Sanity check: Bad Sub ? (REW Sweep results)

[TLS Guy]

Unless I'm misremembering Erin's measurements, the Q 11 Meta's port tuning is 35 hz or so. Their response is dropping like a stone below 30 hz. Feeding them the lfe channel seems pointless and potentially risky.

It could be for movies, but in music there is little content below 30 Hz. I run my rig with all speakers except the ceiling speakers full range. I have never had a problem. I highly doubt his left and right speakers will be damaged though. Minimizing the crossover settings improves the transient response. When I use crossover to all the speakers, it is unquestionably a downgrade. The crossover from the mid lines to the bass lines is acoustic, and the LFE added. After all before this bass management craze we always ran our speakers full range, and I never had a problem, which is most of my life. I run my family room speakers full range and push them hard at times. Mind you, KEF B 139s are really robust and capable speakers. Falcon acoustics are producing them again, and reporting brisk sales. That driver goes back to 1969. I don't think a modern driver of that size, and many that are larger, can compete with them.

 

[Linwood]

...common sense is a much better starting point. Just because it is a fancy program with a name given by an ad. exec, it does not mean it is right by a long shot. So don't slavishly trust them ever.

Thanks. My current mode is first I have to figure out what it is doing before I start undoing anything.

 

[Linwood]

Crossovers:

So I've listened to hours or videos and read a bunch of writeups and I honestly do not know how crossovers work.

Some (and google's AI) say that you set them in the manual section of speaker setup before using Dirac Art, but other indications are the closest you get is using the F Support Low which is a cutoff at which that speaker does not provide "support" (which is the cancellation of standing waves as I understand it), but it is not clear that's the traditional crossover point, and certainly it doesn't indicate it matters where you start.

I listened to one Direct Engineer discuss some related things, like subwoofer phase, and his reaction was "doesn't matter, it will figure it out and do the right thing". Though he did say the support cutoffs needed some attention.

So to the original question: I have no idea how it's set, and whether the setting before you start matters (because it's painful otherwise as I don't know how you get back into that setting if you wanted to change it shot of some kind of reset).

I am starting to understand ART's philosophy, though there's so much contradictory advice out there (e.g. whether or not to let the center support other speakers; the Dirac Engineer seemed to say for all speakers they should provide support unless the speaker is already too close to running out of capacity just to play).

And another mystery to me is that ART is supposed to understand the acoustic structure of the space in the room, where standing waves will interfer or re-inforce so it can act on them at different frequencies, and to do this at different locations -- but there's nothing in the measurements that provide precise location information. Is the space to the left at 2', 3', 1.5'... the system doesn't know. All the measurements seem relative locations, not quantitative distances.

 

[Kingnoob]

It could be for movies, but in music there is little content below 30 Hz. I run my rig with all speakers except the ceiling speakers full range. I have never had a problem. I highly doubt his left and right speakers will be damaged though. Minimizing the crossover settings improves the transient response. When I use crossover to all the speakers, it is unquestionably a downgrade. The crossover from the mid lines to the bass lines is acoustic, and the LFE added. After all before this bass management craze we always ran our speakers full range, and I never had a problem, which is most of my life. I run my family room speakers full range and push them hard at times. Mind you, KEF B 139s are really robust and capable speakers. Falcon acoustics are producing them again, and reporting brisk sales. That driver goes back to 1969. I don't think a modern driver of that size, and many that are larger, can compete with them.

With my avr to to run mains full range, I think I have to select double bass or the sub won’t get that lfe channel the mains are doing ? My icons only go down to 35hz or so I run them at 60. Ancient onkyo.
Dirac Art cost a lot you need a new avr that supports it and multiple subs ?

 

[Linwood]

Dirac Art cost a lot you need a new avr that supports it and multiple subs ?

The Denon x3800h supports it, and getting two SVS PB-1000's, second coming tomorrow.

From a lot of reading and watching I'm convinced (a) I have a terrible room setup but don't intend to fix it, and (b) it does the most to fixing bad rooms. Not that it can be a complete fix.

 

[lovinthehd]

Crossovers:

So I've listened to hours or videos and read a bunch of writeups and I honestly do not know how crossovers work.

Some (and google's AI) say that you set them in the manual section of speaker setup before using Dirac Art, but other indications are the closest you get is using the F Support Low which is a cutoff at which that speaker does not provide "support" (which is the cancellation of standing waves as I understand it), but it is not clear that's the traditional crossover point, and certainly it doesn't indicate it matters where you start.

I listened to one Direct Engineer discuss some related things, like subwoofer phase, and his reaction was "doesn't matter, it will figure it out and do the right thing". Though he did say the support cutoffs needed some attention.

