Try to EQ Sub plus L/R powered mains?

[NCDave]

Funny, I've read a bunch over here but I guess this is my first post.

I've played with REW and BFD for a while, and am not having much success getting my room EQ to settle down. I think I know why, but don't know how to attack the problem. I did a few searches here and found that it's not a good idea to try to EQ differing subs. Also multiple subs in general are hard in weird-shaped room. I have all three.

I have the "powered bass stands" (PBS, discontinued from AV123), with their ELT-1 bookshelf speakers (also discontinued.) The PBS are essentially 8" sealed, powered subs, in tall boxes that double as stands for the L/R speakers. I knew I had a poor room response (very heavy around 40 Hz) from previous sweeps I'd done with REW. Never tried to EQ it, because the subs would need separate EQ, and right now they get a speaker-level signal. [The PBS get a full-range signal from the receiver--set as "large"--and I use the PBS' crossovers to send the upper range to the ELT-1's. There is an option to send the PBS a separate line-level signal, and there is a line-level out for daisy-chaining.]

I can supply a link to AV123's Retired Products archive with the PBS, if you need...not sure I can do that with 0 posts.

Anyway, this week, I added a sub from the Outlaw sale (LFM-1 Compact). This sub gets me the low end (mid-20's) which was missing with the PBS. I like hearing these, especially when playing low-volume jazz--I can hear the bass line now BUT I hear definite resonances in the room at a number of frequencies. REW has confirmed this.

I've been trying various phase and level settings on all three "subs" (counting the PBS as subs), trying to tame the response, but this is just a confusing mess. I worked out some filters on the BFD and got nowhere near the predicted response, even with the PBS turned off.

So...I need to attack this problem in pieces, I think, and maybe just give up on the PBS(?) I would appreciate your comments on these ideas & questions. I need to step back and develop a plan...

1) Don't try to EQ everything at once (sub plus main/PBS). For now, maybe disconnect the PBS and use just the Outlaw by itself or with the L/R ELT-1 bookshelves. Set the receiver crossover to 80Hz.

-- Should I turn off the L/R speakers for this test, too?
-- Should I use the sub cal signal or the speaker cal signal?

2) If I keep the PBS in the test with the mains, maybe set the fronts to "small" to prevent the PBS from contributing much? This might be like not using the PBS at all.

3) Give up on the PBS/ELT-1 combo and go buy non-powered, near-full-range speakers for L/R. I feel like I need a near-full-range because this is used for 2-ch music as well as HT, and the bookshelves plus sub are a bit lacking in the midrange for 2ch.

I don't know. I'm a bit at a loss here. Not even sure I've outlined the problem clearly.

Thanks,
Dave

[Wayne A. Pflughaupt]

Welcome to the Forum, Dave!

I have good news for you. You are making this way too complicated! I found your speakers at AV123's site. The PBS isn't a subwoofer, and AV123 doesn't even call it a subwoofer. They call it a woofer. So set them to "small." It won't be like they're "going to waste". The ELTs only have 5-1/4" woofs; the PBSs' are needed to fill in the mid bass between the ELTs and the new sub.

So, with only one real sub in the equation, just use REW to find the best placement, running sweeps with the sub only. Once you do that you can add back the mains, adjust the phase, etc.

Regards,
Wayne

[NCDave]

Thank you for the quick reply, Wayne.

Ok, well, I hope it IS as easy as you say...but there are still "sub-like" adjustments on the back of the PBS (crossover, phase, level). And you are right--I do tend to make things way too complicated, but not intentionally.

I thought more about this as I was studying the PBS manual a bit more just now. I may have misunderstood how these PBS work when using the speaker-level inputs. I thought the internal crossover would prevent the ELT-1 from seeing the lower frequencies (HIGH LEVEL OUT from the PBS), but I think now that all it does is prevent the PBS woofer from seeing the higher frequencies present on HIGH LEVEL IN.

