Suggestions for small listening room

Conrad · « **on:** 16 Sep 2018, 07:42 pm »

I'm interested in people's opinions on my frequency response charts and suggestions for what they would do if it were there results.
I should start by describing my set up.

The room is an upstairs room converted to a "cinema", so suspended floor over joists, garage below. Front wall has a window which is covered by a 1/4" plywood board. The room is carpeted with deep pile carpet and is 8' wide by 12' long. The front wall has a retractable screen handing down about 6' from the wall. Front and side walls, and ceiling are covered with devore fabric to about 6' out. It's not super tight to the wall. Door is in the back left corner. Walls are drywall, but the front and right walls are external so cavity with brick outside.

The system is a pair of B&W 802 Nautilus, not spiked, each about 6" (yes, inches) from the side wall and 8" from the front wall. In the middle is a Velodyne DD15 subwoofer crossed over at 60Hz. Between the speakers and the sub on one side is a PC which is my source (Windows 10, RME HDSP 9632 Sound Card, JRiver) and on the other is a Classe CT5300 amp. Both the PC and the amp are on brass feet and 100lb concrete slabs. The brass feet are mostly to allow some air to circulate under the amp, they're under the PC so they match visually.

Center speaker is a B&W HTM3 Nautilus sat on the sub. The only other things in the room are a two seater sofa about 2' from the rear wall and a small table in front of that. Distance from LR speakers to LP is about 8'.

All that said, here are my frequency responses and a waterfall.

That's a raw sweep from REW at the listening position and then, based on the three filters it suggested, a treated sweep. It's tamed some peaks at 450Hz and 6kHz but didn't do much at the 47Hz (I'll probably remove that filter).

That's the waterfall with the filters. Unfiltered looks largely the same but with the peaks as per the SPL charts.

So, what do you think? I'm intrigued by the idea of some treatments but, if I'm reading things right, I only really need treatment to tame that 45Hz peak. I'm guessing that and the 100Hz trough are room modes and can't be affected with EQ.

One other thing, I've tried EQing the speakers nearfield which REW gave me more filters for, and I've tried Dirac, but I didn't like them. Dropping that treble out at 6kHz seems to make things veiled and muddy. These latest filters are certainly the best I've heard, so I'll see how that goes. Maybe I just like a bright sound and a lively room, if I acoustically treat, I might ruin that.

Thoughts, comments and opinions welcomed (I'm looking forward to them!)

Thanks, sorry for the long, rambly first post.

Hipper · « **Reply #1 on:** 17 Sep 2018, 04:42 pm »

There are three ways to change the frequency response (FR): positioning, of both speakers and listening chair; room treatment; and EQ/DSP. I use all three.

Have you tried positioning? Is it practical for you? It can make a big difference. I've no experience of subs and their positioning but I understand moving them about may change things (it could help with the 45Hz problem). You could use REW to find the best locations. Concentrate on the 20-300Hz range. I use 'The Thirds' as described here:

http://www.barrydiamentaudio.com/monitoring.htm

Room treatment is mainly about sorting the issues in this region too. Bass traps in the corners are the usual solution, the more the better. Unfortunately to be effective at low frequencies they need to be big, say 40cm square. A couple of sources with lots of info on room acoustics:

https://www.gikacoustics.com/

http://realtraps.com/

For DSP/EQ, I assume you have got these filters you mention from REW's EQ pages?

What you could do is measure for the range 20-300Hz only, using four sweeps to get an average, and measuring both speakers at the same time, microphone being positioned pointing down the centre of the room from where you ears would be.

Apply those filters in JRivers DSP Studio. Re-measure. You should get a pretty flat response for that range.

Listen! You may like what you hear without further work.

For the 100Hz issue you could add some EQ to see what happens. Maybe nothing but maybe it will change things. I shouldn't think adding 8dB will do any harm.

For the apparent ringing at 45Hz, first check the REW Spectogram - it's accessible on the tabs top right. If the 45Hz signal is shown in the negative time area it means the sound existed before the REW signal started. This means it is not caused by ringing but from an external noise, like traffic.

