Question Can this sub response be EQd? Should I add subs?

SenduB

Standard Member
Joined
Jun 29, 2007
Messages
39
Reaction score
3
Points
5
Age
44
Location
Cambridge
I'm planning a dedicated AV room. 5.3x4.4x2.4m. Probably getting Martin Logan 15A speakers, which are electrostatic panels on top of active subwoofers that go down to 22Hz.

I want to know if I need to budget for and prepare wiring (electrical sockets, speaker cable outlets) for subs before constructing the room.

I've used REW simulator but my main issue is not knowing if my room will be considered "sealed", because that changes everything. The room will be sound-proofed as much as possible and not have windows, but if I put in any kind of small vent for fresh air, will that give me real-world measurements like the unsealed simulations?

Just Front L/R speakers:
UJeAtjNh.png

Without subs could you EQ this flat down to 20Hz by pulling everything above 55dB down to 55dB? And then raise everything back up 20dB?

Is that just normal, no big problem for your standard AVR (I might be getting one with Dirac Live)? Does it mean your power amps then have to do an extra 20dB of output to get back to normal volume? And does that mean you'd need more powerful amps than you might otherwise have thought?

Same, but sealed room:
3MUWHRmh.png

Yikes!

The best wall position for 1 sub was mid-rear wall:
W2obOXUh.png

Also EQable at 55dB.

The only 2 sub position I could find that was as good as or better than the above was 2 rear corners. All asymmetric options introduced a lot of seat-to-seat variation.
9zFJKPYh.png

Still only EQable at 55dB.

Sealed room causes a null at about 25Hz:
74Atagoh.png


I could find no positioning of 4 subs that was as good or better than the above options.

What would you pick amongst these?
 
What are your main speakers, are you using it for movies or music or a mixture of both? And what subs are you looking at (and what's your budget).

Room sim will give you a pretty good idea but what you get in room won't be exactly what you've got there.

Having a vent doesn't make your room not sealed. Even an ajar door would be ok. Non-sealed means having an opening to another room, like a kitchen diner or large internal opening. Do you have the ability to measure your existing set-up? If so I suggest you try it with the door closed and then open and see how much difference it makes.

I don't know how much flexibility you have with placement, but sitting in the middle of the room with the speakers 1/5 from the back and side walls gives a much better response. Putting a sub in the front left corner and crossing over at 80 gives pretty good extension down to 20Hz. Moving further back from your speakers also gives a better response than you have.

Cutting by 20dB in multiple places and then raising the volume by 20dB isn't a great idea unless you have awesome amps. The power required to add 20dB at 20Hz is immense. I don't know the numbers but it's probably double or quadruple power, someone here will be able to confirm.

Also, asking your 22Hz rated speakers to dip down to 20Hz in a room of that size is asking a lot, in my opinion. Don't worry too much about crossing over higher than you think you should. In my room (much smaller than yours) my speakers go down to 30Hz ish, but I cross them over at 100Hz and it sounds good. Let the sub(s) do what they do best and let the speakers breathe and do the higher frequency stuff.
 
My main thought is that I wouldn't buy those speakers. Principally as it's totally unnecessary to have mains speakers that go as low as as 22 Hz if you're going to add subwoofers. At least one ML design like that actually also has relatively poor measured performance, according to comments I've seen from Floyd Toole. Possibly you've heard them and loved them in which case that's obviously relevant, but they still don't seem appropriate for what you have in mind to me.

As for your sealed room question, the model is approximate but I think your room is likely to be closer to being sealed. REW is only ever going to be a starting point though. When running a simulation I'd tick 'Align subs individually', assuming you'll have equipment capable of doing this. Note also that where you site your listening position will affect the response as well.

You should be able to get a better response than you have I think. Just having a very quick play and one sub in the front left corner and a second in the bottom right would give a better response than you have right now (assuming main speakers that don't run full-range).
 
What are your main speakers, are you using it for movies or music or a mixture of both? And what subs are you looking at (and what's your budget).

Per the OP, I'm looking at the main speakers being Martin Logan 15As. Don't know what subs I might use.

Having a vent doesn't make your room not sealed. Even an ajar door would be ok. Non-sealed means having an opening to another room, like a kitchen diner or large internal opening.

Hmmm, do you have references on that? Because an opening to another room just means it is a non-rectangular room, and these simulations assume a 4 walled rectangular room.

Do you have the ability to measure your existing set-up? If so I suggest you try it with the door closed and then open and see how much difference it makes.

The room doesn't exist, and the equipment has not been purchased.

I don't know how much flexibility you have with placement, but sitting in the middle of the room with the speakers 1/5 from the back and side walls gives a much better response. Putting a sub in the front left corner and crossing over at 80 gives pretty good extension down to 20Hz. Moving further back from your speakers also gives a better response than you have.