So to the original question: I have no idea how it's set, and whether the setting before you start matters (because it's painful otherwise as I don't know how you get back into that setting if you wanted to change it shot of some kind of reset).

I am starting to understand ART's philosophy, though there's so much contradictory advice out there (e.g. whether or not to let the center support other speakers; the Dirac Engineer seemed to say for all speakers they should provide support unless the speaker is already too close to running out of capacity just to play).

And another mystery to me is that ART is supposed to understand the acoustic structure of the space in the room, where standing waves will interfer or re-inforce so it can act on them at different frequencies, and to do this at different locations -- but there's nothing in the measurements that provide precise location information. Is the space to the left at 2', 3', 1.5'... the system doesn't know. All the measurements seem relative locations, not quantitative distances.

It's a fairly simple concept, just where high pass filter and low pass filter slopes "cross" so as to create a relatively flat response. Then there are different slopes and approaches to automated setting of ideal crossover (even if slope is limited, so....).

 

[Kingnoob]

The Denon x3800h supports it, and getting two SVS PB-1000's, second coming tomorrow.

From a lot of reading and watching I'm convinced (a) I have a terrible room setup but don't intend to fix it, and (b) it does the most to fixing bad rooms. Not that it can be a complete fix.

yeah I meant it’s not cheap for Dirac Art I can probably get my room sounding good without fancy eq and measurements. Can’t afford a $1500 receiver right now.
Ahh I wish I could get better subs but only able to afford a single if I got one. Not happening til something breaks.

 

[ski2xblack]

Happy turkey day! @Linwood It seems you're beginning to appreciate the challenges.

Think about your initial impressions when your speakers arrived, when you first noticed how good the bass was, when your sub was not even plugged in! That's because your speakers are pretty much full range, with the bass integrated into a cohesive whole, courtesy of the engineers at KEF.

Now think about what you're faced with by adding subs. You're in the unenviable position of donning your own speaker engineer hat, to add subs while maintaining/preserving that integrated, cohesive whole (no small task). Or let Dirac tackle it for you-always verify using REW.

High passing your mains will have some benefits, mainly reduced modulation distortion, and allow a bit more dynamic wiggle room. But that also means the pressure is on to get the crossover correct, lest it result in less than integrated, cohesive results and audible tells.

We've already mentioned some of the options to you, some which sidestep having to conjure the perfect crossover, at least somewhat.
-The subs to main L and R are most important to attend to, so start there.
-You should probably let the Q 11s run full range.
-Let the SVS subs handle the lfe (leave the lfe channel lpf at 120 hz)
-Experiment with the sub's lpf settings and see how it measures. Try 40-60 hz or so, measure, tweak, remeasure... Follow the procedure from the video you posted, it had some great calibration advice.

So under this scenario, rather than a textbook crossover (e.g. LR4 implemented at 80 hz, the most typically recommended target), your subs and L and R speakers will overlap, providing four sources of bass which should aid with modal smoothing.

After that you can decide whether or not to high pass your center and surrounds. They're all ported designs, so the inherent phase shifts could be an issue, so experiment as needed.

You could always try high passing all your speakers, as is typically done, and who knows, it might end up working better in your situation. It would be a bit safer, if you're blasting it at disturb-the-neighbors spls. Just realize, higher crossover settings means dividing the fundamentals and harmonics of recorded instruments and divvying them out to separate speakers, which is ripe ground for errors, so you (or Dirac) need to match the competence of those KEF engineers to arrive at that integrated, cohesive result. With the subs essentially colocated with your L and R, getting the crossover right is easier to pull off, but might not help as much with modes.

 

[Linwood]

Thank you, @ski2xblack , but I have a big challenge with that since getting Dirac Art.

Dirac Art seems to turn off all crossover control, take it over, and replace it with something that's different, the low/high "support" ranges that don't appear to be crossovers in the traditional sense because they indicate which speakers can "help" which other speakers with resonance cancellation, as opposed to a handoff of responsibility. There are hints that the crossover setup before you start has some relevance, but after even 1 filter is applied, you can no longer see them or change them, so they can't be adjusted for additional filters (I guess without a factory reset?). So I doubt they are relevant as something this engineered would not forget about being able to adjust.

So all the discussion on crossover type issues seems moot with ART? Or am I just missing a control?

By the way, I do not in any mean to ignore that these issues you mention exist, and am trying to learn to understand them better. But having took the ART plunge... I don't see how to apply any of that advice? It very much seems to be a "trust the magic" combined with a change from thinking crossover to thinking support ranges and target curves.

 

[ski2xblack]

I wish I knew the answer, I have zero experience with DLBC or Dirac ART. But it's a powerful tool that takes care of all those messy details for you, so maybe take a "trust the magic, but verify" approach? Let it do it's thing and see (hear) how it sounds.