Beyond 300Hz you may not need to do anything. You may like wall reflections although I now don't so I employ treatment to stop them.

I would suggest not having the table between you and the speakers. If your chair is leather it might be best to cover it with fabric. Both these can cause reflections.

There are also possible ceiling reflections. That depends on how your speakers fan out the sound.

Conrad · « **Reply #2 on:** 17 Sep 2018, 06:36 pm »

Thanks for the considered response. It's great to have confirmed some of my thinking, your post has also corrected me in some areas.
Let me share a photo with you to show you what I'm up against, this it (clearly) my front LCR speakers and sub.

This is taken from just in front of my listening position. Not a lot of room for manoeuvre in terms of re-positioning. I could move the sub left of right but looking at a room mode calculator, my room length gives a mode at 47.5Hz, which is exactly where I'm seeing the spike.

The article you linked to was interesting and I might be able to move the couch forward slightly so that I'm just outside that equilateral triangle it talks about, I'm slightly outside that the moment (6 feet from speaker to speaker, 8 feet from speaker to listening position). I can move forward a foot so it's 6 feet from speaker to speaker and 7 feet from speaker to listening position.

The couch is already fabric. I was considering changing to an Eames lounge chair but I might shelve that idea now. I can look at my options for removing the table as well.

I'll try and add stronger filters at 45Hz, but I always thought that room modes can't be tamed with EQ and need re-positioning or absorption (which you're also suggesting). I'll look at some absorption calculators to determine what thickness I need to 45Hz. If I move the couch forward I should have space for a corner bass trap, although only on one corner (rear left) as there's a door on the opposite corner.

Would a loose absorber work, so I could close the door and put the absorber in place, then move it again when I open the door?
Also, I'm assuming that I need absorption and not diffusion?

I don't think that the 45Hz is caused by an external noise, but I'll check, it's a good call.

Are you seeing wall reflections on the waterfall over 300Hz then? I thought it looked pretty flat and even? Ringing only seems to happen until 120-180ms. That's pretty good, isn't it?
That said, if I clap in the room I do hear that fast flutter echo. I wonder if, because of the size of the speakers and sub (probably too big for the room but such a good price I couldn't pass them up) I'm actually nearfield listening and not hearing wall reflections as much as I would in a larger room? Pure speculation.

Thanks again, I have a few things to try out.

FullRangeMan · « **Reply #3 on:** 17 Sep 2018, 06:39 pm »

Maybe this help:
https://www.audiocircle.com/index.php?topic=107352.msg1183456#msg1183456

Conrad · « **Reply #4 on:** 17 Sep 2018, 06:45 pm »

Thanks, that's useful, I'll take a read.

mresseguie · « **Reply #5 on:** 17 Sep 2018, 06:49 pm »

Hi, Conrad.

There seems to be a gravity issue in your listening room. I've managed to correct the issue for you.

Michael

Edit: I thought for sure your original picture was upside down, but now I see it isn't.

Conrad · « **Reply #6 on:** 17 Sep 2018, 06:50 pm »

I saw that and took another photo. Do you think that the room rotating like that is affecting my listening pleasure?

richidoo · « **Reply #7 on:** 17 Sep 2018, 07:56 pm »

Issues I see:
Ringing (slow bass decay) centered at 45Hz. It is over 1/2 second. As you suspected, this is the room. Use tuned traps or thick dense absorption like OC705 4" several layers on rear wall, etc to damp this resonance. Put kraft paper or thick plastic drop cloth material or even thin masonite board over the absorption to prevent it killing the HF. Use your mic to find the spots with the lowest amplitude response at 45Hz and put your traps there. Low amplitude = high velocity, and absorption works by reducing air velocity.

Diffusion can't work that low. The only other option I know is active cancellation, where you put bass driver at the back of the room and delay the launch so it collides with front wave at a position behind you and they cancel each other. You sit in front of the cancellation zone and only hear the front wave. Your room is 25 feet long it could work for you. AC member Scotty is an expert on this. Search for his posts about bass treatment, etc.

