Jump to content

REW measurements for speaker/LP positioning


Recommended Posts

Am after some advice regarding using REW to aid speaker/LP positions, I've just bought a UMIK-1 & mic stand.

I assume the 'All SPL' measurements is what's used and what smoothing is best for a novice?

I understand there will be compromises with every location but is there a guide somewhere as to what is preferred or what should be avoided.  I did some googling but no luck.

 

Won't be doing any EQ correction, I have bass treatments (tuned wall hanging resonators & corner helmholtz units, both spec'd per measurements taken by an acoustic engineer) & general absorption panels.

We recently got a new lounge which better suites the room and I then managed to get permission from the Mrs to move the speakers further into the room :)  (the accoustic engineer had suggested some placements but the Mrs wasn't impressed at the time!)  So far moving the speakers forward & inwards as was suggested has improved the soundstage, vocals are also more intelligible.

 

Any advice would be greatly appreciated :) 

Edited by David.M
  • Like 2
Link to comment
Share on other sites



1 hour ago, David.M said:

I assume the 'All SPL' measurements is what's used and what smoothing is best for a novice?

 

REW's psychoacoustics is good to use. 1/6 smoothing is good too.

 

How do the plots look, measuring from main listening position?

Link to comment
Share on other sites

11 minutes ago, rand129678 said:

 

REW's psychoacoustics is good to use. 1/6 smoothing is good too.

 

How do the plots look, measuring from main listening position?

cheers.  mic boom only arrived this afternoon so will do some better measurements tomorrow.  I've done a couple with the little tripod that comes with the umik (close to LP), comparing those @ 1/24th smoothing with the measurements taken yrs back by the engineer I'm not sure they're better or worse 🙄, both speaker & LP are now in different locations.

 

 

  • Like 1
Link to comment
Share on other sites

  • 2 weeks later...

Hi David,

 

"All SPL" will give you the frequency response of your speakers + your room's reflections at the spot where your microphone was for the measurement.

For room measurements you will often place the mic where your head would be when sitting in the primary listening position.

Let's assume for a bit you only want to focus on the primary listening position - later you might want to achieve better sound across a listening couch, or across multiple rows in a home theatre...but not yet...

 

On 26/05/2022 at 9:18 PM, David.M said:

mic boom only arrived this afternoon

I'm assuming that means you bought a microphone stand with a boom? like this one ICON MB-01 Height Adjustable Boom Style Microphone Stand w/ Tripod Base | Microphone Stands - Mannys Music // Mannys Music

IMO a microphone stand with a boom is an essential bit of kit for a measurement rig - without the boom you can't place the mic in the spots you need to measure from...like where your head/ears are when sitting in the listening chair!

 

On 26/05/2022 at 9:18 PM, David.M said:

I've done a couple with the little tripod that comes with the umik (close to LP), comparing those @ 1/24th smoothing with the measurements taken yrs back by the engineer I'm not sure they're better or worse 🙄, both speaker & LP are now in different locations.

Changes in speaker position and listening position will result in big changes to measurements - it's not worth spending time on analysing the differences between measurements taken years ago to new measurements if the speakers/mic has moved.

 

The key takeout is if say adjusting speaker position take measurements with the mic in exactly the same position while you move your speakers.

Do the measurements in 1 session without moving the mic at all. If you didn't tighten the boom angle adjuster enough and the boom sags, start over...been there, done that :(

 

Try different smoothing - 1/3 octave smoothing is approx what the ear actually "hears", but looking at the graphs with less smoothing eg 1/6, 1/12 etc shows more detail...so look at them all, but understand that the ear doesn't necessarily "hear" sharp peaks and troughs in frequency response, even though they look bad on a graph - the human ear is way more sensitive to gentle changes in the frequency response, which is why 1/3 octave smoothing is useful.

 

A smooth frequency response is generally agreed as desirable for good "in room" sound, and it's a bit contentious, but IMO a straight but tilted frequency response smoothly falling from 20Hz to 20kHz of 6-12dB works for me (called a "room curve")

 

On 26/05/2022 at 7:31 PM, David.M said:

So far moving the speakers forward & inwards as was suggested has improved the soundstage, vocals are also more intelligible.