In the manual, it says this: (you can skip ahead to *** if you like)

"Other Features:
8” woofer is high passed at 31Hz <<< for protection, I guess
0-180 phase adjustment, gain, and
crossover adjustments (40Hz-150Hz)" <<< highest freq the PBS woofer plays

and for connections:

"If you have a subwoofer...
Take the speaker cable from your amplifier or receiver and plug it into High Level inputs on the back of the PBS amplifier. Next take the supplied jumper cable and go from the High Level output terminals to the binding posts on the back of the speaker."

(Is does not mention high-passing the speaker).

They also say there are two methods of connecting the PBS (and I'm about to ignore it all, but hang on):

"Option one:
Set your receiver or processor to MAIN speakers = LARGE. You may need to adjust the SUB setting in your receiver or processor to say ENHANCED or MAINS + SUB or something to that effect. This ensures your sub will be working in tandem with the woofers in the PBS."

(Option 1 makes it sound like they are high-passing the ELT-1 inside the PBS.)

"Option two:
If you receiver or processor offers very low crossover points to the sub, you can try setting your MAIN speakers as SMALL but with a 30 or 40Hz crossover. Set your SUB to YES. This will give you a lot of low pass in your PBS’s but not the lowest of the low. You may prefer this to option one but it's user dependent. If you tend to play your system at reference levels (very loud!), this is probably your best option."

The lowest my receiver's crossover goes is 80Hz, so Option 2 does not apply.

(***)

So if I understand your suggestion, I set my fronts (plus PBS) in the receiver to SMALL, and set the PBS crossover all the way up (150Hz). Connect the ELT-1 to the speaker level out on the PBS. So the whole ELT+PBS will act like just a wider range speaker (which is what I want), and is not going to get much below 80Hz. The PBS woofer will play from 80 to 150Hz and the ELT-1 above that. Makes sense. And that's actually one of the combinations I tried. I think.

But, I was not trying to EQ the sub alone, which is the key to all of this I think. Ok, then after THAT, I put it all together and adjust the PBS volume to get the midrange where it needs to be. Not sure yet about the PBS phase, but one thing at a time.

If I can get any before/after plots which are not too horrible, I might post them. In either case, I'll report back with my findings.

Thanks again for your help, Wayne.

Dave

P.S. Nice icons on this forum!

[NCDave]

Just a quick note to say I am continually amazed at the terrific info on these forums. I got a new laptop with XP on it and have never gotten a proper calibration on the sound card.

Well, one thread in the EQ section on how to set up the Creative Live! 24 bit External sound card and presto, it works. Problems I hit which were resolved:

1) Problem: Level too low (I've had this issue from the beginning, even with the older laptop).
Solution: You cannot plug-and-play this card. You must install the drivers (and more, read on). Suddenly the level is fine.

2) Problem: feedback, giving the classic swept-sine calibration curve (wavy line vs. straight line).
Solution: The sound card monitor is ON by default. Install the Creative software (not just the drivers!) to get access to all the controls, and read and follow Post #2 in this thread:

http://www.hometheatershack.com/foru...nal-setup.html

3) Problem: CRSS will not turn off.
Solution: Same solution as #2 above.

And now I have a nice cal curve with great levels.

Hopefully will get a chance to actually try the EQ suggestions tonight.

Thanks again,
Dave

[Wayne A. Pflughaupt]

Thanks for the info, Dave.

NCDave wrote:

"Other Features:
8” woofer is high passed at 31Hz <<< for protection, I guess
0-180 phase adjustment, gain, and
crossover adjustments (40Hz-150Hz)" <<< highest freq the PBS woofer plays

and for connections:

"If you have a subwoofer...
Take the speaker cable from your amplifier or receiver and plug it into High Level inputs on the back of the PBS amplifier. Next take the supplied jumper cable and go from the High Level output terminals to the binding posts on the back of the speaker."

(Is does not mention high-passing the speaker).

Hopefully it does that.

So if I understand your suggestion… the whole ELT+PBS will act like just a wider range speaker (which is what I want), and is not going to get much below 80Hz. The PBS woofer will play from 80 to 150Hz and the ELT-1 above that.

That's correct.

But, I was not trying to EQ the sub alone, which is the key to all of this I think.