That steep rolloff above 6kHz. This is something system-related, like a DSP filter, or rolled off measurement test signal. Hopefully nothing's broken.

The huge peak at 450 and huge dip at 3kHz will both seriously mess up the tone of acoustic instruments, voices, etc. But fix the 6kHz rolloff first, it might cure these also. 450Hz is only 30 inch wavelength so it's not room mode. Maybe some other reflection?

Your speakers are capable of far higher performance than seen in your measurement. Above 300Hz you should be looking a lot better than you are.

Peaks at 60Hz would be suspicious because appliances make 60Hz noise, but at 45Hz it's all room, especially because the offending wavelength matches a room dimension.

You don't have to calculate what thickness trap for 45Hz. You will need far more than you have room for. Just do what you can, read how bass traps work so you can optimize them to your specific concern. Frequency tuned traps might work more efficiently than broadband since you have a specific freq to damp. Place absorption at locations with low amplitude at the offending freq because low ampitude = high air velocity, absorption works by slowing velocity.

Maybe EQ can reduce room resonance?
good luck!
Rich

Conrad · « **Reply #8 on:** 17 Sep 2018, 08:03 pm »

Quote from: mresseguie on 17 Sep 2018, 06:49 pm

Edit: I thought for sure your original picture was upside down, but now I see it isn't.

It was. After about 10 minutes of editing I was able to get it to show up the right way around.

Conrad · « **Reply #9 on:** 17 Sep 2018, 08:08 pm »

Quote from: richidoo on 17 Sep 2018, 07:56 pm

Issues I see:
Ringing (slow bass decay) centered at 45Hz. It is over 1/2 second. As you suspected, this is the room. Use tuned traps or thick dense absorption like OC705 4" several layers on rear wall, etc to damp this resonance. Put kraft paper or thick plastic drop cloth material or even thin masonite board over the absorption to prevent it killing the HF. Use your mic to find the spots with the lowest amplitude response at 45Hz and put your traps there. Low amplitude = high velocity, and absorption works by reducing air velocity.

Diffusion can't work that low. The only other option I know is active cancellation, where you put bass driver at the back of the room and delay the launch so it collides with front wave at a position behind you and they cancel each other. You sit in front of the cancellation zone and only hear the front wave. Your room is 25 feet long it could work for you. AC member Scotty is an expert on this. Search for his posts about bass treatment, etc.

That steep rolloff above 6kHz. This is something system-related, like a DSP filter, or rolled off measurement test signal. Hopefully nothing's broken.

The huge peak at 450 and huge dip at 3kHz will both seriously mess up the tone of acoustic instruments, voices, etc. But fix the 6kHz rolloff first, it might cure these also. 450Hz is only 30 inch wavelength so it's not room mode. Maybe some other reflection?

Your speakers are capable of far higher performance than seen in your measurement. Above 300Hz you should be looking a lot better than you are.

Peaks at 60Hz would be suspicious because appliances make 60Hz noise, but at 45Hz it's all room, especially because the offending wavelength matches a room dimension.

You don't have to calculate what thickness trap for 45Hz. You will need far more than you have room for. Just do what you can, read how bass traps work so you can optimize them to your specific concern. Frequency tuned traps might work more efficiently than broadband since you have a specific freq to damp. Place absorption at locations with low amplitude at the offending freq because low ampitude = high air velocity, absorption works by slowing velocity.

Maybe EQ can reduce room resonance?
good luck!
Rich

All very interesting. Unfortunately my room isn't 25' long, it's 12' long, not sure if that was a typo?
Either way, I'll investigate the bass trappings, I'm currently looking at Helmholtz-Resonators (not something I knew existed). They seem cheap and easy to build and fairly easy to hide. Even if they don't work, it's not much effort or expense to find out. I have to wonder though, if they're so effective, cheap, and easy, why isn't everyone using them.

