Speaker Position Problems Magnepan 20.7's

[Bilbo]

Since upgrading from Magnepan 1.7's to the 20.7's I have had a hard time finding the right position to place them.  Despite the 20.7's having the reputation of better bass response I have yet to experience any substantial gain in bass weight or slam. (There is a slight improvement but not what I was hoping for).  The 20.7's have been run in for over 450Hrs and demonstrate superiority over the 1.7's in areas such as mid to high frequency response, detail, sound stage width and depth however the bass is making me feel underwhelmed.

 

When I run REW sweeps of the room affected response of the 20.7's and compare them with sweeps I did of the 1.7's it appears the 20.7's have bigger frequency response swings and do not have as smooth a response as the 1.7's.

I have moved them into a position that sounds best so far which is within 30cm of the 1.7's best position.  This position was selected after a large number of sweeps in 10cm increments as having the smoothest response between 50Hz to 400Hz.  The problem is the large dip at 68Hz and worse, the chunk missing from 120Hz to 200Hz.  I can hear this missing chunk as a thinness in upper bass and lower mid-range affecting the timbre of voices and instruments operating in that frequency band.

 

Here is a screen shot of the two responses.  The Red line is the 1.7's and the Blue line the 20.7's. The plots are smoothed to 1/12th and the serious dips in response can be easily seen.  The mic position is almost the same and does not appear to affect the response much below 200hz even if I move it +/- 1/2 a metre. The room has acoustic treatment such as corner traps,front wall and rear wall absorbers & diffusers and first reflection absorbers and sounded pretty good with the 1.7's. Dimensions in metres are 9.18L x 4.42W x 2.54H plus a 1.5 x 3m deep alcove at the back left side of the room.  I would like to get the response to within/- 5dB in this band (50 - 400Hz) so clearly I have some work to do.  I am however quite frustrated as I can't seem to work out where to go from here.

Any help would be very appreciated! 

 

 

PS*  I can't find any way of attaching the REW .mdat file to this post.

 

[HdB]

On the fly, as they say, I'd suggest the 68Hz is possibly the peak created by the room width (1000/68 = 14.7ft or approx. 4.5 metres)  just the numbers seem to be too convenient and the narrow width of the trough seems to indicate a cancellation point

 

The 2 dips in the lower mids about 150 and 180 don't seem to be related to anything except maybe the floor/ceiling height (1000/2.54 x 3.3 = 119Hz) unless you've got some humungous couch in the listening position or a large hard topped table maybe - the plots of the 2 speakers are fairly similar so it appears that your room is playing a significant part in the response 

 

At the risk of sounding fatuous, maybe get one of those diffuser panels and hold it on the ceiling above one speaker with a stick, 2 broom handles, etc to see if any change in response - also maybe about mid distance between speaker and listening position (on the ceiling again) to create interference to 1st ceiling reflect point.

 

Another silly one but maybe tilt the speaker back about 5 - 10 degrees and see if any change - don't think so, but I sometimes find lifting the speakers off the floor a bit more can be useful too ...

 

Unfortunately, I don't have any experience with any Maggies, so offering just guesswork - 'normal' speakers behave a bit differently

 

You may find it helpful to actually reduce some of the absorbtion and diffusion to see what changes, particularly the wall behind the speakers - if you can produce changes in the freq response without changing speaker position, it's an option or fresh starting point

 

Did you get a good coat of sealer paint on the foam diffusers?   The ones here just seem to suck up paint like a sponge and I'm thinking about making up a smooth cream consistency of that wall patching 'polyfilla' to fill the little holes first, or something to that effect

[Addicted to music]

Add a sub.....

[Bilbo]

Thanks for the comments.

I forgot to mention the room has ceiling/corner absorbers and I do have 2 sub woofers (Rythmik 12SE's) controlled by an Antimode 2.0 DSP unit. The measurements were taken with the subs muted.

I am trying to get the 20.7's response as flat as I can because I only want to use minimal sub reinforcement.  I also don't want to push the sub crossover beyond about 50Hz as this tends to start muddying the very clean bass delivered by the 20.7's.  Here is the measurement with the subs.

 

 

From this you can see that I can correct some anomalies below 100Hz but the dips between 120 - 200Hz are more audible and I particularly want to try and get rid of.

There is a possibility that the equipment stand I am using encourages some minor peaks and dips but I don't think that is the main cause.

