"Official" Audyssey thread (FAQ in post #1) - Page 1801

Someone moving from XT to the 818 will have made a big jump in performance. And that is, indeed, a very good deal. But I remember the disappointment from us AND Chris when it was learned that an early implementation did not include Sub EQ HT. So some disappointment that it is now being omitted after a good solid period of it being included is understandable.

Of course, this "disappointment" is pretty much limited to enthusiasts on fora who have watched the evolution from 2EQ to XT 32. The A/V buying public won't know the history.

Extra annoying, but maybe a testament to the marketing power of Audyssey, is that, IMO, they are using it to step people up to the next unit.

Welcome back, Keith.

Jeff

I agree, but don't think the 818 user is sacrificing much worth having. Level setting is more or less irrelevant as it is easy to do it manually. Distance setting can be done manually, and Audyssey often needs it tweaking anyway for best results. And as Craig says, it's only right for one seat anyway. The most important thing is two EQ the subs as one, and the 818 does that.

When you say it has been included as a standard part of XT, you are right of course. But for $2,500 not less than $1,000. If I was on a budget and looking for an AVR right now, or a second one for another room, I suspect the 818 would be top of my list. And for people with one sub (most?) the whole issue is irrelevant anyway.


Thanks Jeff. Good to be back.

Hey keith, glad you're on the mend!

^One more thing-has it been verified that the subs be manually set to different delays within the 818? And even if they could, that would not have been taken into account with the Audyssey filter on the sub channel. At this point I cannot in good conscience recommend an 818 as a good one box solution for 2 nonequidistant subs.

But can it though? I remember from your subwoofer optimization experience (at least the very first time with Pro), that you not only had to vary the distance of each individual sub from what Audyssey had set, but also that Audyssey had set different distances for each non-equidistant sub. I think it's both of those capabilities that are lost on the new Onkyo. How much differently would your rig sound and measure if you needed to choose a single subwoofer delay, say 14 feet, for both units? Especially in your case where you have little freedom to move either sub.

Craig suggests this scenario is not a concern but it's hard for me to understand how it wouldn't be. Not saying anyone is right or wrong on this topic, just that I haven't yet been convinced either way. My dual subs are co-located so I'm totally in theory land here. :-)

PS...welcome back!

You would turn the subs DOWN more to make the receiver turn them up. Unless you have a problem like auto on not working correctly there may be no great benefit to changing. But if you do you must repeat the entire autosetup process AFAIK.

How does that 9 band EQ work with audyssey? I am not sure how to put this, I guess I am wondering what roll it plays to get the correct speaker response.

It doesn't work with Audyssey. If engaged, Audyssey is turned off. Not recommended.

Jeff

I guess what they say is true..."The Shadow knows!"

Seriously, thanks a lot for this information.

Mark

Hehe.... I do IT tech support for a living.

Glad to help my AVS family

I can't quote the bit of your post I am replying to because of your unconventional quoting method in your reply, but the point I was making about Audyssey getting the distances 'wrong' was that we know from experience and measuring that the distance settings can be changed for the better in, it seems, more cases than not. So it may not be all that consequential that the 818 doesn't set the individual delays. And, as Craig pointed out, the distance settings can only be right for one seat anyway.

This brings me to the conclusion that the 818, for under a grand, is a great way to get XT32, with most of the benefits and disadvantaged only in two areas (levels and distance) which are perhaps of lesser importance than we customarily assume. Certainly the levels issue is trivial to resolve manually, so it boils down to the issue of the delays not being set for each individual sub - something I believe is of less importance than you do. We're probably both right

That is correct.


It's true that the 818 uses the 'Y-cord' method, but this can give perfectly good results so long as a few basic premises are followed. I agree that not everyone (me!) can place subs equidistantly, but Craig's post references 'gurus' who seem to suggest that this is not a matter for great concern. I agree that the full fat implementation of XT32 is the gold standard, but it also costs around $2,500 to enter the club. The 818 for less than $1,000 seems to me to be a great way into XT32, with a few compromises, but in areas where the impact is not savagely felt. And for people with only one sub, the 818 offers the full fat experience anyway. For a grand!


Thanks Tandy - good to be well again!

Welcome back Keith,

Yes, I concur that it would seem that the exclusion of SubEQ HT from the 818 is not that critical. As someone else has already pointed out, the Welti and Geddes papers are based on multiple subs receiving the same signal, so in essence 2-4 subs all running off the same output.

As more than a few folks with measurement equipment and multiple subs (myself included) have noted, tweaking the sub distances can improve results over the Audyssey settings, it would seem that the lack of individual sub distance settings may not be as much of a drawback as perceived, especially given the 818's pricepoint.


Max

The multiple sub scheme might value mode-free smoothness over coherence. But then I've never seen any testing that suggests that a multiple sub/multiple distance setup lacks coherence. Perception of something like that is likely impaired from the chest-punching and arse-kicking those systems typically have .... Craig.