I've got complete flexibility. The current placement of the speakers is according to ML recommendation. Note that electrostatics are best far from the front wall because of their diploar nature. I'll have a play around though based on your suggestion, thanks.

Cutting by 20dB in multiple places and then raising the volume by 20dB isn't a great idea unless you have awesome amps. The power required to add 20dB at 20Hz is immense. I don't know the numbers but it's probably double or quadruple power, someone here will be able to confirm.

Ah, shame. So what is EQable without having to worry about these sort of issues? What's the max difference between peaks and dips, or is there another way to consider it?
 
Quick thought: as it sounds like you have a significant budget here, have you considered paying a professional to design/spec the home cinema for you? I think it's worth you considering.
 
Apologies, I missed the speakers and the fact that you haven't built the room yet.

I don't have references on the sealed room thing, just personal experience. Multiple rooms with different speakers and subs, opening or closing the door makes limited difference. You're right that having an opening to another room changes the size of the room. Subs don't really see "rooms", they see spaces that they try and pressurise.

I think that 1/5 of the way in to the room is actually further in than you have them, but as you and Ultra say, play about with where the speakers are and then where you sit relative to them. It's a shame they have built in subs. If the speakers benefit from being away from the walls then you won't get the boundary gain that you get by placing a sub against a wall or in a corner. If they have built in subs then they're prob ably genuinely good down to 20Hz and it's a dedicated unit that's doing it. Normally, taking the bass out of the main speakers means they have less work to do which gives you cleaner output in the midrange. By using a sub crossed over at, say, 80Hz, you won't get that benefit, you'll just be wasting the subs in the MLs. Do they do a version without subs?

For EQ, most receivers allow you to cut by 12dB and boost by 3dB. The suggested order for optimal response in a room is usually:
- appropriate speakers
- placement
- room treatments
- EQ

If you're designing the room from scratch then room treatments should be on your list. I had success with Mafia Panels (very helpful, good build quality, cheap) but most people recommend GIK. They'll take your plans and advise you on panels for full bandwidth help. They'll suggest bass traps in the corners and a mix of absorption and diffusion on the walls. They might suggest that you get everything first and then send them your REW file. That way they can look at your decay times as well, something that won't show up in the frequency response and with which standard absorption panels make a huge difference with.
 
My main thought is that I wouldn't buy those speakers. Principally as it's totally unnecessary to have mains speakers that go as low as as 22 Hz if you're going to add subwoofers.

Well, that's one of my questions. Do I need to add subs?

At least one ML design like that actually also has relatively poor measured performance, according to comments I've seen from Floyd Toole. Possibly you've heard them and loved them in which case that's obviously relevant, but they still don't seem appropriate for what you have in mind to me.

I've heard them and compared them to everything else suitable (high-end speakers with a matching centre speaker) I could find in the UK (B&W, Kef, Focal) and they were the best.

They do measure well:
with ARC applied, this is one of the flattest spatially averaged in-room responses I have measured in 27 years of performing this test

What do you find inappropriate about them? Do my graphs suggest they won't work well in my room?

When running a simulation I'd tick 'Align subs individually', assuming you'll have equipment capable of doing this.

My understanding from the REW docs is that you don't want this on because you negate some of the benefit of having subs at different positions; aligned subs won't interfere with each other as much to cancel out peaks or increase nulls. In any case, it happens not to make much difference in the positions I came up with.

You should be able to get a better response than you have I think. Just having a very quick play and one sub in the front left corner and a second in the bottom right would give a better response than you have right now (assuming main speakers that don't run full-range).

Hmm, that's not what I saw, but I left the main speakers as being full range.
 
I don't have references on the sealed room thing, just personal experience. Multiple rooms with different speakers and subs, opening or closing the door makes limited difference. You're right that having an opening to another room changes the size of the room. Subs don't really see "rooms", they see spaces that they try and pressurise.

Did you ever compare your measured results to what the simulation said? With your door closed or open, which simulation was it more similar to? Sealed or unsealed?

Do they do a version without subs?

Not really, no. (I mean, they do an entirely different speaker that is full-range electrostatic, but that is massive and too expensive.) The electrostatic panels cross over to the subs at 300Hz, so they're necessary for basic sound reproduction.

For EQ, most receivers allow you to cut by 12dB and boost by 3dB

Thanks for all the info. I'll have another play with REW when I get the chance.
 
I didn't do that check, but I'll do it when I get home. I need to re-eq my subs anyway. I'll do my mains alone (good to 30Hz), with the door open and closed, and then see what REW sim says. It won't be until next weekend though.

Fair enough about the subs going up to 300Hz. You'll get some benefit to adding dedicated subs then as the subs in the MLs will now be doing ~80-300Hz and the subs you get can do <80Hz, plus you can put the subs in the "right" place and get the boundary benefits for ULF.

