REW response best placement?

goatlips

Prominent Member
Joined
Nov 22, 2004
Messages
3,411
Reaction score
1,479
Points
707
Age
51
Location
Henfield
Good eve guys,

I have attach my subwoofer REW response in the right corner of my room measured every 90 degrees on its axis.

Green driver facing left wall
Blue driver facing in room
Red driver facing right wall.

Which one is the best placement and why?
20200825_214730.jpg
 
Depends on your crossover, but potentially green is your best option.

Worth doing your front LCR with your subs on to see what it looks like, 10-150hz on REW
 
This is quite a significant difference. I was always under the impression oreintation in the same spot wasn't supposed to be much if any different.

Thanks for sharing the response, very interesting. This has made up my mind to add another down-firing BK rather than move to side facing 1961's, looks like there could be some risk in doing so.
 
You'll always get a difference, but only as much as you're changing the distance between the driver and a boundary wall.

If we assume that the sub above is against or close to the right wall, that null that facing the right wall is creating is likely the result of the sound bouncing off the wall and cancelling the direct sound from the driver. Rotating the sub changes those relative distances and therefore moves the null (as seen on the green trace). When pointing into the room it moves again, as seen on the blue trace. It's present in all responses, it's just at different frequencies as the distances between direct and indirect sound change.

Don't think that DF removes this issue, sound moves in three dimensions. I guess the sound radiates from all sides of the sub in a DF. That could make it better or worse, depending on placement and the room. Certainly DF don't lend themselves to a consistently better response, otherwise everyone would make DF subs and only DF subs.

You can test it by turning your Monolith DF on it's side and taking a measurement.
 
I understand why, as you say the boundary wall reflection, my belief (based on user comments) was that it wouldn't actually manifest itself as dramatically as it shows in the OP's responses. General reasoning has always been cited as the standing wave lengths of bass frequencies are in the meters region. Clearly orientation does in fact have a bigger part to play.

No DF is not better, it's simply my setup (and I really didn't want to drag the thread off on a tangent in my direction) has great response as it is with a single DF. I want a second sub for co-lo gains but also want to change colours of my sub - it therefore represents a good opportunity to move from BK to Arendal. But the limitations in placement orientation means that I potentially could get unlucky and induce a null - which I don't presently have. 2 x DF's = no risk for my specifics.
 
You'll always get a difference, but only as much as you're changing the distance between the driver and a boundary wall.

If we assume that the sub above is against or close to the right wall, that null that facing the right wall is creating is likely the result of the sound bouncing off the wall and cancelling the direct sound from the driver. Rotating the sub changes those relative distances and therefore moves the null (as seen on the green trace). When pointing into the room it moves again, as seen on the blue trace. It's present in all responses, it's just at different frequencies as the distances between direct and indirect sound change.

Don't think that DF removes this issue, sound moves in three dimensions. I guess the sound radiates from all sides of the sub in a DF. That could make it better or worse, depending on placement and the room. Certainly DF don't lend themselves to a consistently better response, otherwise everyone would make DF subs and only DF subs.

You can test it by turning your Monolith DF on it's side and taking a measurement.

No matter what I do with the sub I've got a 75hz null and have a 30hz gain just the way my room is I guess?

Would you choose the green line?

I'm not sure why but its currently running the red one, i took the measurements earlier in the year and i didnt act on it lol.
 
I understand why, as you say the boundary wall reflection, my belief (based on user comments) was that it wouldn't actually manifest itself as dramatically as it shows in the OP's responses. General reasoning has always been cited as the standing wave lengths of bass frequencies are in the meters region. Clearly orientation does in fact have a bigger part to play.

No DF is not better, it's simply my setup (and I really didn't want to drag the thread off on a tangent in my direction) has great response as it is with a single DF. I want a second sub for co-lo gains but also want to change colours of my sub - it therefore represents a good opportunity to move from BK to Arendal. But the limitations in placement orientation means that I potentially could get unlucky and induce a null - which I don't presently have. 2 x DF's = no risk for my specifics.

I completely agree, and low frequency waves are meters long, you're right.

My thinking is if you've got a sub that's 1m away from the left side wall, 3m away from the right side wall and 3m away from the MLP, if it's pointing at the MLP the indirect sound radiating sideways might not be enough to interfere with the direct sound, it's a cone with the strongest being direct. If you face the left side wall there's now minimal direct sound but the reflected sound is ~5m, 1m to the side wall, ~4m reflected to the MLP. Pointing at the right side wall makes the reflected sound ~7m, 3m to the right wall and ~4m to the MLP. That's a significant difference.

I'm speculating here, but that was what I was thinking when I wrote my initial response.

If you can measure your sub in each potential position you can model dual subs in REW so you can buy with confidence knowing what your response is going to look like. As I say, you can lay your Monolith on it's side to get an idea of what the arendals would do. Take one measurement facing one way and the other facing the other way and merge them in REW. Another two at the other potential location and you'll get two pairs that represent the two subs.
 
No matter what I do with the sub I've got a 75hz null and have a 30hz gain just the way my room is I guess?

