Audyssey's Next Generation of Room Correction (MultEQ-X)

[GalZohar]

I disagree with this. SubEQ chooses delays for each sub, including the relative delay between them, that optimizes frequency response for the subwoofers.

Maybe you were referring to subs+speakers? Audyssey can't nail that down because the user is free to change crossover settings, which differs from Dirac.

How does it choose each subwoofer delay? As far as I'm aware its only capability is to measure the acoustic delay of each subwoofer relative to some timing reference internal to the AVR. As far as I'm aware, there is no guarantee that this will result in optimal summation of the 2 subwoofers, although I can see why theoretically one might expect it to. In reality what I understand is that the sound is mostly reflections and not direct sound, so delay measurement is not that useful for getting optimal response, although in some (or many?) cases it might be good enough. I don't have statistics to tell.

Just an example of some complaints about SubEQ HT:
https://www.reddit.com/r/hometheater/comments/a0kwwu
I do have personal experience with subwoofer-main integration on my system, where the Audyssey distances made good integration of the subwoofer with the left speaker, but very bad integration with the right speaker. Increasing the distance so that the right speaker becomes optimal made the left speaker bad. Increasing distance some more got a result that's almost as good as the best result for the left speaker, and also almost as good as the optimal result for the right speaker.
What makes it even more strange is that the subwoofer is actually very close to the right speaker, the subwoofer is at the corner and speaker a bit away from the corner (woofer ~1m from each wall, maybe a bit more than 1m from right wall), and both the right speaker and subwoofer have rather similar responses at the crossover region (so they "activate" about the same room modes). But based on the result I suppose none of this actually matters for the ability of Audyssey to perform the integration (if you can even call it "integration").
In any case I agree that the crossover must be chosen before subs and mains can be integrated. But if the Audyssey approach of just measuring delay and setting distance accordingly was a good one, then the crossover would actually not matter (as everything is "time aligned", which apparently is just not good enough).

 

[Chromatischism]

How does it choose each subwoofer delay? As far as I'm aware its only capability is to measure the acoustic delay of each subwoofer relative to some timing reference internal to the AVR. As far as I'm aware, there is no guarantee that this will result in optimal summation of the 2 subwoofers, although I can see why theoretically one might expect it to. In reality what I understand is that the sound is mostly reflections and not direct sound, so delay measurement is not that useful for getting optimal response, although in some (or many?) cases it might be good enough. I don't have statistics to tell.

I honestly don't know. I just know that in my experience over the last 7 years in 2 houses, it gets over 90% of the way there and when I make tweaks I can only make very small improvements that I'm not even sure are audible.

I do have personal experience with subwoofer-main integration on my system, where the Audyssey distances made good integration of the subwoofer with the left speaker, but very bad integration with the right speaker. Increasing the distance so that the right speaker becomes optimal made the left speaker bad. Increasing distance some more got a result that's almost as good as the best result for the left speaker, and also almost as good as the optimal result for the right speaker.

I hope you weren't expecting it to because it wouldn't be possible to get a perfect sum with speakers in different locations. I would instead align them the best you can to L+R (not individual speakers, most bass is mixed to mono) if you are majority music, or to the center if you're mostly watching movies with the system.

 

[GalZohar]

I hope you weren't expecting it to because it wouldn't be possible to get a perfect sum with speakers in different locations. I would instead align them the best you can to L+R (not individual speakers, most bass is mixed to mono) if you are majority music, or to the center of you're mostly watching movies with the system.

What I was surprised is how great of a difference there was between optimal delay for aligning with left speaker, vs right speaker, vs the optimal delay for best compromise that gave as close to optimal as possible for both speakers. I don't have a center, but once alignment was almost-optimal with both L and R, obviously L+R gave almost optimal results as well. Optimizing L+R signal is problematic as it may neglect issues that appear in only 1 of the channels, while my method of "best compromise" tries to fix as many of those issues as possible which also results in a pretty good L+R result. In my case, while the distance for "best compromise" was quite different than both optimal distance for L as well as the optimal distance for R, the final "best compromise" result of L, R and L+R (representing "center") was quite close to the the individual optimal results for each of the 3 channels.

As for subwoofers, for some reason I doubt that if physical distance is so bad for speaker-sub integration, why would it be good for 2 subwoofers? The only thing the subwoofers have going for them is identical drivers, meaning identical anechoic phase response. But still, as far as I'm aware in a room the phase response is greatly affected by the acoustics, which may make using pure delay like Audyssey does not optimal. But I guess it works at least fine for many cases, or else it would have never been implemented.

 

[Chromatischism]

I will say that my subwoofer placement is the best that you can do in this room but if I can input a still squiggly line with ~17 dB variance and get a more or less straight line with +/- 2 dB, I'm pretty happy. Maybe I'm just lucky.