@jrhill  is probably right in that the wall/ceiling could be the issue but it's already well covered with absorbers.  Also due to the speakers behaving as line sources the ceiling should be a lesser problem I would have thought. 

[HdB]

I had a brief read up on these Maggies and can't see why you'd ever need a sub at all - but maybe I'm missing something

 

As you mentioned, both the line source nature of the drivers and their size and would probably reduce any ceiling reflection issues - plus the horizontal nature of the diffusers would enhance this too - perhaps this might be part of the problem, unlikely though it may seem - the double dips in the low mid octave appear to be interference but your 'waterfall' doesn't seem to indicate a significant energy imbalance - seems remarkably smooth actually

 

There was a mention of the extended 'break-in' time of the drivers but I don't think that exhibited here

 

No much further to add unfortunately - perhaps the new drivers are creating a mismatch with the amps?  (scratching around here!)

 

Back on your first post, the graph shows a large peak in the bass response at a bit below 40Hz and your 'waterfall' indicates a quite pronounced standing wave, if I've read it correctly - looks to be about 10dB - what's creating this?  

If it's the length of the room resonance problem, you'll need a limp mass, or hermholz, trap to get rid of that, I think, rather than back wall diffusers - absorbers will suck out too much low mid and indirect reflections -  my opinion here, mind you ...

 

Is there a dedicated Maggie website and has this problem arisen before?

[Bilbo]

The 40Hz peak is due to the 9.18m room length and effect of the alcove.  To get rid of it requires as you suggested more elaborate bass trapping or careful placement of the sub to counteract it.

That's why that particular peak doesn't concern me as much as the dip between 120 - 200Hz.  I tried tilting the Maggies using a 12mm strip of wood and ran some sweeps of it as below.  I have changed the frequency scale up to 20KHz to show where the greatest changes occur.  Above 1.5KHz it certainly changes the response but not so much in the problem area.  The dip at 135Hz becomes pronounced when you tilt them back which would indicate ceiling as the primary problem.  The speakers vertical position is in purple, tilted forward is green and tilted back is blue.

 

[sir sanders zingmore]

23 hours ago, Addicted to music said:

Add a sub.....

 

Wont help a lot if it's a suck out 

[HdB]

Well that's bloody interesting - haven't seen that before

So peak no 135 is touchy about speaker vertical, eh - somehow, there's a cancellation going on here unless it's a change in the microphone's field - try something a bit on the weird side - put a cushion/pillow on top of one speaker and see if this alters anything - then try the other to see if the same change occurs, if at all

 

Another thought ... as the half wavelength of 135Hz is a bit over a metre (3.7 feet, no?), how far are the panels away from the side walls (ignoring the absorbers for now) and are both sides the same? 

Is the other panel a bit less than a metre (2.8ft) that might correspond to the approx 180Hz trough?

 

You might drop a line to Bob Prangnell at Mad Scientist Audio in NZ - he runs big maggies and has recently moved house so will be going thru a similar process of optimising the speaker position.

Who else up there (Brisbane) runs the 20.7's?

[frankn]

Can you put a reasonably scaled schematic of your room, speaker placement, mike(sitting?) position and significant physical changes in the room, windows, doors, large furniture pieces  etc. 

[Robin.perry]

I don't have maggies but am really interested in them, I heard them at the Melbourne show last year. My reason for adding a comment may still be valid though.
I have a room I now use as a home cinema separate from the audio room. This used to be the main Hi-Fi listening room. In this room my Hi-Fi never sounded right no mater what I did hence moving the Hi-Fi into a different room and resurrecting the cinema equipment. The cinema room consists of M&K main speakers with a REL Sub. No matter where the sub is in the room, and I mean anywhere, there is a measured -20db flat spot in the bass (around 40-50hz) at the same point in the room which just happens to be the listening position, above and below it's fine. I've even tried the sub at different heights to no avail. Always at the same spot the same frequencies just seem to disappear. I resigned the room does not work and as noted above took the Hi-Fi out and endure cinema with a void in the bass which I don't notice so much with movies. After a number of house moves and my Hi-Fi following me to each, I've learnt that the room makes the biggest difference. I really miss a large rented house along the way because my Hi-Fi used to sing in it.
Maybe your like me and the room just sucks.

[Bilbo]

 

@Robin.perry Yeh - the room does suck.  With the length close to double the width and that in turn close to double the ceiling height it's not a great room.  However it's what I've got and at least it is a dedicated music room so I'll just have to deal with it.

