One Summit "out of phase" with Audyssey measurements

[sleepysurf]

Since getting my Denon 3808CI for home theater use, I've been playing around with Audyssey MultEQ XT. Every time I've run the measurements, Audyssey gives a warning that the left Summit is "out of phase" which I've just ignored, per Audyssey's recommendations (since they say dipole speakers may yield a false positive phase warning). I finally had time today to swap the right and left Summits, and am now consistently getting the phase warning for that same speaker (now on the right). Apparently there *IS* something wrong with that one Summit, and it's not due to the dipole/room interaction. In the past I had to pull out the bottom woofer of that speaker to fix a loose connection, so I checked today to make sure the red and black wires were properly reconnected, and they are.

Now what? Should I risk disassembling the entire speaker to check all the other connections, or just keep ignoring the error (not sure if it makes an audible difference or not). Of note, my Summits were bought used, so it's possible the previous owner may have disassembled them for some reason, and switched the wires for the front woofer and/or stat panel.

 

[Audiophiliac]

If the panels were out of phase, it would likely be blatantly obvious. If you reverse the wires on one woofer, does it say its still "out of phase"? I would try that first. Then maybe dig into it a bit deeper.

 

[Brad225]

Alan I would find it hard to believe that one of your panels is out of phase with internal wiring.
When I listened to them they had a strong clear centered image and this would not be true if one were out of phase.

Try wiring one speaker out of phase and listen you will hear more sound at the sides of your listening position and no center image.

 

[mikesinger]

What do your ears tell you ?

 

[Statman]

As posted just reverse the positive and negative terminals on the "out of phase" speaker and try the calibration with Audyssey again and see what it reads?

 

[sleepysurf]

To my ears (and using a Chesky Test CD), the speakers are, overall, in phase. However, I'm wondering if just ONE of the woofers might be wired wrong. Interestingly, I took off the jumpers to test the Stat panels alone, and Audyssey then said they were BOTH out of phase. When I repeated the measurement with both panels and woofers, only the suspect one gave the out-of-phase warning. I didn't have time tonight, but ASAP will repeat measurements with the woofers only, as well as purposely reversing the speaker cable polarity on the suspect speaker.

I know Seth just got the Denon AVP-A1HDCI, and hope he'll comment on how the Audyssey measurements went with his ReQuests.

Anybody else running Audyssey with their ML's?

 

[Craig]

If your woofers were out of phase it would result in reduced bass. I would suspect that the dipole effect from the large panels confuses the phase reading the Audessey measurements.

Keep in mind there is a major sound wave coming from the front of the panel and a delayed sound wave from the back of the panel reflecting off the front wall. The typical cone box speaker doesn't quite have this effect. The distance of the speaker from the front wall most likely has an effect on the phase. And you may not need to move the speaker that much.

I'll bet if you relocate the speakers closer or further from the front wall would effectively change the Audesseys phase readings. Try moving one speaker at a time and then retest for phase. It may be that the Audessey could be used to help with speaker placement. Just a thought.

 

[randallp]

I own a 4308ci and ran into the same problem where my left channel reported out of phase. I switched +/- and found a tremendous improvement in overall balance. Before correcting, my left channel seemed drawn back and I would correct by turning up the left channel. After, no problems at all.

 

[randallp]

I own a 4308ci and ran into the same problem where my left channel reported out of phase. I switched +/- and found a tremendous improvement in overall balance. Before correcting, my left channel seemed drawn back and I would correct by turning up the left channel. After, no problems at all.

One thing to note here...all my audio speaker connections were connected correctly

 

[sleepysurf]

Craig nailed it! I slowly moved the "suspect" speaker BACK towards the rear wall, and at ~3' the phase errror went away! Previously, I had pulled them as far out as possible, usually about 4-5' from the wall. I then re-measured Audyssey for 8 different listening positions, and am quite happy with the result. If anybody else encounters this issue with Audyssey, play with the speaker placement first before all else!

 

[JonFo]

Craig nailed it! I slowly moved the "suspect" speaker BACK towards the rear wall, and at ~3' the phase errror went away! Previously, I had pulled them as far out as possible, usually about 4-5' from the wall. I then re-measured Audyssey for 8 different listening positions, and am quite happy with the result. If anybody else encounters this issue with Audyssey, play with the speaker placement first before all else!