 

[chych7]

I will say that my subwoofer placement is the best that you can do in this room but if I can input a still squiggly line with ~17 dB variance and get a more or less straight line with +/- 2 dB, I'm pretty happy. Maybe I'm just lucky.

How's the response at places other than MLP?

 

[Chromatischism]

How's the response at places other than MLP?

I don't have measurements to share, but it's pretty good, though I can hear a difference. I don't sit in the other seats so it's not really important to me. Maybe later on it could be.

 

[GalZohar]

As far as I'm aware, statistically the Audyssey way of time aligning the sub works "OK", but a more precise method like REW or MSO will find the optimal delay, which Audyssey simply can't do. It just hopes that the delay it calculates will be optimal but in practice it's just often enough close enough to optimal. Again I base this on experiences I gathered online (you are an example of a case where it worked pretty well, although you should probably try find a way to compare to optimal alignment, and the above reddit post is an example for a case it worked poorly), and the technical capabilities of each of the methods (Audyssey being limited to choose a delay without calculating how that will actually affect the frequency response).

 

[peng]

How does it choose each subwoofer delay? As far as I'm aware its only capability is to measure the acoustic delay of each subwoofer relative to some timing reference internal to the AVR. As far as I'm aware, there is no guarantee that this will result in optimal summation of the 2 subwoofers, although I can see why theoretically one might expect it to. In reality what I understand is that the sound is mostly reflections and not direct sound, so delay measurement is not that useful for getting optimal response, although in some (or many?) cases it might be good enough. I don't have statistics to tell.

Just an example of some complaints about SubEQ HT:
https://www.reddit.com/r/hometheater/comments/a0kwwu
I do have personal experience with subwoofer-main integration on my system, where the Audyssey distances made good integration of the subwoofer with the left speaker, but very bad integration with the right speaker. Increasing the distance so that the right speaker becomes optimal made the left speaker bad. Increasing distance some more got a result that's almost as good as the best result for the left speaker, and also almost as good as the optimal result for the right speaker.
What makes it even more strange is that the subwoofer is actually very close to the right speaker, the subwoofer is at the corner and speaker a bit away from the corner (woofer ~1m from each wall, maybe a bit more than 1m from right wall), and both the right speaker and subwoofer have rather similar responses at the crossover region (so they "activate" about the same room modes). But based on the result I suppose none of this actually matters for the ability of Audyssey to perform the integration (if you can even call it "integration").
In any case I agree that the crossover must be chosen before subs and mains can be integrated. But if the Audyssey approach of just measuring delay and setting distance accordingly was a good one, then the crossover would actually not matter (as everything is "time aligned", which apparently is just not good enough).

There seem to be a lot of talk about XT32+Sub EQ HT not being able to integrate/align multiple subs. I have been using XT32 SubEQ HT since 2013 and always have good results in the alignment of my two different SV Subs. Still, I do intend to try MSO eventually but it will be just for fun. As for the minidsp, in my experience it is completely unnecessary and it has been put back to the box after experimenting, though I do suspect if I have 1 or two more subs then the mini might have help. Like Chromaticism, I am getting +/- 1.0 to 2 dB (about +/- 3.5 dB if not using the App+Ratbuddyssey) or better in the 20-200 Hz range ignoring a couple narrow sharp dips, even when averaged for many mic positions within the listening bubble, any slight improvements will not result in audible difference to me anyway. So I am not saying MSO etc., can/will help with integrating subs and/or subs+speakers, but while the results may well be visible, they may not (it may) really make audible difference.

To be clear I am not defending Audyssey and/or Sub EQ HT, just want to share my own experience and would welcome D+M switching to Dirac as long as they include the version that can handle two or more subs. To those who complains about poor result using Sub EQ HT to integrate their multiple subs or subs + speakers, I would suggest they check the results with REW first to see what's going on, then go from there and take whatever remedial action (MultEQ X, Editor App/Ratbuddssey, minidsp, MSO etc.) necessary based on the findings.

Just an aside, one thing I did find is that while Audyssey seemed to produce reasonable impulse response for me, Dirac did seem to do visibly better consistently. I only have the 2 channel PC standalone version so I was only comparing the two in stereo.

Regarding "complaints", among SubEQ HT users, there are going to be complaints of all sorts for sure. Just check out many Amazon reviews, even among the 5 star ones you will find many complaints. Some may be valid, but many are not, or anecdotal at best.

 

[GalZohar]

The fact is while MSO, ARC and probably other algorithms could time align the sub in a similar way to Audyssey, they don't do it. I think it's not far fetched to assume the other methods provide better results and the Audyssey way is a compromise which is a result of system limitations.