 

@jrhill  Balancing pillows on top of the Maggies could be tricky and possibly dangerous around the ribbon tweeters - a gust of wind and all that....

Changing the distance to the walls does not greatly affect the band in question in fact the better response comes when the speakers are closer together.  I'm also not aware of any other 20.7 owners in Brisbum.

 

@frankn  I'll get to work on that and post it for review. Thanks.

[Black Orange]

If you can't fix the room issues maybe think of getting a Room EQ and speaker correction unit like the DEQX. 

Fantastic units, they helped me fix my room issues. Also DEQX will be able to time and phase align the Subs with the maggies. 

 

[Bilbo]

6 hours ago, Black Orange said:

If you can't fix the room issues maybe think of getting a Room EQ and speaker correction unit like the DEQX. 

Fantastic units, they helped me fix my room issues. Also DEQX will be able to time and phase align the Subs with the maggies. 

 

 

I have thought of going down this road but the thought of more boxes and cables has kept me from taking the plunge.  I got the 1.7's response reasonably well tuned into the room but the problem with the 20.7's is a new one and has caught me off guard.  

[aechmea]

17 hours ago, Sir Sanders Zingmore said:

 

Wont help a lot if it's a suck out 

Well, yes and no.

 

It helped considerably in my case.

 

Despite the attention given to room modes in the literature and internet chatter, modes were not the cause of the peaks and troughs I had in the low freqs in my room with my speakers.  Instead my 'problems' were actually comb filtering (aka SBIR) rather than the standing waves of modes.

 

In short SBIR is the interaction of the direct wave with a strong reflection.  If it is half a wavelength (or odd multiple thereof) out of phase there is a cancellation at that frequency.

 

Planar speakers have an equal amplitude and opposite polarity backwave, so when that interacts with the direct wave after a bounce or two off the front and back walls then a typical comb filtering response might be seen.  The typical comb response graph can be obscured by other reflections from floor, ceiling and side walls.  I used an Excel spreadsheet, room dimensions and some trigonometry to calculate path differences from speaker to chair via various wall reflection paths.  These calculated path differences matched pretty well with many of the peaks and suck-outs.

 

So I moved both subs to unusual places around the room.  The sub that I placed at the wall behind the chair (rear wall) filled in the suck-out that was caused be the Maggies bouncing off the rear wall back to the chair.  Rear wall to chair = 2.5m ==> path difference  = 5m  ==> frequency of that as half wavelength (cancellation) = 35Hz-ish.

 

I place the mains and chair for the best soundstage, imaging, clarity and that sort of thing then place sub(s) to fill in the bass 'holes'.  Sub positioning is never anywhere near where the mains are, but maybe that is just my room's geometry.  My use of subs has nothing to do with bass reinforcement as such which seems to be the HT approach, rather a way to even out the bass in support of the room geometry and mains.

 

 

----------------

 

@Bilbo Hi Bill

 

I would think that your suck-outs at 150Hz and 180HZ would almost certainly be room geometry/positioning/reflection/SBIR.  But what exactly is the culprit.

 

180Hz is approx a path difference of 0.95m.  150Hz is a bit longer at 1.14m.    So the reflection from what surface/object is able to give a 1m-ish path difference.

 

Unlikely to be floor bounce.  Floor bounce for a 3m listening position 1m off the floor is about 300Hz and higher freq for seats further away.  Ceiling bounce is a possibility but only if the ceiling is 2.3m or less and even then planars don't have anything much off-axis to contribute to floor or ceiling bounce.  Planars have a natural null at 90º and 270º so normally side walls don't matter much either, if at all.  Its really all about the directwave and the reverse polarity backwave.

 

Hmm, so I would be looking at something roughly 500mm from the speaker or the mic that the positive direct wave might be reflecting from or something 1m away that the reverse polarity backwave can reflect off. [how far from the front wall are the bass panels?]

 

---------

 

On a different tack but all effectively changing the direction of the first reflection of the backwave and the directwave.

 

1.  You might try crossing the speaker axis in front of your chair.  This has the effect of angling the back wave in a different direction and maybe reflecting off an entirely different set of walls.

 

2.  Point them directly down the length of the room.  Same reason.

 

3.  Somewhere in between.

 

4.  I tilt mine downward significantly, so that the top and bottom of the speaker are equidistant from the ears.  Somewhat disconcerting to look at.  They are very tall and my ears would be below half way were the speakers upright.  However it's unlikely that such a small path difference between top and bottom would be anywhere near 180Hz.  But it does change the path of the backwave and directwave.