I'm going to re-measure the room. I'm now wondering if the >3kHz is due to the temporary positioning of the monitor on the table to do the sweeps blocking the signal. I'll either reposition or use the projector next time.

I'll also give the taming of the 45Hz and 450Hz peaks, and a big boost at 1kHz to see if I can flatten things out.

That combined with an adjustment of the listening position might be enough.

Thanks for the advice, looks like I've got a day's measuring to do!

FullRangeMan · « **Reply #10 on:** 17 Sep 2018, 10:35 pm »

Quote from: richidoo on 17 Sep 2018, 07:56 pm

Ringing (slow bass decay) centered at 45Hz.

Rich I think post-ringing is not a prob in bass as it mix with the harmonics, pre-ring is more noticed in the mid-range and treble.

dB Cooper · « **Reply #11 on:** 18 Sep 2018, 05:50 pm »

Ever hear the old joke about someone "having a great face for radio"? I think what we have here is the perfect room- for headphones. Not so much for speakers IMHO,,, at least not the ones that are in it. A pair of mini monitors and maybe a sub (not too big) would probably work out better. At least then you could get the mains speakers out of the corners, which is probably causing many of the problems you are struggling with. Then the sub is piling on. Too much bass in too small a space. Maybe looking at speakers designed for corner placement also, but there really aren't many I know of.

Conrad · « **Reply #12 on:** 20 Sep 2018, 11:15 am »

Ha, thanks for that.

As you can see, it's a home cinema so I'm not sure that headphones will work, but I understand your point. The speakers and sub are waaaay overkill for that room. Like I said, I got them for a good price. I also don't need a 5x300WPC amp, but that was a bargain too. I'm buying with the future in mind. Eventually my wife will either come around, or leave me, in which case I'll get a bigger room

That said, I'm not sure that a 45Hz room mode is too bad of a problem to deal with.

I did some additional research. It seems that my SPL meter (RadioShack) isn't useful above 10kHz, which is why we're seeing the roll off. I did let REW generate filters to get it flat with +-1db all the way to 20khz. It sounded absolutely awful, so I know that the treble is working. Without an accurate mic though the whole 1kHz+ process is for nought, so I'll stop there.

This also explains why Dirac sounded bad to me. I'm actually really pleased about that as it means I can give it another go later. I'll get myself a MiniDSP USB mic and retry.

Once I realised that that I focussed on the 10-200Hz range and had some successes dropping that 45Hz peak and filling in that 100Hz trough (which is a floor reflection, apparently). I also tried it with and without the table, with the listening position further forward and nothing really makes much difference. I think those changes are so small compared to the speakers and room as to be not noticeable.

I'll post some more screenshots of the bottom end when I have them.

Thanks for the help so far.

guest61169 · « **Reply #13 on:** 20 Sep 2018, 03:18 pm »

Have you tried running the speakers full range while adjusting the sub so it starts rolling off later (ie. at 90Hz)?

Conrad · « **Reply #14 on:** 20 Sep 2018, 03:34 pm »

To widen the overlap and try and get rid of that 100hz dip, you mean?
No, I haven’t. I can do that in jriver, it allows me three crossover options: remove the bass altogether, move the bass to the sub, or copy the bass to the sub. I’ll give “copy” a go and adjust the crossover up from 60.

Thanks, that’s an interesting idea. Where there’s additional bass I can deal with that in jriver or in the sub itself.

mick wolfe · « **Reply #15 on:** 20 Sep 2018, 04:23 pm »

Quote from: dB Cooper on 18 Sep 2018, 05:50 pm

Ever hear the old joke about someone "having a great face for radio"? I think what we have here is the perfect room- for headphones. Not so much for speakers IMHO,,, at least not the ones that are in it. A pair of mini monitors and maybe a sub (not too big) would probably work out better. At least then you could get the mains speakers out of the corners, which is probably causing many of the problems you are struggling with. Then the sub is piling on. Too much bass in too small a space. Maybe looking at speakers designed for corner placement also, but there really aren't many I know of.

