"Official" Audyssey thread (FAQ in post #1) - Page 1775

I was thinking of a polarity switch, not a cable swap

Audyssey cannot change polarity or phase. It will sometimes detect a phase problem and put up a warning, but that's as far as it goes.

^^^
Audyssey has been known to fail to detect a sub phase error. Even with the pro kit (right, Keith?). And I gather did not do so in this case when the subs were intentionally reversed. That suggests some other method is needed to get the sub phase correct. I would suggest the old 50's trick of reversing the polarity of one of two side-by-side speakers and listening for the drop in bass. Ian's mains and subs are adjacent so that should be easy. Set the mains to large and the sub to max bandwidth and play something with lows in it and set the sub for the strongest lows. Switching it while playing for a good A-B comparison. Do that for both sides. Done.

I'm assuming that Audyssey would have detected a phase error in the mains so make them the reference and set the subs to match. I'm also assuming that the sub distance correctly indicates very little lag in the sub. I.e., the delay isn't enough to foil the above test.

How low does the receiver noise generator go? I seem to recall it working as a source for this test.

The DVE blu ray has a track for checking phase between two subs.

I have read that it's easier to find the point where the sub output is LEAST then simply switch polarity of one sub. Apparently its easier to detect the maximum cancellation than the maximum summation, but I've never tried it personally.

Yes the DVE track does this by frequency .

+1. Moreover, as is oft, oft discussed here, Audyssey does not set the crossover frequency. The AVR manufacturers' logic controls that choice, often reaching a conclusion that Audyssey's Chief Tech Officer frequently mentioned he disagreed with.

My main speakers are ported. Being subject to the same physics that all speakers are, they are 180 degrees out of phase at their tuning frequency, gradually going fully in phase above that point. AFAIK, since Audyssey doesn't know where the crossover will be, it could not properly correct phase at the splice, which is where it matters. Maybe someday they'll implement additional algorythms to adjust the sub delay to account for relative phase versus one or more of the other speakers, automatically adjusting based on the crossover point chosen for whichever speaker it seems best to focus on. Relative phases between sub and each speaker could differ to a relevant extent unless the difference in distance between sub and each of the other speakers is the same at the primary listening position.

It's not a good idea to have auto wakeup on when running the Audyssey calibration. See this part of the FAQ for sub adjustments needed prior to running Audyssey:

3. How do I set the controls on my subwoofer before running MultEQ? (click here)

It sounds as if you have done things properly but it may be worth reading through that checklist and coming back if there is anything in there you haven’t done or anything that sparks a new thought.

I can’t think of an answer to your problem as to why the sound is better if you flip the phase of both subs 180 degrees. Are you sure it is 'better' and not just that there is 'more' bass - the two are not always the same thing?
If you play a piece of music with a good, prominent and sustained bass guitar in it (I usually use Stanley Clarke) which way sounds best in terms of accuracy, musicality, rhythm, timing and so on - 0 degrees or 180 degrees?

If you are sure it sounds better set at 180 for both subs, then for now I'd leave it that way and enjoy. It may be worth rerunning Audyssey with the phase of both subs set to 180 degrees and then seeing if you get an improvement by flipping them back to 0. Without independent measuring gear I'm not sure you will get to the bottom of this one.

I'm curious what the Audyssey "after" graphs show?

This thread reminds me of my mother's soap operas; you can miss a few days and you really haven't missed anything.

The graphs do not show what they purport to show ... what most would assume they show. They show the weighted average (as determined by Audyssey's proprietary algorithms) of the system response at all of the spots that went into the calibration. While there might be somewhere in the measured area that has that response, that is not really what the graphs show.

The before and after graphs are tools, that's all.

Jeff

They are "useful idiots" IMO. ( to Gooddoc)

Just joking Jeff . I think the topic has been flogged to death...

Feri, it's pretty simple to do the 'battery test' on a sub. Just connect the + (Pos) of the battery to the pin of the RCA cable connected to the sub and the - (Neg) of the battery to the shield of the RCA cable. If the woofer moves outwards when connected this way, then the polarity is correct (ie a positive amplitude). If the woofer moves inwards then the polarity is reversed.

Those batteries with the terminals on top would be easy to connect in this way, using a little piece of twin wire to make the connection to the RCA cable.

But Keith, between the RCA jack and the driver there is an amplifier. You can't run DC through an amplifier. Right?

Disconnect the amp before testing if you're worried.

Come on Guys, why on Earth would you want to connect a battery to an analog RCA input of the sub's amplifier. That input only takes audio frequencies at 100 mV/ 10 kOhm, typically.

OK tip of the day! Why fuss with the battery trick, play DC off a test tone disc!

The battery offers a great way to measure phase. While 9V into a line level input is a bit high, the idea is to determine whether all the speakers are moving in the same direction when the input to the device/box/amp or whatever sees the proper polarity. An amp in the circuit that is always on should not be turned off.

Let's take a look at the circuit diagram of a typical powered sub.



Do you see "IN" on the left side? OK, that's the RCA jack. Now, do you see C1 (0.22 uF)? That's a capacitor. A capacitor always blocks DC voltage. Full stop.

The only way to do the battery test on the sub is to open up the box, access the terminals of the sub driver and connect them to the battery.

Agree?

Fairly off topic, but I'm sure most here will find this article interesting, and a lot to discuss..

Enjoy!

24/192 Music Downloads... and why they make no sense.

And lots of links to other interesting reads...

Someone suggested hooking a battery to an amp input to check speaker polarity????

The 5.5 Volt is AC, the 3 Volt of the battery is DC. That's a big difference.

Fabulous!

@pepar: Jeff, scroll down to the header near the bottom of the page "Confirmation Bias, the Placebo Effect and Double Blind". There is some terrific stuff in there that is 100% relevant to the 'debunking' thread you started and which I am subbed to.

More like "marketing" tools. :-)

I just read the same thing. You and I and most others don't need any more evidence, and the ones who do won't be swayed.

I liked your DC test signal ...

Ah, right. Asked and answered then - thanks!

You are correct that the amp will not sustain output with a DC input. However, there will be a transient "thump" resulting from the step change in input voltage and you will see the cone move significantly. You'll probably have not much over a second before the cone has returned to rest, but it should be enough to see which direction the cone moved.

A step change in the input voltage contains all frequencies from DC on up to infinity but is dominated by DC and the lowest frequencies. Hence you will see the cone move in the direction of the rise in voltage.

That said, I would not do this and especially not with more than 1.5v given that the amp input is probably rated for not much over one volt.

So you can use the 'battery test' on a subwoofer then? Darn! I just conceded to Feri! That's a beer you owe me, Feri

True enough. There was a link in the article on how to conduct a double blind test - I posted the link in the other thread. It's a useful read.

Hi Harrison, you are, of course, right about the transient "thump", but I would still not recommend to do a sub driver test this way. By coincidence it can/could do more harm than good (especially for non-engineers) if it is repeatedly done coz the guy couldn't catch the movement of the cone the first time.