Not sure what you mean there. You wanted to EQ the PBS' along with the subwoofer? Sorry, can't be done, unless there are line level inputs for the PBS. And it would require a separate EQ.

Ok, then after THAT, I put it all together and adjust the PBS volume to get the midrange where it needs to be.

Yup.

Not sure yet about the PBS phase, but one thing at a time.

Ditto. Not sure any phase adjustment would be needed, but, the fact that it's there might be is an indicator that the PBS is probably high passing the ELT. Using REW's RTA mode you could see the effects of the phase adjustments in real time.

Regards,
Wayne

[NCDave]

Wasn't sure I could upload pictures as a new member, so I padded a couple posts.

I think I am getting pretty good at the sub-cal-and-EQ procedure. I just can't believe how sweet that REW software is, once you get to know it. And the Midi support for the BFD is terrific.

So, I understand the sub-only calibration. But I cannot figure out how to properly integrate the levels of the main speakers (and PBS) with the sub. I've turned up the volume on the PBS (left/right powered woofers) all the way to maximum. This boosts the 100-200Hz range a bit. Not much. Above that to 1kHz the level is a LOT lower than the sub.

The sub seems to sound about the right level, at least with music. It is not at all over-powering and actually sounds a bit lower than when I first set it up. The speakers are louder than what I'm used to at the AVR level I usually use because the PBS volume is all the way up. (So I turn down the AVR, no biggie.)

So what's the deal with the graphs which run a speaker cal sweep, showing the bass as so much higher than the midrange? (And I didn't bother with anything above 1kHz.)

Here are the pictures:

1) Sub-only, relative to target level, without any EQ. No smoothing.

2) Sub-only, after adding 4 automatic filters (awesome feature), plus 4 more boosts, staying away from the nulls. Bummer those nulls...

3) Sub plus L/R speakers, including the PBS. 1/12 octave smoothing. The PBS are turned up all the way, and I performed a new, speaker-type calibration before the sweep.

Flipping the phase on the PBS makes (3) worse.

Any ideas what this means? Is any further action needed?

Thanks,
Dave

[brucek]

Any ideas what this means? Is any further action needed?

Well, I feel the level of the entire area from ~20-200Hz is too high. At the very least the sharp jump at 200hz should not occur. I would lower the level of the mid bass pair.

brucek

[NCDave]

I actually had it lowered before the last measurement, bit since I'm running the AVR crossover the to mains at 80Hz (mains as "small") the PBS only seem to contribute in the 100-200Hz range. I ran several sweeps, just running up the volume on the PBS to verify this. I would like the PBS to play higher than the max crossover setting in the PBS.

The bass just seems too hot, according to the graphs...or the midrange is too low.

I thought of a problem with using the PBS in the upper bass area above 80Hz: bass becomes directional there. The PBS have the woofer on the SIDE, so the woofers actually do not point toward the listening position. It's worse when the mains are toed in. So the meter could be reading low.

I'm going to pull the PBS out and see what happens with just the sub and bookshelves.

There is clearly an issue with level (sub vs. mains), but I don't know how to fix it when using REW according to the instructions. I guess I could just choose a lower target SPL level than 75dB, assuming I have a high enough level for measurement.

I have a few more things to try.

[brucek]

The bass just seems too hot

Why can't you simply turn it down?

I'm going to pull the PBS out

Why not just turn them down.

I guess I could just choose a lower target SPL level than 75dB

Nope - you never do that. Simply turn the bass down.

brucek

[Wayne A. Pflughaupt]

NCDave wrote:

1) Sub-only, relative to target level, without any EQ. No smoothing.

2) Sub-only, after adding 4 automatic filters (awesome feature), plus 4 more boosts, staying away from the nulls. Bummer those nulls...

Your sub level was way too high, at least 18 dB above the Target Curve, which I’m confident resulted in some pretty drastic equalizing. The Target should have been reset via the “Set Target Level” function or by lowering the subwoofer level before running REW’s auto EQ function. I imagine those boost filters you added were only needed to counteract some of the effects of the drastic cuts REW recommended.