 

[TLS Guy]

Thank you, @ski2xblack , but I have a big challenge with that since getting Dirac Art.

Dirac Art seems to turn off all crossover control, take it over, and replace it with something that's different, the low/high "support" ranges that don't appear to be crossovers in the traditional sense because they indicate which speakers can "help" which other speakers with resonance cancellation, as opposed to a handoff of responsibility. There are hints that the crossover setup before you start has some relevance, but after even 1 filter is applied, you can no longer see them or change them, so they can't be adjusted for additional filters (I guess without a factory reset?). So I doubt they are relevant as something this engineered would not forget about being able to adjust.

So all the discussion on crossover type issues seems moot with ART? Or am I just missing a control?

By the way, I do not in any mean to ignore that these issues you mention exist, and am trying to learn to understand them better. But having took the ART plunge... I don't see how to apply any of that advice? It very much seems to be a "trust the magic" combined with a change from thinking crossover to thinking support ranges and target curves.

That is what Dirac does, which is why I did not buy it.

I can't emphasize enough to go by the dictum: -" Keep it simple Stupid!"

You liked the sound of your new mains. So run them full range. You will get the best phase response that way. Now for the moment forget all this auto Eq, and add in your sub gently below 60 Hz, see what level sounds best to you and integrates the sound well with the mains.

Once you are happy, then you can set levels and distance with Audyssey. Listen to the rig for a few days. Then activate Audyssey Eq. and see if you like it better, or if it sounds worse. If it sounds worse or no better, then leave it off. There are a lot of sound engineering reaons as to why those sort of programs are bogus.

With your system, and your excellent mains, Dirac will muck it up for sure, as its bass management makes no sense, none whatsoever, which is why I did not buy it.

I do not use Audyssey Eq, on any of my three systems. I can do better without it, although in my AV 10 pre/pro it was neutral which is an improvement over previously, where it made it worse. Anyhow I switched it off to save processing power. I get really good results without resorting to those programs. I wonder all the time how many would be better off without them, and I have a feeling it is a lot, especially for those with really good neutral speakers.

 

[Linwood]

I appreciate the advice but I'm going to keep trying to figure out Dirac ART. I may fail, but going to try.

 

[lovinthehd]

Might be some good info in this thread https://www.audiosciencereview.com/forum/index.php?threads/active-room-treatment-art-by-dirac.40743/

 

[TLS Guy]

I appreciate the advice but I'm going to keep trying to figure out Dirac ART. I may fail, but going to try.

That makes zero sense. The inability to run your speakers full range will be a big impediment with your KEF speakers. if you use Dirac then the speakers have to be crossed over. That is BAD.

In any task you start with the simplest solution and only add complexity when you have to. At the moment you have no idea what the native sound of your system is. That is always the appropriate starting point. If you don't do that then you won't know how much those programs are mucking your system up, which they often do.

The better your equipment, the less you need those programs. In other words you are looking for a solution for a problem to which you have no idea if you have one.

 

[Linwood]

That makes zero sense. The inability to run your speakers full range will be a big impediment with your KEF speakers. if you use Dirac then the speakers have to be crossed over. That is BAD.

I need to see that it really does that. My brief experiments does not show that the fronts are NOT running full range, in fact I can also lower their support of the subs down into their lower range, say 35 vs 50 it defaults to. I need to get a ART run I like, then do some actual measurements to see what it did (as opposed to what it says it is doing).

I get the push for simple. But please understand what's simple and obvious to you is not to me, and ART promises (to be seen if it delivers) to provide a narrowing of the possible steps toward correcting for an awful room.

 

[TLS Guy]

I need to see that it really does that. My brief experiments does not show that the fronts are NOT running full range, in fact I can also lower their support of the subs down into their lower range, say 35 vs 50 it defaults to. I need to get a ART run I like, then do some actual measurements to see what it did (as opposed to what it says it is doing).

I get the push for simple. But please understand what's simple and obvious to you is not to me, and ART promises (to be seen if it delivers) to provide a narrowing of the possible steps toward correcting for an awful room.

You are a supreme optimist in that case. We did very well without all those auto programs, in my case more then 60 years, and I still don't use them, as they can not improve my systems. In fact I have found their benefit is marginal to none, and easily harmful. They add huge complexity and not simplicity. More likely than not you will "chase your tail."

There is nothing simpler than starting off, how I suggest. I have a very strong hunch with the equipment you have bought it will sound good. Then you can see of you can improve it with those complex programs. There is a high chance you won't. I can't stress enough that knowing the native sound of your equipment before working with Audyssey or Dirac, is not optional, but absolutely essential.