 

Unfortunately REW doesn't have measurements for imaging, soundstage, vocal intelligibility :( 

 

REW is an amazing tool, but interpretation of the different measurements REW can take is required to improve things like imaging/soundstage.

 

An example of a REW measurement that might indicate why speaker imaging is poor would be if you measure each main speaker individually, and the Energy Time Curve (ETC) for each main speaker shows a strong 1st reflection in 1 speaker but not the other, or the general shape of the ETC is quite different between speakers, indicating your room doesn't have even absorption on each side of your listening room (eg 1 side is glass doors, the other side is a besser block wall).

 

I found Acoustic Frontier's guide on "ACOUSTIC MEASUREMENT STANDARDS FOR STEREO LISTENING ROOMS" to be useful:

Acoustical Measurement Standards for Stereo Listening Rooms (acousticfrontiers.com)

 

My room doesn't meet any of their standards - but it's a target with metrics you can compare your room with.

 

cheers

 

Mike

 

 

  • Like 2
Link to comment
Share on other sites

On 05/06/2022 at 5:34 PM, almikel said:

Hi David,

 

"All SPL" will give you the frequency response of your speakers + your room's reflections at the spot where your microphone was for the measurement.................

 

 

Hi Mike,

thanks so much for that detailed reply, quite a bit of information (& typing!) :)

A week ago I ended up getting the acoustical engineer back in (he hadn't been back since I installed the bass treatments he organised).  He took some measurements and said the wall hung resonators were doing their job but suggested I shorten the ports in the corner helmholtz units (gave me a specific length for 45Hz).  He said room was sounding pretty good, any case I'd need to start removing bookcases & LP storage to put more treatments in, not WAF friendly!   He suggested I now work on fine tuning speaker placement (hard work @ 100kg each!)

 

At the moment pretty happy with the sound, a couple of test tracks with bass chords that originally used to fluctuate in volume are now smooth & constant plus I'm getting more background detail.

 

I'll have a read of that Acoustic Frontiers guide and look at my room measurements.

 

Overall imo I've gotten a great return on $ spent getting a qualified professional in to measure/advise, if we ever move house it's one process I would definitely repeat.

Edited by David.M
  • Like 1
Link to comment
Share on other sites



9 hours ago, David.M said:

Overall imo I've gotten a great return on $ spent getting a qualified professional in to measure/advise, if we ever move house it's one process I would definitely repeat.

👍

IMO it's also a great way to fast-track your understanding/knowledge of what's going on in your room if you ask questions/and the pro is happy to share...

 

...now you have a measurement rig, if you're inclined for a bit of testing, I would do some "sanity" checks of the following:

9 hours ago, David.M said:

He took some measurements and said the wall hung resonators were doing their job

Do you know what "job" the wall hung resonators were installed to do? eg absorb over a particular freq? - If you don't know it would be entirely reasonable to ask them what they were designed for.

Run a measurement with the mic at the listening position with the wall hung resonators in the room, and again with the wall hung resonators not in the room, leaving the mic in exactly the same position.

If the wall hung resonators are working you should expect increased decay in the REW waterfall chart in the frequency range the wall hung resonators are designed to operate.

 

9 hours ago, David.M said:

suggested I shorten the ports in the corner helmholtz units (gave me a specific length for 45Hz).

Same deal - measure before shortening the ports, then measure after shortening the ports - notate each measurement so you can discern between each measurement.

 

I'm not in any way trying to diss your acoustic guy, my post is about how to check for yourself that the things the pros suggest do the things they're meant to.

You may find that your acoustic guy becomes more engaged when he knows you've undertaken a process to better understand the acoustics of your room!

 

Mike

  • Like 1
Link to comment
Share on other sites

On 07/06/2022 at 7:59 PM, almikel said:

👍

IMO it's also a great way to fast-track your understanding/knowledge of what's going on in your room if you ask questions/and the pro is happy to share..................................

Thanks again Mike for your interest & info.

I have written reports from the engineer.  The wall hung resonators are tuned to 78Hz (they're 1200 x 800 & weighed around 34kg each so not really keen to take them down).

 

Had a browse through the Acoustical Measurement Standards doc you linked, it's great :)  I need a bit more reading to better understand some of it plus working out some of the REW info.