Exactly. Wrong speaker for that space.....especially when tucked in the corners as shown. A smaller Audio Note monitor may work as they're designed for corner placement.

richidoo · « **Reply #16 on:** 20 Sep 2018, 05:30 pm »

Test mic accuracy is very important to getting good room treatment results. With such a big error in the HF, I wouldn't trust the LF measurements either, at least not enough to build tuned treatments, which target a specific frequency. I was recently recommended this mic.

I assumed the 25 foot length from your statement, "my room length gives a mode at 47.5Hz." A mode at 47.5Hz requires at least one room dimension of 23.8 feet, which is the wavelength of that frequency.

If your room length is 12 feet then your fundamental (lowest) mode for that dimension is at 94Hz. Sounds cannot bounce, ring, resonate, etc between surfaces that are spaced closer than the wavelength of the sound. (EDIT: but their harmonics can!) There are no modes below a sound's wavelength. The modal region starts at the wavelength corresponding to the room's largest dimension. So the 47Hz peak is caused by something else, not a room mode. I would ignore it until you get a better test mic.

You can use modecalc program from realtraps.com to learn what room modes and harmonics you can expect to see based on your room dimensions. With an accurate test mic, you should be able to correlate your test data with the room dimensions to know what actions are necessary to change the room response. If you remove the smoothing from your bass sweep graphs you'll be able see more detail.

Your main speakers have more than enough SPL and bass extension for your room, so you really don't need a sub. You'd do better plugging the reflex ports of the B&Ws to eliminate the phase error caused by the ports, then EQ them for flat in-room response. With the ports plugged, you can EQ below the port tuning freq, if needed.

Conrad · « **Reply #17 on:** 20 Sep 2018, 06:05 pm »

It was the minidsp umik 1 I was looking at already, albeit not a calibrated one. I'll see what my options are in the UK.
I already have a behringer ECM8000 which came with the velodyne, but it's old and I believe that mics can "go off" with age. Plus, to run it from the PC I'd need phantom power which would mean getting a mic pre amp. If I'm going to spend money I'd rather get a decent mic than a pre-amp for a mic that may or may not be good.

Apologies, for the room mode I used the first calculator I came across: https://amcoustics.com/tools/amroc?l=12&w=8&h=7&ft=true&r60=0.6
That places the first mode at 45Hz for my room, but I see now that it looks like it might be for another purpose.
I'll check again with the one you suggested. However, that's actually really good news as it means that, if that spike at 45Hz or thereabouts exists, it can be tamed with EQ. But I need to see the actual landscape before I can do anything about it.

As for not needing a sub, as mentioned, it's a home theater as well as a listening room. I can switch the sub off for 2 channel, or just have JRiver play it back as the source number of channels which takes the crossover out of the equation. The sub does give an extra 5 - 10 Hz extension at the bottom end, which is good for movies that have content that low. That's without the ports plugged or the speakers EQd though, I'll look into doing that as well.

Conrad · « **Reply #18 on:** 20 Sep 2018, 06:47 pm »

So I just used the realtraps room calc app and it put the first mode at 47.08 for a 12 foot long room.
The second mode is at 94.17, which lines up with the above. Am I missing something?

Edit: I did some more reading. Looks like a mode can occur for half the wavelegth: http://www.acousticfrontiers.com/room-modes-101/

richidoo · « **Reply #19 on:** 20 Sep 2018, 07:50 pm »

The fundamental mode is the lowest frequency mode that will ring in the distance between boundaries, but the upper harmonics (multiples) of the fundamental will also ring within that same dimension. aka overtones, etc. Just like you can hear overtones ringing in a piano or guitar strong besides just the nominal note. You'll see that modecalc also displays these harmonics so you know if they gang up to make an especially strong mode.

modecalc is only useful for seeing the axial modes: ringing reflections between parallel boundaries. It doesn't consider tangential or oblique modes, but it can explain some of the modes you might see on a test sweep.

Suggestions for small listening room

guest61169