Jeff

Incoherence??? Impaired??? Sounds like the last time we got together and you made "Weapons-Grade" Margaritas!!!

Those files are not accessible at this time.

So nice to converse with you again, Keith!

Just to add a bit to the "only right for one seat" thought, I ran a multiple-point REW measurement after optimizing the subs using distance tweaks, and all of the measurement points showed improved bass response versus the same REW measurements prior to the distance tweaks. My point is, while we might only use one point as we are making the adjustments, the improvement seems to be over a wider area. I encourage you to take similar measurements to see for yourself.

looks great! nice job as always Keith.

Poor Keith. He's barely out of bed and we want him to take more measurements.

But, your points are sensible to me. If one is worried about matching two subs, the "one right seat" is not a serious issue since the sources of the low frequencies are not identifiable. I.e., there won't be a sound stage shift if the two sub distances are not correct at your seat. But, for maximum SPL, the two subs should be as close to in-phase across the seating area as possible which means they should be in-phase at the MLP. However, recall that the wavelength at 80 Hz is just 14 feet so if you are 7 feet further from one sub than the other, you will be sitting in a null at 80 Hz (for the direct wave). If you are off by 3.5 feet, the subs are off by 90 degrees; measurable and maybe noticeable but not a disaster. If the difference is less than about 2 feet the dip would be hard to measure. The bottom line is you do want the subs in phase at the middle of the listening bubble, but unless the bubble is very large, this won't be hard to do and doesn't need to be perfect.

Matching the subs and mains is a somewhat different issue. Again, you want the two 80 Hz sound waves to be in-phase at the MLP and not too far out of phase in the other seats. But here you are mixing sound from two mains and a center with that from two subs. There's going to be less loss of phase coherence at the side seats since there are five sources involved (and a host of reflections). But, if three of those sources (L, R and C) are not in phase with the two subs, you will have a dip and it will probably affect all seats more-or-less equally. Theoretically more at the off-center seats, but not by much (probably only a db or two). So I'd suggest this tweak is more important than having the two subs well aligned.

This has made me always question putting the subs on the side walls. It seems that keeping the subs in phase across a three-wide seating setup would be difficult. The left seat might be 6 feet further from the right sub than the right seat. And vice-versa for the left sub. Seems problematic. Maybe the reflections tend to swamp out the direct waves in that sub arrangement making this concept moot.

How does the fact that the oft quoted research was conducted without time-aligning multiple subs bolster the notion that time-aligning them doesn't matter?

For example in this scenario, is it mere coincidence that the "best" placements are those where the subs are equidistant from the listening area (i.e. time-aligned)? The author goes on to conclude, "Symmetrical locations seem best." Would mimicking symmetry by adjusting delays in asymmetrical subwoofers result in different recommendations? We'll never know from this study, but this one certainly doesn't provide a conclusion on that topic either way.

And a brief review of this seems to suggest that a 5 or 10ms delay among subs is actually required to optimize response for both the 2 and 4 subwoofer configurations in their simulation. The author goes on to report, "Adding gain–delay–filter optimization lowered the MSV by an average of 13.9 dB2 when two subwoofers were used". Gain-delay-filter optimization reads Sub EQ HT to me.

Since the work of these scientists is increasingly being used as a reason for not worrying about time-aligning individual subs, are there studies by these folks directly addressing this? If so, could someone please send me a link, as what little I've reviewed so far seems to indicate the opposite--that symmetry and/or independent delays matter.

Just trying to learn, and understand if I've been "duped" by paying the extra $1K for Sub EQ HT...

I sometimes use the embedded response technique when discussing multiple issues in a post as I find it quicker and clearer. I didn't realize it had the added benefit of making my points difficult to counter.

I very much appreciate your response and all the viewpoints being expressed on this complex topic of optimizing the bass region. I hope to test out some of these techniques the next time I pull out the Pro kit.

I have a non symetrical room. Three walls opening into a larger space. I have my subwoofer placed in the right corner and my surround speakers to the side of my listening position on stands.

For my AVR I have the Denon 4311ci. For my speakers I have the Polk Audio RTi A7 main speakers Polk CSi A6 and for rear speakers Polk Audio Rti a6. For my subwoofer I have a Rythmik F15HP.

My problem is with how Audyssey is reporting the -3 roll off point to my receiver. I know that the avr ultimately decides where the crossover is set but its set way to high for speakers. My Polk Audio RTi A7 are rated down to 20hz ( I realize this is a liberal estimate) more realistically down to 30-40hz.

my crossover is getting set between 150hz to 200hz on all 5 speakers. I have ran audyssey 5 times and get the same results. It also cranks my subwoofer up really high. On my receiver when it starts auto eq it gives you a level reading of the subwoofer and I have to turn my subwoofer gain up to 3/4 max volume before it reads correct. After the calibration is done the subwoofer is hot by like 20db according to my radio shack meter.