Would you choose the green line?

I'm not sure why but its currently running the red one, i took the measurements earlier in the year and i didnt act on it lol.

Green or blue. Blue has worse response in the 70-90 region, but better in the 90-100 region. It's worth measuring the mains and seeing if that area needs subs or not.

If you can measure all as you have them now, with a timing reference, and measure the LCR with no subs and the same timing reference, you can model the integration in REW. Post up your MDAT file and I can take a look and show you how.
 
Green or blue. Blue has worse response in the 70-90 region, but better in the 90-100 region. It's worth measuring the mains and seeing if that area needs subs or not.

If you can measure all as you have them now, with a timing reference, and measure the LCR with no subs and the same timing reference, you can model the integration in REW. Post up your MDAT file and I can take a look and show you how.

Thanks for your response, can you please explain about timing reference??
 
I'm speculating here, but that was what I was thinking when I wrote my initial response

No I agree with it, and I think the difference from what I read is largely underplayed. I think only the other week someone asked about FF vs DF and I don't recall there being anything to advise of this potential, which is depending on what the FF is 'facing' it could actually be quite different.

I could measure my sub in various orientations to get an idea, I probably will on the weekend - I'm of the opinion though that given the response I have is excellent now anyway then it's a bit of a no-brainer to just add another. I've emailed BK to see if they could sell me an empty XXLS400 chassis along with another new one, that way I can transfer my components to the bare chassis, probably the cheapest way I think to get 2 gloss black subs in the room!

To the OP, I'd go for the green orientation, makes sense to get the largest most consistent output in the region you want it, i.e 80hz and under really.
 
No I agree with it, and I think the difference from what I read is largely underplayed. I think only the other week someone asked about FF vs DF and I don't recall there being anything to advise of this potential, which is depending on what the FF is 'facing' it could actually be quite different.

I could measure my sub in various orientations to get an idea, I probably will on the weekend - I'm of the opinion though that given the response I have is excellent now anyway then it's a bit of a no-brainer to just add another. I've emailed BK to see if they could sell me an empty XXLS400 chassis along with another new one, that way I can transfer my components to the bare chassis, probably the cheapest way I think to get 2 gloss black subs in the room!

To the OP, I'd go for the green orientation, makes sense to get the largest most consistent output in the region you want it, i.e 80hz and under really.

Where are you putting the second sub, are you colocating?
 
Thanks for your response, can you please explain about timing reference??

A timing reference is a signal from another speaker that you record consistently at the beginning of each measurement. It allows REW to align two different responses to a common reference point.

I use my surround back left (SBL) as my timing reference when I take a measurement.

So I measure my left speaker and a I get a blip from SBL and then the full left sweep.
Now when I measure my sub using SBL as a timing reference I get a blip from SBL and a sub sweep.

When those two measurements are loaded into REW the application can align the two "blips" and produce accurate relative timing information for the left and sub channels. That means that when I apply a 2ms delay to my sub in REW, the predicted response matches the response I get when I add 2ms to my sub in my room.

If you don't have it REW considers all responses to start at 0 which won't be true given the delays in the receiver, the position of the speakers in the room, and any DSP in the chain out to the sub (minidsp, the sub itself, etc.)
 
Where are you putting the second sub, are you colocating?

Yep, I've no nulls to fill in really - just need (ok want!) a boost so I can cut the total output to be smooth, and have large amounts of headroom! Pretty sure I did you a graph on another thread the other day ;) or I'm suffering from serious deja vu!
 
A timing reference is a signal from another speaker that you record consistently at the beginning of each measurement. It allows REW to align two different responses to a common reference point.

I use my surround back left (SBL) as my timing reference when I take a measurement.

So I measure my left speaker and a I get a blip from SBL and then the full left sweep.
Now when I measure my sub using SBL as a timing reference I get a blip from SBL and a sub sweep.

When those two measurements are loaded into REW the application can align the two "blips" and produce accurate relative timing information for the left and sub channels. That means that when I apply a 2ms delay to my sub in REW, the predicted response matches the response I get when I add 2ms to my sub in my room.

If you don't have it REW considers all responses to start at 0 which won't be true given the delays in the receiver, the position of the speakers in the room, and any DSP in the chain out to the sub (minidsp, the sub itself, etc.)

And does REW do all this??
 
Yes. You need to select acoustic timing reference when taking your measurement. REW does the rest.
 
Yes. You need to select acoustic timing reference when taking your measurement. REW does the rest.

Ok I will look into this over the weekend and report back.

So shall i move my sub to face the other wall (green line) and start from there with my mains with dirac off??

How did you manage to output the sound to the LCR?
 
Let me know if you need help.

Green or blue. Consensus seems to be blue is preferable, it'll depend on the response of the mains and the integration I guess.

In REW make sure you're connected via HDMI and use the FlexASIO output. Set windows to multichannel out and make sure the AVR is on a multichannel processing mode. If that's all correct all the channels will be available to you (LCR, SL, SR, SBR, SBL, SW). Choose L and turn off the subs and you'll get just the crossed over L response.

