bogus, and the sighted aspect is usually only one of many factors that makes typical audio comparisons by audiophiles unreliable.  The lack of level matching is another huge factor.  Going back to A has usually been, for me, and decent control anyway.  Most often the differences I heard with "B" were mainly because I had just heard "A" and now my senses are primed to listen for something different.  But then whatever I thought I heard anew with B just because of listening experience tend to still be there when I go back to A.  So whatever the other limitations, and A/B/A test is far better than a A/B test, and it tends to moderate the sense of huge differences with people who have at least some degree of ability to change their minds when exposed to evidence.

I was reminded of this when I found how to fix the frequent dropouts in my 2nd FM tuner during this vacation week.  I had been noticing dropouts in my second FM tuner, a Kenwood KT-6040, going back some time.  I'm not sure how long this problem had been present.  I keep the Kenwood tuned to one of two college radio stations mostly.  Tuner #1, the Pioneer F-26, is usually tuned to the classical station I listed to more often.  But this being a vacation and holiday week for me, I'd been noticing the dropouts more and more and it was beginning to drive me crazy.

So first I tuned in the same college radio station on my Pioneer F-26.  It had no dropouts.  Well that was not a good comparison because they use different antennas also.  So I connected the F-26 to the antenna I normally connect the KT-6040 to.  Still no dropouts, and at this point I had to reflect on how good the F-26 was sounding on this station, better than I had remembered, not only more musical than the KT-6040 but about equally quiet.  Because of sophisticated multiplex IC's which became available in the late 70's and appeared in tuners mostly in the 1980's, generally newer tuners are considerably quieter than old ones, but they often don't sound as good.  Older tuners tend to have ganged air capacitors which make far better machinery for tuning than the varactor diodes used in all "digitally tuned" FM tuners.

Anyway, at this point, with the F-26 not having the dropout problem (a dropout about every 10 minutes or so--after an hour of listening there's no need to A/B or A/B/A or ABX to be sure, the difference between a channel with frequent dropouts and one that doesn't is absolutely apparent if you listen to the channel long enough for there to be no doubt) I was beginning to think it was that superiority of tuning capacitor in the F-26.  I was thinking which analog tuned tuner might become my next #2.

But just to be sure I wouldn't be relegated to only using analog tuners, I decided to try a second digital tuner, a Yamaha TX-1000.  Interesting enough, it did not have the dropout problem either.  I didn't like the sound as well as that of the F-26, nor even the KT-6040 when it is not dropping out, but it was good enough to be the new #2, though I do like a lot of things about the KT-6040 better, including especially the ability to scan for new stations using the remote control.  Nevertheless, I continued listening to the Yamaha on both college stations all night, just to be sure it didn't show the dropout problem, and it didn't.

Well now I thought there were two possibilities that the KT-6040 was actually OK after all.  One was that when I was attaching Antenna #2 to the Yamaha, I had found the twinlead caught below a power cord.  I had to fix that just to pull out the twin lead far enough.  So perhaps the fix was not actually changing the tuner but just re-routing the antenna lead.

The second was that I changed the Sonos input level.  I listen to each of the two tuners through Sonos, by connecting the tuner outputs to Sonos Connect inputs.  I find Sonos with uncompressed I/O to be very transparent, so I don't bother bypassing it as I used to do (both tuners are actually in the living room, I could simply hook them up to the Lavry AD10, and in fact I used to do that).  One of the cool things about using Sonos this way is that I can set the level exactly for various kinds of devices.  But I can also set it wrongly, and unfortunately Sonos doesn't provide any feedback like a "clip" light, even in the computer interfaces.

I noticed when I was connecting the Yamaha that the input level for Tuner #2 was 6, which assumes a very low voltage level.  Meanwhile, the Pioneer F-26, which has < 1V output and 0.58V for full modulation, was set only to level 2.  Well I couldn't believe that the Yamaha would have far less output than the Pioneer, in fact I would think the Yamaha being a late 1980's design, and the F-26 a late 1970's design, that the Yamaha would have more output, inspired by the 2.0V output of CD players.  Prior to the introduction of CD players, most line level sources were less than 1.0V, sometimes a lot less.  CD was launched with higher output as to trumpet the high measured dynamic range.  So I changed the level to level #1.

So anyway, both or either of these changes could have been responsible for the drop-outs on the Kenwood, as they were changed when I changed to the Pioneer and Yamaha tuners.  I switched back to the Kenwood, keeping the level at #1 and making sure the route the twin lead far from other cords.  And now, there were no drop-outs with the Kenwood.

I was really glad I had checked, because I do like the sound of and remote control features of the Kenwood better.  But it would still be interesting to distinguish between the antenna change and the Sonos level difference.  But I was thinking I don't need to do this test, I have the Kenwood working now, and I'm sure it should have Sonos set to the highest level, and I'll always be careful about twin lead routing.  Plus I didn't think the antenna cable routing would be responsible for this big a difference, I was sure it had to have been causes by the Sonos level setting.

But it was easy to change the level back to 6 so I did, and now thee were still no dropouts.  So it does appear that it was antenna routing that had caused this problem, and it probably started when I added the new DAC and didn't route the power cord and antenna to avoid each other.

So this shows how it is, to really find out you have to go back and try many additional tests, rather than making assumptions, or just settling with good sound without worrying about how it came about.

And that's probably how it ought to be, but I won't promise to always do more tests because I don't always need to know…I can maintain a "not sure" about any belief so I won't be trapped by it in the future.  I'll find out, if I need to, by doing the extra tests when I need to do them.

I won't commit to either approach, only to doubting a result if it hasn't been fully verified.

No comments:

Post a Comment