Do you have any other location options for the sub that might eliminate those nulls?

The sub seems to sound about the right level, at least with music. It is not at all over-powering and actually sounds a bit lower than when I first set it up. The speakers are louder than what I'm used to at the AVR level I usually use because the PBS volume is all the way up. (So I turn down the AVR, no biggie.)

So what's the deal with the graphs which run a speaker cal sweep, showing the bass as so much higher than the midrange? (And I didn't bother with anything above 1kHz.)

The bass just seems too hot, according to the graphs...or the midrange is too low.

I agree, the graph does appear to be low end heavy, but you say it sounds about right. Only you can be the judge of that. But you might try to get things adjusted so that they look more correct and listen to it that way for a week or two. Sometimes people aren’t used to the way a properly calibrated system sounds, if they’ve never had one...

Regards,
Wayne

[NCDave]

Wayne A. Pflughaupt wrote:

Your sub level was way too high, at least 18 dB above the Target Curve, which I’m confident resulted in some pretty drastic equalizing. The Target should have been reset via the “Set Target Level” function or by lowering the subwoofer level before running REW’s auto EQ function. I imagine those boost filters you added were only needed to counteract some of the effects of the drastic cuts REW recommended.

Do you have any other location options for the sub that might eliminate those nulls?

I agree, the graph does appear to be low end heavy, but you say it sounds about right. Only you can be the judge of that. But you might try to get things adjusted so that they look more correct and listen to it that way for a week or two. Sometimes people aren’t used to the way a properly calibrated system sounds, if they’ve never had one...

Regards,
Wayne

Thank you Wayne and brucek.

I'm a bit discouraged, actually. Did a lot more messing with this last night and did not get anything like what I wanted.

Yes, the bass is too high. Yes, I can lower it. But I bought a sub, which contributes to the low end for HT, which is nice even though I have no "HT"--I have a CRT TV on which I can watch movies...but at least I get to hear some cool low stuff.

And then there's the low-volume listening when the kids are in bed. The new sub allows me to hear the bass better.

So I turn it down and both of those features go away.

Yesterday, I tried running without the PBS (just the sub and the ELT-1's). With a proper 80-100Hz filter on the AVR, the sub stops its contribution but the ELT's don't pick it up and there is a hole in the upper bass. So I turned up the AVR crossover to 200Hz, and got it closer to flat. But you know what happened, of course (and I learned): then I heard the bass coming from over in the left corner-yuck.

I have three other places the sub could go. None are in corners--or at least no corners that the SPL meter "thinks" are corners, according to the level. One spot is right in front. The other two are off to the right and that side is pretty much all open. In any of the three other locations, I lose a lot of the sub output and extension.

After the fiasco last night, I just said, [forget it]. I have the fronts (with PBS) as small. AVR crossover at 80Hz. I turned the PBS crossovers up as high as they could go. I turned the PBS volume to 1/3. I set the sub crossover at max (which might be only 90Hz--there is conflicting info in the manual and on the sub) I set the sub volume at 1/3. It sounds ok to me. It probably looks like garbage on REW.

The test tones (pink noise) from the AVR--if I run them to get 75dB on my meter--indicate that the sub is way too loud. If I turn down the sub volume or the AVR level going to the sub to "equalize" the SPL using the meter, then I can't hear the bass.

It's a conundrum.

[NCDave]

I'm going to try one of the other sub locations anyway. Who knows what will happen? And maybe the big null goes away, but I think I calculated that is coming from one of the room dimensions.

[NCDave]

I could not stay away.

I didn't move the sub yet, but decided to try a few things.

1) Set the knobs to sound "pretty good to me."

2) Use REW full sweeps to see the "big picture" [full sweeps only to 1kHz]

3) Turn down the sub and PBS volumes. See if any phase adjustments help (no, they didn't).

4) Sub only now. EQ that.

5) Add back the mains. See if volume adjustments help (not really). See if phase adjustments help (no--usually made things a lot worse.)

6) Be happy. Some things in Life cannot be fixed, like the low 200-350Hz section.