Yep what you are affecting is the rear wave intensity and/or direction. Also, you are affecting its relative delay. As I've shown in the metrics in the measurements and placement thread, that rear wave can be a significant percentage of volume relative to the front, so easy to see why it would fool the Audessey algorithm.
Also, I believe Audeyssey might be using time windowing to detect phase. If the rear wave is more than 12ms delayed, it might read a phase error, but if less than 10ms it won’t (assuming that those reflections are artifacts from other nearby objects).

Another way to deal with this is to damp more of the rear wave with absorption, because if it's strong enough to fool the measurement system, it's definitely strong enough to be audible in a negative way.

 

[randallp]

Yep what you are affecting is the rear wave intensity and/or direction. Also, you are affecting its relative delay. As I've shown in the metrics in the measurements and placement thread, that rear wave can be a significant percentage of volume relative to the front, so easy to see why it would fool the Audessey algorithm.
Also, I believe Audeyssey might be using time windowing to detect phase. If the rear wave is more than 12ms delayed, it might read a phase error, but if less than 10ms it won’t (assuming that those reflections are artifacts from other nearby objects).

Another way to deal with this is to damp more of the rear wave with absorption, because if it's strong enough to fool the measurement system, it's definitely strong enough to be audible in a negative way.

This thread peaked my interest as I experienced the same problem and corrected it by swapping the +/- on my left Ascent i channel back in January 08. Since then, I added absorption material after reading the thread on acoustic panels. Last night I swapped the +/- back on the left channel to put the Ascent i's back in phase. What I heard in my short audio audition sounded great (only had five minutes to listen after the phasing of the system). The mids were out front and the highs were crisp/clean. I plan on listening more this evening and running the Audyssey test again. Thanks for the great topics and conversations!

 

[Statman]

Glad you've figured it out!

I own a 3803CI and haven't used Audeyssey yet. I just used my Radio Shack meter and my ears. Maybe one of these days when I have time to experiment.

 

[sleepysurf]

This thread illustrates how the addition of EQ processing further complicates the delicate relationship between room acoustics/treatment and speaker positioning/performance. Throw room EQ measurements into the mix, and you REALLY end up with lots of variables to tweak! This will certainly keep me busy for quite some time. Someday, I hope to take a break from all this tweaking and do some actual listening!

 

[SteveInNC]

Thread hijack, sorry...

So do you various Denon owners like your 3808/4308 units? I started a thread inquiring about the 5308... and got zero feedback. I've also considered the above units as reasonable alternatives along the price/performance curve (plus I'm looking to build a house in the immediate future, so I have budgeting trade-offs). Any feature(s) you wish your unit had that makes you regret not buying further up the lineage? Doing the N-way comparison on the Denon site, I get stuck with the "yeah, I'd really like to have that, but is it worth an extra $1-2K?" mental gymnastics.

 

[sleepysurf]

...Doing the N-way comparison on the Denon site, I get stuck with the "yeah, I'd really like to have that, but is it worth an extra $1-2K?" mental gymnastics.

That's how I ended up with the 3808CI. The only feature the higher lines offer, that I *might* have used, is HD Radio, but frankly, the myriad Internet Radio options are more than sufficient. All the Denon models (even the vaunted AVP-A1HDCI separate) have a LOUSY interface for streaming music and accessing Internet Radio, so a Squeezebox is much better for that. For driving ML's you really need a separate two or multi-channel amp anyways, but the Pre/Pro section of the 3808, especially with Audyssey MultEQ XT, is excellent. I presume the AVP-A1HDCI might sound even better, but am not yet convinced of that from reading the early reviews.

The main remote, BTW, is identical for all the Denon's (even the $7000 separate), and is a horrible design. You'll probably want to get a Harmony or other universal remote. I'm starting to look for one now.

To summarize, IMHO, the 3808 provides 90+% of the functionality of the higher models, at a much sweeter price.

 

[SteveInNC]

That's how I ended up with the 3808CI. The only feature the higher lines offer, that I *might* have used, is HD Radio, but frankly, the myriad Internet Radio options are more than sufficient. All the Denon models (even the vaunted AVP-A1HDCI separate) have a LOUSY interface for streaming music and accessing Internet Radio, so a Squeezebox is much better for that. For driving ML's you really need a separate two or multi-channel amp anyways, but the Pre/Pro section of the 3808, especially with Audyssey MultEQ XT, is excellent. I presume the AVP-A1HDCI might sound even better, but am not yet convinced of that from reading the early reviews.