I've managed to replicate some of the simple measurements from the AMS doc which suggest things aren't too bad.  I'm stuck with different left/right sidewalls, one side is a fireplace/large mantle, window/blind & LP storage, other side are bookcases & 2 doorways, assume that would explain the different L & R measurements.

 

Midrange Frequency Response for Left & Right speakers (1/3 smoothing)  is within spec I think

 

T20 & T30 midrange decay is within spec (engineer report says for my room dimension my RT target is 244ms)

 

I couldn't work out how to do the Low Frequency decay chart nor the ETC graphs (which I can't understand right now anyway).

 

I could do with some more low bass absorption but space & WAF are in the way.  To start with I'll experiment with moving speakers.  As room doubles for HT I'm a little limited with listening position, already HT is compromised in favour of stereo. 

 

 

 

 

Left Right Freq Resp.jpg

Left T20T30.jpg

Right T20T30.jpg

Left Waterfall bass.jpg

Right Waterfall bass.jpg

Edited by David.M
Link to comment
Share on other sites

Looks like the engineer has aimed at a studio flat response and slightly over damped Rt - nothing wrong with that especially if AV use included but might look at ways to vary this 'ideal' freq response if your ears aren't attuned to flat freq response - mine certainly aren't for home hifi (at home)

 

The freq area below 250Hz is where many room acoustic problems occur but your 'waterfall' diagrams illustrate exceptionally good linearity and reasonable decay rates - if the quoted 244mS Rt target is the overall figure, you might look towards re-evaluating this for stereo reproduction.

 

... just my 2 cents!

Link to comment
Share on other sites

On 14/06/2022 at 5:07 PM, HdB said:

Looks like the engineer has aimed at a studio flat response and slightly over damped Rt - nothing wrong with that especially if AV use included but might look at ways to vary this 'ideal' freq response if your ears aren't attuned to flat freq response - mine certainly aren't for home hifi (at home)

 

The freq area below 250Hz is where many room acoustic problems occur but your 'waterfall' diagrams illustrate exceptionally good linearity and reasonable decay rates - if the quoted 244mS Rt target is the overall figure, you might look towards re-evaluating this for stereo reproduction.

 

... just my 2 cents!

For background the engineer initially identified 2 immediate issues, one a lot of energy returning to the room from the lobby/stairway (which has hard flooring/walls) and the other reflections from the back wall/windows (which was approx 8db below & 17ms behind the direct sound).    Pending curtains I bought a lot of Polymax XHD 50mm which I had freestanding across the back of the room and also made a few 1200 x 600 framed stands for side reflections (room is 7.1 x 5.5 x 2.4).

 

I had a doorway installed to close off the lobby/stair area and curtains were installed (naturally I went for the heaviest backing & wife chose the facing material which is a heavy linen type fabric).   When the engineer came back he said my room was now too damped and had me remove all the Polymax except 2 of the framed stands which I now have against the front wall behind each speaker.  The only other absorption I have in the room are some acoustic foam panels glued to the ceiling at 1st reflection positions and a large Persian rug (on tiled/concrete floor).   Side walls are pretty much diffused via bookcases/fireplace mantle/LP storage.

 

That document which Mike provided a link to suggests optimal midrange decays times are .2 to .5ms which I'm within but yes at the lower end.   It also says that diffusion can yield absorption effects but without creating a 'dry and sterile' sound associated with overly short decay times.,  I'm assuming I fit into that category as I have a great soundstage depth & width, with the right music the speakers aren't there and at least to me the sound is very engaging.

 

Over the last 2 weeks I've experimented with the speaker position and found a position a bit closer to the front wall (front baffle 1510 from wall) which improved the low bass (more even and not as bloated) & I couldn't sense any adverse affect to soundstage etc.   The measurements above are from that speaker position.  I don't know that I can better the current speaker position, would like some more bass treatment but where they need to go is a WAF issue :(

Edited by David.M
  • Like 2
Link to comment
Share on other sites

  • 1 month later...

I got around to fixing a couple of rattling doors, 1 used to make a noise with the heartbeat opening of DSOTM UK 1st (only the 1st, not with UK 2nd/3rd, MFSL, 30th, AU or SACD).  No doubt they were rattling at other times but I didn't hear it over the music.