I have no idea whats happening here. Is it possible that my mic is bad or something? What else could be causing this. I thought maybe the mic was in a null. I ran some sweeps and I hear no major dips in frequency.

So confused please help!!!

Audyssey multEQ XT, REW and sub placement?

I am trying to figure out the best place for my sub, any suggestions to figuring this out? How should I run audyssey to get the best sub results for the main seating position vs good for multiple seats?

Under the new AVS Forum platform to be introduced next week (?), including "nested quotes" will be the default setting when pressing the "Quote" button with the option to remove them if desired via your "My Profile" section, so it will be much easier to respond to SOM. (his post manually created below with my sample responses)

Thanks Max.


That's certainly my take on it too. Personally, I'd prefer the full fat XT32 but I can see a lot of instances where XT32 for less than a grand would be a very attractive proposition, albeit without this functionality which some of us think is very important and some of us think is much less important.

Thanks AJ! Good to be back.

I take your point fully and have no reason to disagree. I just think the 818 is being dismissed too readily. The functionality it lacks is, well, lacking for sure. But my view is that it is less important than some of the other guys think. In an ideal world, I'd want XT32 with the ability to set individual distances and levels - but for me it would not be a deal breaker in a unit that costs less than $1,000. XT32 benefits are still present and all that is lost is the distance setting for individual subs, which I know from my own experience can be taken with a grain of salt anyway. As it stands, the 818 is by miles the cheapest way into XT32. Once the Denon 4311s have all been sold, it will be the only sub-$2,000 way in AFAIK.

Thanks batpig. Appreciated.

LOL!!


I'm beginning to come to the conclusion that the guiding credo for all this is "YMMV". I'm not sure there is a right and wrong answer - so much seems to depend on individual rooms, equipment, circumstances and setups. I'd hesitate, for example, to suggest anyone reverses the phase on one sub, but that has given me the best bass I have ever experienced in my room. Experimentation, testing and listening seem to be the way forward, on an individual basis.

^^

+1

Use the guide and, if after allowing a few weeks to adjust to "reference", and if one feels adventurous, start experimenting ... and if the result sounds better ... do it .... after all, in the end it's "preference" that should rule.

Note that in the first link you quoted, the sub positions are symmetrical, but not necessarily equidistant. As the paper points out, they first ran the simulations with the seating positions centered in the room, they then reran the simulations with the seating positions biased towards the back of the room where they were more likely to be placed in a real room and the effectiveness of multiple subs was still evident. It has more to do with the excitation of the modes in the room than equidistant positioning from the listening positions.

They also tested the results in a real room, and noted that the construction of the room and differing rigidity of the walls resulted in some differences compared to the simulations' predictions.

In addition, if you look at the Geddes multisub optimization, unlike the Welti symmetric, even number simulations, Geddes advocates using 3 subs, asymmetrically placed (1 in a corner close to the main speakers, 1 midwall on a side or rear wall, and a 3rd preferably away from the first 2. 1 sub should also be elevated). As with the initial Welti/Devantier simulations, the Geddes placements are based on all subs receiving an identical signal.

The goal of both is to minimize the seat-to-seat variance (note that the goal isn't to initially achieve the flattest frequency response at any one seat). The idea (and practice) is that when seat-to-seat variance is minimized (even if the resultant frequency response is quite uneven), that uneven frequency response can THEN be equalized flat/smooth, and the result will be a similarly smooth at the multiple seating positions.

This is why it's been mentioned in this thread, that even with Audyssey, if a user has measurement equipment available AND some flexibility with subwoofer and especially multiple subwoofer placement, that the goal is to place the sub(s) pre-EQ such that the seat-to-seat variance in the room is minimized. This is because the resulting frequency response is the easiest for Audyssey to equalize. Instead of reading different frequency responses at different measurement locations and needing to decide which peaks and dips it should attempt to correct (at the expense of potentially exacerbating peaks and dips at other locations), if all the measurement locations show similar peaks at the same frequencies and dips in the same places, then Audyssey can simply flatten the FR and the result will be flat at multiple positions.


Max

Yes I left Geddes out because his technique didn't support my case. ;-)


Totally agree! Flexibility is critical to getting the very best (and least variable) pre-EQ response. In that sense, independent distance controls on the subs are another tool in the chest and offer more flexibility towards achieving that goal...especially when one can't be flexible on placement.

Were I currently and forever planning to use a single sub, equidistant subs, and/or non-equidistant subs that are already known to measure well when receiving an identically-timed signal, I would be quite comfortable buying an AVR that omits the individual sub delays. Short of that, though, I would rather pay for the flexibility and tweakability and not need it than need it and not have.

Is it alone worth an extra 100% to the MSRP of an AVR? Heck no, but I have to suspect the new Onkyo receiver is also making many other compromises compared to the 4311 to get to such a low MSRP. But that's in/for another thread all together!

Thanks for the additional insights!