You'll need to disconnect the L speaker and measure the sub to get the sub only. You can use channel 4/LFE, but that won't apply a crossover so isn't any use for testing integration.
 
Thank conrad for you help.

Where is flexASIO and how do you set windows to multi channel?
 
In windows, right click on the speaker icon in your system tray area and choose 5.1 or 7.1 from the available settings.

FlexAsio comes as part of the later versions of REW. I suggest getting the latest beta. It's rare that there's an issue with it and there's often new and useful features.

In preferences -> Soundcard, there's a drivers dropdown. If you select Java you'll only be able to use L, R, or L+R channels. That's fine if you're just doing sub and L/R integration. The other option is ASIO, select that. There should then be an option for ASIO control panel, open that and make sure the right output device is selected. Once that's done hit reload and the output dropdown should have all the channels listed separately.

FYI, if you have access to a mac, they expose all these channels natively and it's slightly easier to work with. I've had trouble connecting my minidsp to my mac though so it's not great for implementation of filters.
 
In windows, right click on the speaker icon in your system tray area and choose 5.1 or 7.1 from the available settings.

FlexAsio comes as part of the later versions of REW. I suggest getting the latest beta. It's rare that there's an issue with it and there's often new and useful features.

In preferences -> Soundcard, there's a drivers dropdown. If you select Java you'll only be able to use L, R, or L+R channels. That's fine if you're just doing sub and L/R integration. The other option is ASIO, select that. There should then be an option for ASIO control panel, open that and make sure the right output device is selected. Once that's done hit reload and the output dropdown should have all the channels listed separately.

FYI, if you have access to a mac, they expose all these channels natively and it's slightly easier to work with. I've had trouble connecting my minidsp to my mac though so it's not great for implementation of filters.

It seems my cheap HP laptop doesn't have multi channel sound card. Only outputs in stereo via HDMI.
 
Personally I'd go with blue. It's the smoothest overall. The green has a null at 90Hz and the red just above 70Hz. Yes the blue has a null at 155Hz but I would assume that is well above your crossover point.

But as conrad mentioned, using a timing reference to make sure the subs play nicely with the mains is possibly more important than the sub's individual response.

For me, using the 'Alignment tool' and the 'match phase slopes at cursor' function (with the cursor at the crossover point) in REW made a huge improvement.
 
Personally I'd go with blue. It's the smoothest overall. The green has a null at 90Hz and the red just above 70Hz. Yes the blue has a null at 155Hz but I would assume that is well above your crossover point.

But as conrad mentioned, using a timing reference to make sure the subs play nicely with the mains is possibly more important than the sub's individual response.

For me, using the 'Alignment tool' and the 'match phase slopes at cursor' function (with the cursor at the crossover point) in REW made a huge improvement.

Thanks for your input.

Yes I thought blue as LFE rolls off at 145hz anyway so the later null wont matter.

I dont understand the time alignment tool? Wont dirac correct that for me? How would I import that into the Avr?

Can you please explain more about the time alignment?
 
Ah, sorry I didn't realise you were using dirac. Yes dirac will do it's best to correct that for you and it will do a pretty good job, it's the best of the 'automated' solutions IMO. But with some practice and knowledge of REW you might be able to do slightly better doing it manually.

Very simply, time alignment means getting the delay/distance settings (in your AVR) just right for each speaker and sub so the sound waves from different speakers and subs match up correctly when they get to you.

There are a few tools built into REW to help you do this. Happy to give some starter instructions if you wanted to try it.
 
Last edited:
Ah, sorry I didn't realise you were using dirac. Yes dirac will do it's best to correct that for you and it will do a pretty good job. But with some practice and knowledge of REW you might be able to do slightly better doing it manually.

Very simply time alignment means getting the delay/distance settings just right for each speaker and sub so the sound waves from different speakers and subs match up correctly when they get to you.

There are a few tools built into REW to help you do this. Happy to give some starter instructions if you wanted to try it.

Starter instructions sounds good.

My real purpose of this post really is to get my sub sounding it's best before running dirac. Dirac picked up that massive null at 70hz and I just wanted to improve that area, also my mains crossover is set at 80hz.

You say you'll might get a better job manually, but how would you import the new settings into the AVR??
 
Starter instructions sounds good.

My real purpose of this post really is to get my sub sounding it's best before running dirac. Dirac picked up that massive null at 70hz and I just wanted to improve that area, also my mains crossover is set at 80hz.

You say you'll might get a better job manually, but how would you import the new settings into the AVR??

If you're going to run dirac anyway there's no point in doing the time alignment yourself. It's an either/or thing, either you let dirac do it or you do it yourself. Doing it yourself may well yield better results but it takes time and trial and error and so on. I'd say let dirac do its thing, forget about doing it manually for now, come back to it next time you get the itch to change things.

You 'import' the setting into the AVR by manually adjusting the individual speaker/sub distance settings, that's it. The complicated part is using REW to figure out the exact settings for best results.
 

The latest video from AVForums

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