Was curious if there would be any measured difference & yes there is an improvement in decay below about 70Hz, no idea whether it's an audible difference?  (before waterfall plots are above)

 

 

Left Waterfall.jpg

Right Waterfall.jpg

  • Like 2
Link to comment
Share on other sites



  • 5 months later...
1 hour ago, Batty said:

I have no idea what these are telling me.

You can see that you have fairly loud ringing in the low bass, and that the bass is relatively flat (about t+/- 5dB) .... but the scale is too coarse to see much.

 

Next time.... On the first one, you can see from ~75dB, down to 50 .... resize the chart so you can see at least a range of 40dB below the max (so 75 down to 30 ish)

 

On the second, resize it so the chart spans about 30dB or so.

  • Like 1
Link to comment
Share on other sites

  • 3 months later...

Edit: Just realised this isn't a general thread and thus I may be hijacking? 

 

 

Umik-1 arrived today. This may be a new neuroses for me, but hopefully not...

 

Measurement taken at LP with the provided small tripod. From bottom to top, according to SPL overall, is L, R, and L + R speakers. What might account for the >10dB channel L/R difference? 

 

capture1_spkr_measurement.png.c9451c07192426c665efc4812f4083ef.png

Edited by crtexcnndrm99
Link to comment
Share on other sites

34 minutes ago, crtexcnndrm99 said:

Umik-1 arrived today. This may be a new neuroses for me, but hopefully not...

 

Measurement taken at LP with the provided small tripod. From bottom to top, according to SPL overall, is L, R, and L + R speakers. What might account for the >10dB channel L/R difference? 

 

capture1_spkr_measurement.png.c9451c07192426c665efc4812f4083ef.png

 

Yes, I wonder why the difference, Ash.

 

The graphs for L&R are similar - just mostly about 8-10dB different.  :(

 

Can you check that your preamp and your power amp output the same signal level - R&L - at 1kHz?

 

Link to comment
Share on other sites



  • 4 weeks later...
On 18/04/2023 at 8:51 PM, crtexcnndrm99 said:

Edit: Just realised this isn't a general thread and thus I may be hijacking? 

 

On 08/01/2023 at 11:43 AM, Batty said:

I have no idea what these are telling me.

Both are hijacking @David.M's thread...

If @David.M wants, the mods could move...

 

@David.M - most rooms IME measure poorly in the bottom end - your waterfalls look better than most I've seen.

If you have EQ capability you could muck with your baseline frequency response...

...only apply broad EQ and start with EQ cut only, and see what you get.

 

EQ boost of dips can have variable results for multiple reasons - hence if applying EQ you should start with lopping off peaks rather than filling in dips...and applied EQ should always be broad (low Q).

 

Mike

  • Like 1
Link to comment
Share on other sites

On 10/05/2023 at 11:10 PM, almikel said:

 

Both are hijacking @David.M's thread...

If @David.M wants, the mods could move...

 

@David.M - most rooms IME measure poorly in the bottom end - your waterfalls look better than most I've seen.

If you have EQ capability you could muck with your baseline frequency response...

...only apply broad EQ and start with EQ cut only, and see what you get.

 

EQ boost of dips can have variable results for multiple reasons - hence if applying EQ you should start with lopping off peaks rather than filling in dips...and applied EQ should always be broad (low Q).

 

Mike

 

I don't have any EQ capability and hadn't planned to go down that path especially as most of my listening is LP.  

Really happy with how it's sounding, don't think I could do another box plus A/D/A conversion chasing the last few %!

ps I've no issue with others posting here.

Link to comment
Share on other sites

21 hours ago, David.M said:

I don't have any EQ capability and hadn't planned to go down that path especially as most of my listening is LP.  

By "LP" I assume you mean you mostly listen to vinyl and don't want to use DSP for EQ to avoid another analog to digital to analog conversion...fair enough 👍

...when I 1st read your post, I inferred LP meant "Listening Position" in my brain...ie you only listen from a single listening position, which obviously carries a different meaning...

 

...regardless...

21 hours ago, David.M said:

Really happy with how it's sounding, don't think I could do another box plus A/D/A conversion chasing the last few %!