Thanks. I almost bought the 3808 last weekend anyway based on my perceived value .vs need to save for the house, but the local Magnolia just ran out of stock (arghh!!). If I get anything, I'll probably buy through Crutchfield who have a no questions return policy for 30 days (they even pay shipping both ways).

The main remote, BTW, is identical for all the Denon's (even the $7000 separate), and is a horrible design. You'll probably want to get a Harmony or other universal remote. I'm starting to look for one now.

To summarize, IMHO, the 3808 provides 90+% of the functionality of the higher models, at a much sweeter price.

I've had a Harmony 880 for a few years, so I'd likely stick with that for normal use, and use the Denon remote for the occassional set-up tweak. This is what I do with all of the other devices I have. I had a guest over the other day who counted fourteen remotes on the bottom shelf of my coffee table, all superceded for daily use by the 880.

A couple of questions re: 3808:

Various sources say that this unit's Audyssey only supports six sample points, but then I see other mentions of the 3808 specifically allowing up to eight (such as page 26 in the manual). Which is right?

The 3808 as opposed to the 4308 supports AL24 only on front right and left as opposed to all outputs - do you think it matters? My surrounds are inferior anyway, so I doubt it matters to me in the short-term, but I like to future-proof myself.

HD Radio is a non-starter for me. I would probably play with internet streaming, might addd an XM unit if I get tired of the DirecTV feed of same, and the lack of WiFi is also not important. I have a Musical Fidelity A5 Int that I use for the mains in pass-through mode, so the amp stages would only be for the surrounds.

What I'd miss from the 5308 is the extra sub-outs, the improved video scaling, better DACs, and the additional HDMI inputs. Multiple Zones is more of a future-proof thing for me than a "need".

 

[sleepysurf]

A couple of questions re: 3808:

Various sources say that this unit's Audyssey only supports six sample points, but then I see other mentions of the 3808 specifically allowing up to eight (such as page 26 in the manual). Which is right?

The 3808 as opposed to the 4308 supports AL24 only on front right and left as opposed to all outputs - do you think it matters? My surrounds are inferior anyway, so I doubt it matters to me in the short-term, but I like to future-proof myself.

HD Radio is a non-starter for me. I would probably play with internet streaming, might addd an XM unit if I get tired of the DirecTV feed of same, and the lack of WiFi is also not important. I have a Musical Fidelity A5 Int that I use for the mains in pass-through mode, so the amp stages would only be for the surrounds.

What I'd miss from the 5308 is the extra sub-outs, the improved video scaling, better DACs, and the additional HDMI inputs. Multiple Zones is more of a future-proof thing for me than a "need".

The 3808 supports up to EIGHT Audyssey measurement points.

Having AL24 only on the Front L R speakers shouldn't be a problem. I would expect it to be audible mainly for two-channel listening (if that).

I only have ONE sub, so having extra sub-outs wasn't an issue. I suppose you could use a Y-adaptor if necessary.

As for the DAC's, while the top tier Denons use the BB 1796 chips, the 1791A's sound great to me. In fact, switching between the 3808 and my Benchmark DAC-1, I can't discern a difference!

Crutchfield is certainly a trustworthy online source, but you'll pay a bit more. I got a great deal on mine from another authorized online source, as mentioned in this thread...
http://www.martinloganowners.com/~tdacquis/forum/showpost.php?p=60697&postcount=10

 

[SteveInNC]

I've bought stuff before from 6ave without problems; in this case, their web price is no better than Crutchfield. Did you negotiate a lower price on the phone, or was that special pricing at the time?

 

[sleepysurf]

Just a f/u re the phase issue. I spent all morning moving my Summits to and fro, inches at a time, then ran a single Audyssey test for each location. Turns out, I was able to find quite a few positions, ranging 3-5' from wall, where there was no measurable phase error. In most cases, simply adding more toe-in eliminated the phase error. Placing silk ficus trees behind the speakers also helped. Since I ended up with so many more placement options than previously thought, I ended up selecting the best sounding placement without Audyssey, then tweaked the toe-in and ficus diffusion, and ran Audyssey to ensure the Summits still measured in phase. Thus, for those of you using Audyssey, and getting the phase error, consider it an objective measure that should not be ignored. Keep tweaking the speaker to wall distance, toe-in, and use of acoustic treatments, until the error is eliminated.