 

[Chromatischism]

Again I base this on experiences I gathered online (you are an example of a case where it worked pretty well, although you should probably try find a way to compare to optimal alignment,

I couldn't improve on it.

 

[tjcinnamon]

Without DEQ, I’d prefer to not have Dirac. Also, Dirac is a bit of a black box and doesn’t offer the same customization vs MQX. For example, skipping measurements or retaking just one. Also, for home theater applications the Dirac measurements take a very long time.

 

[peng]

Without DEQ, I’d prefer to not have Dirac. Also, Dirac is a bit of a black box and doesn’t offer the same customization vs MQX. For example, skipping measurements or retaking just one. Also, for home theater applications the Dirac measurements take a very long time.

Good point, I wish Dirac would come up with something like DEQ and hopefully a more effective one than Audyssey and YPAO.

 

[Dj7675]

Good point, I wish Dirac would come up with something like DEQ and hopefully a more effective one than Audyssey and YPAO.

Loudness compensation for Dirac appears to be left up to the hardware manufacturer for Dirac and I wouldn't expect that to change. There are models such as the Monoprice HTP1 that have loudness compensation. It is unfortunate that many don't, however.

 

[peng]

Loudness compensation for Dirac appears to be left up to the hardware manufacturer for Dirac and I wouldn't expect that to change. There are models such as the Monoprice HTP1 that have loudness compensation. It is unfortunate that many don't, however.

I think you are right, but I am thinking about the rumor that D+M will be using Dirac instead of Audyssey so in that case if D+M uses the same kind of menu, you might see the DEQ setting only if Dirac is on.

 

[Dj7675]

I think you are right, but I am thinking about the rumor that D+M will be using Dirac instead of Audyssey so in that case if D+M uses the same kind of menu, you might see the DEQ setting only if Dirac is on.

I could be wrong, but I would sure be surprised if D+M offered Dirac as an option. I saw that rumor as well but it would seem like a strange move after the Audyssey PC app and the continued development of it.
Until D+M separates their loudness compensation from boosting surrounds, I consider DEQ broken for surround system usage. I liked it for bass compensation but hated it for surround level boosting so had to stop using it when I had my 2 different Denon units.

 

[tjcinnamon]

I could be wrong, but I would sure be surprised if D+M offered Dirac as an option. I saw that rumor as well but it would seem like a strange move after the Audyssey PC app and the continued development of it.
Until D+M separates their loudness compensation from boosting surrounds, I consider DEQ broken for surround system usage. I liked it for bass compensation but hated it for surround level boosting so had to stop using it when I had my 2 different Denon units.

Agreed. Now that Audyssey has new features I think it needs to be reevaluated vs Dirac.

 

[chych7]

Just an aside, one thing I did find is that while Audyssey seemed to produce reasonable impulse response for me, Dirac did seem to do visibly better consistently. I only have the 2 channel PC standalone version so I was only comparing the two in stereo.

I had found the same in my testing on impulse response; Audyssey does improve it vs. no correction, but Dirac does better. However regardless of this difference, Audyssey still sounded better to me, when both DRCs were running full range. I guess impulse response isn't everything, when it comes to subjectivity. What were your subjective findings?

 

[Chromatischism]

DEQ is only "broken" for surround use if you don't correct what it's doing.

 

[peng]

I had found the same in my testing on impulse response; Audyssey does improve it vs. no correction, but Dirac does better. However regardless of this difference, Audyssey still sounded better to me, when both DRCs were running full range. I guess impulse response isn't everything, when it comes to subjectivity. What were your subjective findings?

When comparing it in my 7.1.4 HT system, I could only compare Audyssey ref vs pure direct without too much delay in between the two modes. To compare Audyssey Ref and DL3 PC standalone version seem complicated and there will be too much delay going from one to the other even if I can find a good way to do it. So I have not even tried yet.

 

[peng]

I could be wrong, but I would sure be surprised if D+M offered Dirac as an option. I saw that rumor as well but it would seem like a strange move after the Audyssey PC app and the continued development of it.
Until D+M separates their loudness compensation from boosting surrounds, I consider DEQ broken for surround system usage. I liked it for bass compensation but hated it for surround level boosting so had to stop using it when I had my 2 different Denon units.

Agreed, though as someone mentioned, it doesn't seem to be as bad lately, may be one of the FW updates changed something or it could just be getting used to it. Regardless, I have been experimenting with lowering the surround channels using the trims and I like it better simply by lowering the levels by 1 to 2 dB. I tried more and didn't like it as much. It should be fixable via FW but I guess D+M won't do it unless enough people complain about it.

By the way, you have the Trinnov Atitude right, does it has a similar feature?