Here it is in pictures:

a) After Step 3 (before any EQ).

b) Sub only, no level adjustments, before EQ.

c) After Step 4 (after sub EQ). Took four cuts and two boosts.

d) After Step 5.

NOW I will leave it a few days or so to get used to the lower bass and see how it goes.

Thank you for your help, guys.

[NCDave]

I forgot to say the 1st and 4th plots have 1/12 octave smoothing, but the low nulls are not much different. It is interesting that the big bad null got better by adding the PBS.

[Wayne A. Pflughaupt]

It might be the smoothing making it look better...

Regards,
Wayne

[NCDave]

Wayne A. Pflughaupt wrote:

It might be the smoothing making it look better...

Bubble-Burster!

I smoothed those two to make them easier to look at. (Smoothing is to the eyes like Advil is to a headache. )

Seriously, the nulls were not appreciably affected by the smoothing. There is improvement in the graphs with the BFD, but not a lot, and to be honest, I used the IN/OUT control on the BFD and don't hear a difference. It would be nice if the BFD had a remote control so I could do this from the seating position, though.

[NCDave]

I thought this was very interesting. In Post #13 of this thread (re-posted below, 1st graph), you can see a bump in the response at 122Hz. I was not able to figure out why that bump was there. Well I found it.

Since that post, I decided that I like the bass boosted a bit from flat. This has to do with listening to jazz at low-volumes; I like to hear the bass a bit better.

So, I boosted the bass to where I like it (by ear), then re-EQ'd the sub...then thought, "Oh, let's just see what each speaker set is contributing to the overall response at a fixed target level."

Here's what I had from Post #13. Notice the 122Hz bump:


Here's what I had after boosting the bass a bit for my preference:


Ok, now the separate speaker sections. Here is the sub alone, but EQ'd with REW and BFD:


Here is the pair of Powered Bass Stands (PBS), alone. One for each front speaker. And I simply cannot figure out how my XR55 is sending that spectrum to the PBS, since the fronts are set to SMALL and filter at 80Hz. It looks like there's no rolloff at all!


Finally, the ELT-1 speakers alone. I have a really crummy room with openings everywhere, so I guess I can't expect a lot. This is 1/12 octave smoothed:


Here are all three now, superimposed. Note that the XR55 volume and soundcard volumes were not adjusted after the sub was EQ'd:


And then, here's the final response with all three speaker types in operation (sub+PBS+ELT). There's that 122Hz bump (1/12 smoothing), but the superimposed plots above kind of told me what speakers were the problem at 122Hz:


Then an idea hit me. I had already tried messing with volumes and phases on the sub and the PBS, in order to deal with that 122Hz bump. Nothing worked. But I had never tried flipping the leads into the ELTs! Here's what I get with the leads into the ELT's reversed. The ELT's are fed through the PBS' crossover (1/12 smoothing):


122Hz bump is gone. Of course now it's a little LOW in that part of the response, but it's getting closer to flat. For comparison, here are the superimposed plots of all speakers with and without the ELT leads reversed (1/12 smoothing):


If you read this far, congratulations! Anyway, I thought this was very interesting. I had never considered flipping the leads from the PBS to the ELT, because these two speakers are meant to work together as one. But it appears that they work better together with flipped ELT phase.

[Wayne A. Pflughaupt]

Interesting. Is the crossover point between the PBS and ELT anywhere in the 122 Hz range? That would probably account for it.

Regards,
Wayne

[NCDave]

One more post with the room layout. I'm trying to figure out if the lower level in the 200-300Hz range is because of my small front ELT speakers (even though they have the PBS, the PBS' highest crossover setting is 150Hz, so it does not contribute to 300Hz), or if the problem is just the room. I've been looking at the Rocket 450's from AV123, but if I put them in this room and they sound the same, I would be really upset. I talked to Jess over at AV123, and he thought the hole in the frequency response was because of the room, but he didn't have the benefit of seeing the room layout.

So here it is... You can see the "front wall" is really a 5' left wall, a 5' right wall, and 8' wide open o the front foyer. The entire right side is open to the kitchen/eating area. The left and back sides are closed.