...that's really the end game right there - you're happy with the sound !

 

I've run DSP since 2010, with DEQX as my pre-amp and active crossover - IMHO my system sounds vastly better with DSP and I couldn't be without it...but that's completely irrelevant in your case.

 

All that matters is:

22 hours ago, David.M said:

Really happy with how it's sounding

👍

 

Mike

  • Like 1
Link to comment
Share on other sites

22 hours ago, David.M said:

I don't have any EQ capability and hadn't planned to go down that path especially as most of my listening is LP.  

 

All of my 'serious' listening is to vinyl - and DSP has made a substantial improvement to what I hear from my spkrs.

 

22 hours ago, David.M said:

Really happy with how it's sounding, don't think I could do another box plus A/D/A conversion chasing the last few %!

 

Sure, if box minimisation is the key issue.  :thumb:

 

But if improved SQ is your goal ... I suggest you need to consider using (a particular type of) DSP to add a "linear phase overlay" - ie. a "FIR" overlay - to the standard IIR filters which your spkrs have ... which have phase anomalies across the XO range.

 

Link to comment
Share on other sites

36 minutes ago, andyr said:

But if improved SQ is your goal ... I suggest you need to consider using (a particular type of) DSP to add a "linear phase overlay" - ie. a "FIR" overlay - to the standard IIR filters which your spkrs have ... which have phase anomalies across the XO range.

Big call @andyr!

Of course I'm a fan of linear phase EQ applied appropriately, and well applied EQ in general (FIR and/or IIR), but I've come back to accepting well designed passively crossed speakers can sound mighty fine also!

23 hours ago, David.M said:

Really happy with how it's sounding,

It really is OK for @David.M to stop there without further improvements/upgrades and not chase diminishing returns... @David.M - sit back and enjoy the music!

 

Mike

  • Like 1
Link to comment
Share on other sites



10 hours ago, almikel said:

Big call @andyr!

Of course I'm a fan of linear phase EQ applied appropriately, and well applied EQ in general (FIR and/or IIR), but I've come back to accepting well designed passively crossed speakers can sound mighty fine also!

 

I didn't say they couldn't, Mike - I merely said that playing music through Roon's "linear-phase" Convolution makes the spkrs sound better than playing the same LP directly, outside of Roon.

 

And I've heard this effect on 3 different systems.

 

  • Like 1
Link to comment
Share on other sites

@almikel thanks Mike, yeah I won't pretend there isn't room for further improvement but I think I've maxed what I can do with the room & speaker/LP placement.  I was a bit lazy not installing the custom bass treatments, when I finally got off my a__ and mounted them as intended I kicked myself for having left it so long :( , the improvement to the whole sound was significant.  Given now the bass is respectable and other measurements (FR/RT) are within targets set out in that Acoustic Frontiers doc I'll try to just enjoy the music!

 

@andyrI really have no idea about filters and all that tech stuff, I wasn't sure if you were talking about my speakers specifically or something more general.  From what Mike has said I'm guessing you mean an issue with all passive speakers?  Wth Rockport Any Payor customises every crossover with the specific driver unit it will be mated to so maybe it's less of an issue or even insignificant, I don't know.   Anything obvious I can listen for, hard I know without being able to A/B.

 

Link to comment
Share on other sites

Hi David,

 

22 hours ago, David.M said:

 

@andyrI really have no idea about filters and all that tech stuff, I wasn't sure if you were talking about my speakers specifically or something more general.  From what Mike has said I'm guessing you mean an issue with all passive speakers?  Wth Rockport Any Payor customises every crossover with the specific driver unit it will be mated to so maybe it's less of an issue or even insignificant, I don't know.   Anything obvious I can listen for, hard I know without being able to A/B.

 

I was talking about spkrs in general.  As Andy Payor's Rockport spkrs use passive XOs (like the majority of spkrs out there) ... he will be using 'IIR' filters.  (Just to explain - a 'crossover' from one driver to the next one higher in the frequency range involves a high-pass filter combined with a low-pass filter; the high-pass filter sends only the high frequencies to that driver ... the low-pass filter sends only the low frequencies to the other driver.)

 

IIR filters exhibit a phase-change across the XO region - this is audible and is the reason so-called "full-range" drivers have become popular ... as they don't use a XO. 

