Axiom Home Page Forums General Discussion Home Theater Double Blind Listening for AVRs- How is it rigged?

OK with my limited knowledge of electronics, I could envision building a simple switch that quietly transfers signal from one set of speakers to another for comparative listening.

However, with all the chatter about which amps and processors possibly sounding better or worse than another, it got me thinking about why people never post that they have tried going back and forth from one set of gear to another. I suspect, the quick answer is somewhere between, I'm having too much fun with my new gear and it's difficult to set up a switching process that doesn't involve cable swaps.

Soooo.
I can see Toslink being a pain. You need specialized equipment to switch fiber optics. However, with Coax, we are back to electrons and a more manual switch would work, but how well? Without even going into the 'rules' of a double blind listening test (of which I don't know much about either) I can still envision some challenges.

With speakers, or even separate amps, a mixer type device could ensure a seamless transition from speaker group A to speaker group B. However, with a digital, multi-channel signal this would not be the answer.

Has anyone built a switch to simply take input from say a DVD player and seamlessly transition it between two processors/receivers?

If so, can you share the basic schematics, theory, complications or reasons why this isn't commonly done at a DIY level?

Work is spookily slow today so I'm thinking too much.

Better to switch it at the outputs than the inputs. But then you wouldn't be able to compare processing, I suppose.

Im sorry that I couldnt tell you how its rigged. But like all DBT's is definetly rigged.

Andrew, although a properly set up double blind test is the only credible way we have to evaluate supposed audible differences between electronic components, it's not that simple to do as merely using a switch. One of the basic requirements is that volume level has to be held equal within very close limits(0.1 dB is commonly a requirement). Failure to be listening at equal levels is the most common reason that we often read of enthusiasm for new players, amplifiers, etc., that in reality don't exhibit audible differences. Setting volume controls at the same point of course doesn't do this since different units have different outputs or gain levels with the same settings. SPL meters aren't nearly accurate enough. The most common procedure is to measure voltage with a voltmeter.


A recent article in The Audio Critic, "Electronic Signal Paths Do Not Have a Personality!" very briefly outlines what basic principles of audio technology teach us in this area, and is worth a read.

Thanks for the info folks. Starting to shape up as I expected but more detail on the standards and procedures for a professional setup would be appreciated. Just curious in any case.

Also, I should have been more clear. I was not trying to join a particular camp in the 'amp A. can sound better than Amp B.' discussions that sometimes take place here. I was just wondering if it was difficult for the average purchaser of a new amp, AVR, processor, or whatever to do a reasonable home evaluation?

I began getting curious because many people on this and other boards discuss how their new equipment has added this and that to the sound they are experiencing but nobody seems to have taken the time to do an A/B comparison with their old. It seems this would go a long way to add to the discussion verses the 'has no effect camp'.

Since I assume everyone is like me and never throws out old electronics until you can't remember exactly what it did anymore, then I had to figure that it would be very difficult to set up a fair comparison if nobody is doing it. This appears to be so.

I'm still interested in how it's done professionally though. What are the requirements to say you did a fair assessment.
For instance, I envision a layman's stereo speaker switch (not a fair assessment) by using a mixing board. This would allow you to set each set of speakers to equal loudness levels and also make the switch. However, I would not have known they have to be set more accurately than a db meter allows. Makes sense, but again....rookie.

What else is required. Does the switch from A to B have to instant so as to be disguised as non-detectable as possible? This might make some sense as well because if you can't hear the difference enough to tell it has switched, then you obviously are not able to identify enough audible differences to say one is different than the other.

Sorry I write long posts. I'm sure I can answer all my own questions if I do some web searching but boards generally provide more interesting feedback as writers are not trapped into technical writing mode.

It's easier to explain if we set up an example. Let's say you want to fairly compare two receivers, A and B. For simplicity, we'll only deal with the front Left and Right channels of the amp. Since we want to keep as many variables constant, we'll use the same source -- a CD player -- and the same speakers during the tests.

Using a "break-before-make" speaker switch, connect the speakers to the common terminals and connect each amp to the A and B terminals, respectively. Connect a pair of Y splitters to the stereo RCA outputs of the CD player, and then connect a stereo pair to the CD input on each receiver. Now, using an SPL meter and a test disc in the CD player, calibrate each receiver so that it outputs to the speakers at the same dB level.

You've now successfully set up an A/B comparison. You could go ahead and test now, but to get truly good results, you'll want to close your eyes or hide the equipment and have someone else switch between the receivers so you can't tell which is playing while you listen.

Peter,

With this type of speaker selector, the switch-over will be audible, will it not?

Sure. The way to get around this is to throw an extra step into the listening session. The switcher should occasionally flip the switch or push the button only halfway and then switch back to the same device. This way, the listener cannot know that a brief interruption in the signal is really a switch to the other device at all.

pmbuko,

What you are describing is the basis of a really excellent double-blind test, the A/B/X comparator, in which the switching device may select "X", which might be "A", "B" or "X" ("X" being A again or B) so the listener simply has to judge whether there is an audible difference or not.

At one time, you could purchase an A/B/X comparator, originally developed by David Clarke, but I don't think it has been available for years.

Regards,