I'm guessing, but I suspect you've settled on the speakers you want and you're asking whether you can make them work in the space. I would say yes, you'll get them sounding excellent. If you want to get solid output in the lower registers, bank on adding subs.

If that's true then I know how you feel. I bought my 802s for a space that's far, far too small for them but I love the look and sound of them and I wanted them. Some treatments and a little EQ and they sound great despite their oversized-ness and terrible placement (right in the front corners) and they make me happy every time I'm in that room. And isn't that what this is all about?
 
I didn't do that check, but I'll do it when I get home. I need to re-eq my subs anyway. I'll do my mains alone (good to 30Hz), with the door open and closed, and then see what REW sim says. It won't be until next weekend though.

Thanks, I'd appreciate it.

I'm guessing, but I suspect you've settled on the speakers you want and you're asking whether you can make them work in the space.

Pretty much exactly. I don't just want a flat response, I want a system I love. Both would be ideal!
 
Well, that's one of my questions. Do I need to add subs?
A significant limitation of relying on subwoofers integral to the main speakers is that it dramatically reduces your ability to optimise their position for bass response. This is one key reason for it being advantageous to have seperate subs, and correspondingly no need for the main speakers to go as low as those you are considering.

Another question I'd have, which I don't know the answer to, is if you did only have the two main speakers would it be possible to send data from the LFE channel of movies to these? I have a suspicion it may not be, and if not this is a clear-cut reason to need dedicated subwoofers.

I've heard them and compared them to everything else suitable (high-end speakers with a matching centre speaker) I could find in the UK (B&W, Kef, Focal) and they were the best.
Did you give ATC a try? They'd be on my cost-no-object shortlist.

What setup did you demo the speakers in? Were they used in conjunction with one or more subwoofers?

What do you find inappropriate about them? Do my graphs suggest they won't work well in my room?
Innapropriate in that, as Conrad has said, if as is likely optimal to don't use the output of the main speakers below about 80 Hz then you're paying a lot of money for capability that you won't use.

My understanding from the REW docs is that you don't want this on because you negate some of the benefit of having subs at different positions; aligned subs won't interfere with each other as much to cancel out peaks or increase nulls. In any case, it happens not to make much difference in the positions I came up with.
The manual is talking about a more advanced approach than I believe you're considering, and I'd always start by allowing the subs to be individually aligned. The only reason not to is if you are then going to manually set time delay differences for the two subs to see give the flattest bass response (not just using what you get by unticking the box). Doing this will though mean that the first arrival times from different acoustic sources is different, and so less optimal from a time-delay point of view. This is getting into a much more complicated subject, with more considerations and options than I've mentioned so far but beyond the scope of REW's Room Simulator.


Hmm, that's not what I saw, but I left the main speakers as being full range.
Yes, trying to include the main speakers going as low as you were will make it much more difficult. This is part of the reason for not doing this.
 

Attachments

  • 1572172406765.png
    1572172406765.png
    50.2 KB · Views: 54
sitting in the middle of the room with the speakers 1/5 from the back and side walls gives a much better response. Putting a sub in the front left corner and crossing over at 80 gives pretty good extension down to 20Hz.

What settings did you use? Did you use my room size? Here's my interpretation of your suggestion, which doen't look good:

Y1SCN7Ih.png


The speakers are also closer to the front and side walls than my version (where they are 1.48m from front, 1.12 from side, and my seats are 1.99m from back wall).

After some more fiddling, I found a 4 sub position that looks good if I add lots of absorption to the walls. Is up to 0.5 a realistic figure if adding absorption panels?

ajipfdZh.png


This should be EQable at 65dB, right? Then it would be flat at middle seat, with only that null at 150Hz for the side seats.
 
Is up to 0.5 a realistic figure if adding absorption panels?
At low bass frequencies, no I'm afraid. Corner bass traps are your best bet but even these have a fairly modest impact unless they're BIG.
 
I was only looking at the main position, and I got this

1572198761805.png


Not perfect, but a great starting point.
I wouldn't worry too much more about room sim, the fact that you can get a good response in the sim means you'll most likely be able to achieve a decent response in room. The sim will only ever be partly accurate and as soon as you put furniture in there, curtains up, pictures on the walls etc, the response you get will change.

Some people swear by the multi-sub swarm/kinesis approach (http://www.audiokinesis.com/the-swarm-subwoofer-system-1.html). I don't think you'll need four subs in there.

Also Ultra's right, unless you employ a very specific processor I don't think you'll get the .1 from movies with just the MLs. I had a look at the speaker manual and they don't have any line level input, they're speaker level only meaning that you'll only be sending the speakers the L/R signals. That doesn't contain the .1 LFE channel on most processors. If you tell the processor that you don't have a sub the .1 will be discarded. The same is true for rear or atmos speakers, for example. If you don't have them, the information isn't sent anywhere.