[NCDave]

Wayne A. Pflughaupt wrote:

Interesting. Is the crossover point between the PBS and ELT anywhere in the 122 Hz range? That would probably account for it.

That was a fast reply, thanks!

The answer is yes, but I can't figure out how to fix that. The PBS' highest crossover is 150Hz (which is where I have it). I would like them to play up a but higher, but that's not allowed, it seems.

Still, the ELT's are supposed to be +/-3dB down to mid-60's or so. That's why I'm thinking something about this room is just sucking down that part of the response...

[NCDave]

Here are the individual speaker responses, again, superimposed, with annotations. It's pretty clear that the ELT's response in this room causes the 122Hz bump. But it's not clear to me why the range from 200-300Hz is so low in the final response, unless it's just because of the weird room.

[NCDave]

And if the XR55 is setting the fronts to SMALL, and crossing them over at 80Hz, how in the world are the PBS' still playing at the level they are, so far down in the response? Is that another room anomaly?

Here are the connections:


XR55----(speaker level)---->PBS w/crossover----(speaker level)---->ELT
|
|----(line level)---->LFM1c Subwoofer

[Wayne A. Pflughaupt]

NCDave wrote:

And if the XR55 is setting the fronts to SMALL, and crossing them over at 80Hz, how in the world are the PBS' still playing at the level they are, so far down in the response? Is that another room anomaly?

You might want to consult your manual or do some further testing. We’ve had people here who had this problem and eventually discovered that their receiver’s bass management only worked for digital sources and signals.

That said, you might also want to take some readings of your L/R separately. I expect that the left speaker has much more pronounced bass than the right one.

Still, the ELT's are supposed to be +/-3dB down to mid-60's or so. That's why I'm thinking something about this room is just sucking down that part of the response...

That’s fairly optimistic extension for a speaker with such small woofers, unless they’re unusually inefficient. And yes, the room makes a difference. Small rooms reinforce bass; large rooms “soak” it up. Keep in mind that the speakers “see” all those areas open to your family room.

Regards,
Wayne

[NCDave]

Wayne A. Pflughaupt wrote:

You might want to consult your manual or do some further testing. We’ve had people here who had this problem and eventually discovered that their receiver’s bass management only worked for digital sources and signals.

Well, I sure hope not. The Panny's manual is pretty bad, and not very specific about the crossover, but I can test that theory with sweeps using different crossovers.

you might also want to take some readings of your L/R separately. I expect that the left speaker has much more pronounced bass than the right one.

Good idea, and you are probably right about that, since the left speaker has two-wall reinforcement. And while I'm at it, I guess I should look at each PBS separately. This could get complicated.

Thanks Wayne.

[NCDave]

I did about 20 more measurements. I'll post a few here. Bottom line is that I have a lot of room gain on the left side, plus I have a nice room mode in the region the PBS are operating.

First question was: does the XR55 apply the crossover to analog inputs? The answer is yes. It's 12dB/octave. I verified this using the ELTs only.

Second question was: are the left and right ELTs (and PBS) significantly different because of room gain? The answer again, is yes.

The first plot is the left and right ELTs, swept individually, and then plotted together, with the 80Hz crossover in the XR55. The blue line is the left ELT. Sure enough, there is the 122Hz mongo peak:


The next plot is the left and right PBS, set at equal volume, swept individually, and plotted together. The purple line is the left PBS. Needs to be turned down, but there is a ton of room gain around 40-50Hz. Remember, they are supposed to be dropping 12dB/oct at 80Hz.


The next plot is the same as the previous one, except I have turned down the volume one notch on the left PBS:


That's better, but the overall response (all speakers plus sub) still has the problems with the left ELT's room mode. I tried most combinations of PBS phase and ELT phase (flipped leads). Most of these obliterated the overall response. The plot below gives two configurations: with and without leads to both ELTs flipped. Blue one is with leads flipped:


I still have the issue in the 200-300Hz range. It's apparently being caused by the left ELT, and I just might not be able to do anything about that. The room's options are limited.