 

In today's DSP world, you can use computer power to create 'FIR' filters instead of IIR filters.  FIR filters do not exhibit a phase change - so by using them ... you get the lack of muddiness typical of a 'full-range' driver, whilst using multiple drivers to really cover the full frequency range.

 

When AP says he tailors the XO to the specific drivers he's using ... what he means is that the XOs in each Rockport spkr will differ slightly, to compensate for the fact that what are supposed to be the 'same' drivers ... measure slightly differently.

 

Link to comment
Share on other sites

@andyr was doing myself a disservice saying "I have no idea" !, I broadly know the purpose of filters/crossover but not the details of how.

Did a little googling and really no more aware of what I should be listening for or am missing, I guess to it's largely going to be system dependent.  Wondering though, if the filter/crossover is a problem wouldn't it need to be either bypassed or the signal processed after the crossover?

My set up is pretty traditional, just phono & a CD/SACD player into amp, part of the reason I'm not keen on pursuing EQ.  Chasing perfection I think would be never ending, particularly since getting the bass treatments sorted I'm comfortable living with what failings are left :)

Link to comment
Share on other sites

2 hours ago, David.M said:

Chasing perfection I think would be never ending, particularly since getting the bass treatments sorted I'm comfortable living with what failings are left :)

 

Different people have different approaches.  :)

 

Some are endlessly trying to improve the sound they hear ... others are happy with what they hear.

 

2 hours ago, David.M said:

Wondering though, if the filter/crossover is a problem wouldn't it need to be either bypassed or the signal processed after the crossover?

 

You can't "bypass (ie. remove) an XO" ... unless you decide to use a (so-called) "full-range driver" - which, by definition, has no XO.

 

(I say 'so-called' because no "full-range driver" exists - which delivers 20-20kHz (the accepted FR we want to listen to) within -3dB limits.)

 

If you decide to use multi-driver spkrs - to get a full-range frequency response - XOs are a fact of life.

 

What you can do, however - wrt to XOs - is use one of several different types:

1.  passive - so the stereo signal coming from the power amp in split into different frequency bands inside the spkr.  Generally (but not always), the frequency splitting involves a HP/LP pair of filters, for each pair of drivers.

 

2.  analogue active

3.  digital active

 

With these, the stereo signal from a preamp goes into the XO - and 'X' channels come out; each output goes into an amp 'channel' ... which is directly connected to a driver.

 

#1, 2 & 3 normally are your standard IIR filter - which, unfortunately, produce a phase change across the XO region.  When you listen to the spkrs, you can't really pick up the phase anomalies - but when you compare them with a 'full-range' driver ... you can pick the difference.

 

Alternatively, with the advent of computers, you can employ active 'FIR' filters - which are linear phase.  Unfortunately, you need quite a lot of processing power ('taps') to create FIR filters ... the miniDSP 10x10HD which I used to have for the 4-way XOs in my active system - and the miniDSP nanoDIGI 8x8 which I have now (for 3-way XOs) - do not have enough processing power to create FIR filters.

 

2 hours ago, David.M said:

My set up is pretty traditional, just phono & a CD/SACD player into amp, part of the reason I'm not keen on pursuing EQ.  Chasing perfection I think would be never ending, particularly since getting the bass treatments sorted I'm comfortable living with what failings are left :)

 

As I said before - if you're happy with the sound ... that's important.  Me - I'm always "happy with the sound I'm hearing" ... but that doesn't mean I'm not interested in making it better.  :)

 

Don't forget - forgetting the XOs themselves - DSP/EQ also enables you to deal with the room modes which every room has.  :o

 

I was very happy using my miniDSP setup to send vinyl to my 2-way active spkrs & subs.  (Which involved an A2D converter on my phono stage, for input to the miniDSP unit, and then 3x DACs after the miniDSP, to drive the 6 amplifier channels.)

 

But I've been able to achieve a significant improvement in SQ by now playing my vinyl via Roon - using my PC's processing power and making use of Roon's "Convolution" capabilities - to turn the IIR filters produced by the miniDSP unit into "linear-phase" FIR filters!  :o

 

Link to comment
Share on other sites



  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...
To Top