The exception to this is the centre channel. If you don't have a centre then it's signal is split and sent to the L/R speakers (and sub, if you have one and there's content below the crossover point).
 
Quick thought: if trying to model the ML sub output in REW the main speakers should probably be placed on the floor, rather than the ear height that would usually be appropriate.
 
Good point Ultra, probably should, although it doesn't seem to make much difference.

1572208108739.png


However, sitting further back gives a decent house curve with a peak at 32 to tame.

1572208033858.png


I got back a bit earlier than I expected today so I had a little play. Not the best response as I'm in the process of setting up a new PC and re-eqing everything, plus no smoothing. Orange is door closed (sealed room), green is door open. Bear in mind though that's with all subs running and they could quite probably pressurise my whole house, although not down to 9Hz.

1572207872075.png
 
Orange is door closed (sealed room), green is door open. Bear in mind though that's with all subs running and they could quite probably pressurise my whole house, although not down to 9Hz.

View attachment 1212485

What does the REW simulation of your room and speaker positions give you, with the sealed checkbox ticked or unticked?
 
Not too much difference, but a definite drop.

Sealed:

1572209186275.png


Not sealed:

1572209209689.png


Aligning subs individually just brings the responses closer across all seats and correcting the one that's sealed makes zero difference.

My subs are dual opposed so I never know whether to list them as two pairs, or just singles. Either way, singles makes very little difference to the overall response either.

1572209401504.png
 

Hmm, neither your sealed or unsealed simulations seem to bear any resemblance to your measured results. I've seen people mentioning it did a really good job of matching their measured results, but not in your case?

I guess sealed is closer to both your door open and closed measurements since the 20-40 range is not down as much vs your increase. (Do you deliberately have your subs set to boost the low end; is that the reason the simulation is so far from reality?)
 
Hmmm. I thought I turned my house curve off for those, but maybe I didn't. As I mentioned, I did it quickly. Let me try it again when I have a chance.

Either way, you can see that opening a door isn't enough to destroy the response.
 
Either way, you can see that opening a door isn't enough to destroy the response.

The same is true in the sim as well; it's only the sim of just my fronts that had such a massive difference between sealed and unsealed. By comparing your real results to the sim, I was hoping to see if a real room behaves like REWs "sealed" room or not.
 
Ok, I finally got round to playing with this properly.
Here are:
- Sweep of left only, one trace with door open, one with door closed (purple closed, blue open)
- Room sim with sealed checked
- Room sim without sealed checked
Door Open vs Closed.jpg
Screenshot 2019-11-02 at 10.43.10.png
Screenshot 2019-11-02 at 10.45.17.png


So either room sim isn't completely accurate or the opening of a door doesn't make the room sealed/not-sealed. As I think I said to begin with, sealed, to me, means that the room doesn't open . on to another room, like a lounge diner. I don't think you need to worry too much about a door being open. You certainly don't need to worry about air vents.

What's good is that the shape of my response it broadly aligned with the shape or the response.
 
A significant limitation of relying on subwoofers integral to the main speakers is that it dramatically reduces your ability to optimise their position for bass response.

Well put! This should be in a sticky post at the top of this sub-forum.
 
Just as a real room example, see my original REW room sim from when I was planning my room, followed by the actual measured response using REW. Delay for the rear subs was critical as was adjusting the overal sub delay relative to the main speakers (not shown in this measurement). I added room treatment behind my AT screen (30cm deep behind the screen, 60cm deep below the screen) and also some around the top edges of the room behind a non AT pelmet.

NOTE: My front subs aren't quite in the corners as I have them in a false wall firing outwards towards each side wall. I did a lot of measuring with REW before the final build of the false wall to confirm what impact this had compared to turning the subs forward/pushing them right into the corners: It was a negligible difference and it works better cosmetically (and allows a gap for black Devore side curtains to slot into when the projector isn't in use).

My initial sim:

REW 4 sub sim.jpg


A subs only measurement at the MPL. Due to the 4 subs, this response remains very even when moving to other seats. While I tend to sit in the MLP, it means that other viewers don't end up sat in a seat with a bass peak and have to suffer booming.

There is a house curve applied to this, so it's more the smoothness of this response that is of interest (no smoothing applied in REW):
REW 4 sub measurement.jpg


Some pre-eq done by my sub amplifiers (iNuke 6000DSP), then Dirac applied over the top of this, then further minor delay adjustments made for the overall sub delay to smooth the response around crossover region.
 

The latest video from AVForums

TV Buying Guide - Which TV Is Best For You?
Subscribe to our YouTube channel
Back
Top Bottom