I have just checked, that Dirac ART "takes full control of your mains and sub". It also states optimistically that it "treats them as one." If you believe that, you will believe anything. I have never used Dirac, so I have to own up to that. But to me it seems like an absolutely ghastly program based on a lot of false premises, which I am certain are not correct. They get the old "Click and Clack: - "BOGUS" as far as I am concerned. So they won't get any funds from me.

 

[Linwood]

There is nothing simpler than starting off, how I suggest. I have a very strong hunch with the equipment you have bought it will sound good. Then you can see of you can improve it with those complex programs. There is a high chance you won't. I can't stress enough that knowing the native sound of your equipment before working with Audyssey or Dirac, is not optional, but absolutely essential.

I have just checked, that Dirac ART "takes full control of your mains and sub". It also states optimistically that it "treats them as one." If you believe that, you will believe anything. I have never used Dirac, so I have to own up to that. But to me it seems like an absolutely ghastly program based on a lot of false premises, which I am certain are not correct. They get the old "Click and Clack: - "BOGUS" as far as I am concerned. So they won't get any funds from me.

Well, it's an interesting excursion none the less.

Here's an example that I am working on right now. This is Left/Right Q11's. Don't blame it for the target curve (heavy bar with control points), that's me. The solid line (sorry, couldn't figure out how to change the yellow, iut's almost invisible) is the measured response averaged across about an 8' wide stage (it has 17 points). The dashed slightly wider line is how it expects to correct. The bulging to the left of the 20hz is because I checked the "infrasonics" box on the left, for the subs; I am not sure why it is trying to boost the L/R below 20hz, but I can turn that off (I think).

It thinks it can smooth out the standing wave dips and peaks in the low frequencies a lot, and a little for the higher frequencies (those are more EQ treatments, it only does the active stuff below 150hz by default).

Now does this all actually work? That's what I intend to look at after breakfast. But it's not doing something to cripple the low frequency response of the mains, as best I can tell. Now I may be over-stressing them with the target curve and trying to pull the < 40hz up a bit. I'm guessing I really need to move my target curve on the 2nd and 3rd control point to the right.

Now for the subs, it did seem to override what I set. The measured response above 150hz is still there, and I had the curve making use of it (perhaps a mistake), and it decided to prune the corrected, combined response (heavier blue curve) down sharply after 100hz. But it also thinks is has cully corrected out the big dip around 55hz (somehow). I think on the infrasonics area (<20) it is ignoring the curve but just letting the speaker mostly play through, though the corrected is not quite the measured, so maybe it is doing something. But it didn't try to correct nearly up to the curve I drew. I won't be hearing that part though. What's a bit interesting is it thinks it can't fix the dip at about 22hz.

Everyone: bear in mind, this is my first attempts, and I have a lot of reading to do, so I may be drawing in nonsense target curves and have the levels all wrong. My point is that it does give visibility into what it is trying to do with each speaker.

 

[everettT]

You are a supreme optimist in that case. We did very well without all those auto programs, in my case more then 60 years, and I still don't use them, as they can not improve my systems. In fact I have found their benefit is marginal to none, and easily harmful. They add huge complexity and not simplicity. More likely than not you will "chase your tail."

There is nothing simpler than starting off, how I suggest. I have a very strong hunch with the equipment you have bought it will sound good. Then you can see of you can improve it with those complex programs. There is a high chance you won't. I can't stress enough that knowing the native sound of your equipment before working with Audyssey or Dirac, is not optional, but absolutely essential.

I have just checked, that Dirac ART "takes full control of your mains and sub". It also states optimistically that it "treats them as one." If you believe that, you will believe anything. I have never used Dirac, so I have to own up to that. But to me it seems like an absolutely ghastly program based on a lot of false premises, which I am certain are not correct. They get the old "Click and Clack: - "BOGUS" as far as I am concerned. So they won't get any funds from me.

I'm not a pusher of any room correction but I do believe that dART, after hearing it, does do really good things to a less than optimal room and setup. I don't use it in any of my systems but you should probably hear a demonstration before completely rulling it out for some. I believe Audyssey's correction with the app is good enough for most, especially with the semi-automated 3rd party software to get the most out of their low frequency.

Trinnov's low frequency correction is the best I've ever heard and that was in a perfect room with a perfect setup, or as close to perfect as it gets. The difference with the processing on versus off was very noticeable.

 

[Linwood]

I'm not a pusher of any room correction but I do believe that dART, after hearing it, does do really good things to a less than optimal room and setup.

Look up "less than optimal" in the dictionary and you find it listed as about 10 stops better than my room. The active component of it is what makes it more interesting to me (active to me is correct but a bit misleading, as there's no feedback involved, but the use of generated cancellation wave structures is very different from what other tools do, which are generally just to make the wave patterns better fit the room).

Is it "better"? Don't know yet. But it is different.