 

5.  I have the ribbon on the outside.  Effectively the bass panels are close together and the backwave bounces off the front wall first and then off to the side, rather than the side wall first and back to the middle.

 

6.  Unless the absorption is quite thick and plentiful it won't have much affect in the freqs that we are considering.  Similarly dispersion is a mid-high freq thing unless the 'skylines' are enormously huge blocks.

 

7.  Have you tried absorption immediately at the back of the speaker to kill the backwave as best you can to see if the cancellations are due to the backwave or not.

 

------------------

 

Having said all of that I am not sure that the 2 major suck-outs would be audible.  There is a shallow dip from 60Hz to 200Hz, then a low hill from 200 to 1000Hz.  I guess that this would give a mid range prominence to the sound, which is sort of what you describe.  The overall look is not unlike mine but in my system I could hit that with a bit of gentle EQ.  Maybe if you reduce 200 -1000Hz by a dB that will make 80 - 200 more wholesome.  One of those supplied ceramic-resistor-things across the mid range panel might help.  Are they meant for the mid range panel, I don't know; the sockets are there.

 

Tis all I can think of for now.  A plan view of the room and positioning as suggested by Frank will help.

 

 

[Bilbo]

@frankn

Here is a turkey scratch of my room.  I left a couple of things out though (1) the listening position is 6m from the front wall and (2) the subs are between the 20.7's and the wall firing backwards when on.  I have also included a picture which is worth a thousand words in my case.  The entrance to the office has no door and is (3.5 x 3.9)m which could well be acting as a giant port!  In the photo you can see the beam running through the middle of the room.  It's 400mm deep jutting out of the ceiling so there another potential problem.  Mike position is at the listening position, ear height.

  

 

 

 

Edited June 19, 2017 by Bilbo
Photos

[Bilbo]

6 hours ago, aechmea said:

I place the mains and chair for the best soundstage, imaging, clarity and that sort of thing then place sub(s) to fill in the bass 'holes'.  Sub positioning is never anywhere near where the mains are, but maybe that is just my room's geometry.  My use of subs has nothing to do with bass reinforcement as such which seems to be the HT approach, rather a way to even out the bass in support of the room geometry and mains.

 

That's exactly the same strategy I am trying to adopt.

 

6 hours ago, aechmea said:

I would think that your suck-outs at 150Hz and 180HZ would almost certainly be room geometry/positioning/reflection/SBIR.  But what exactly is the culprit.

 

180Hz is approx a path difference of 0.95m.  150Hz is a bit longer at 1.14m.    So the reflection from what surface/object is able to give a 1m-ish path difference.

 

The only thing that close is the equipment rack so I took a sweep with absorbers in front of it to see what happens.  It does smooth things a little but clearly not the main culprit.

The red line is the normal configuration and the green line with the absorbers covering the equipment rack.

 

 

I also tried putting the absorbers 1m behind the panels to see what effect that had. The red is normal configuration and the blue line is with the absorbers in place.

 

 

7 hours ago, aechmea said:

On a different tack but all effectively changing the direction of the first reflection of the backwave and the directwave.

 

1.  You might try crossing the speaker axis in front of your chair.  This has the effect of angling the back wave in a different direction and maybe reflecting off an entirely different set of walls.

 

2.  Point them directly down the length of the room.  Same reason.

 

3.  Somewhere in between.

 

4.  I tilt mine downward significantly, so that the top and bottom of the speaker are equidistant from the ears.  Somewhat disconcerting to look at.  They are very tall and my ears would be below half way were the speakers upright.  However it's unlikely that such a small path difference between top and bottom would be anywhere near 180Hz.  But it does change the path of the backwave and directwave.

 

5.  I have the ribbon on the outside.  Effectively the bass panels are close together and the backwave bounces off the front wall first and then off to the side, rather than the side wall first and back to the middle.

 

6.  Unless the absorption is quite thick and plentiful it won't have much affect in the freqs that we are considering.  Similarly dispersion is a mid-high freq thing unless the 'skylines' are enormously huge blocks.

 

7.  Have you tried absorption immediately at the back of the speaker to kill the backwave as best you can to see if the cancellations are due to the backwave or not.

 

I then tried the extreme toe in crossing in front of the listening position and that seems to have had a beneficial effect on the 120 - 200Hz band and smoothed 200 - 350Hz band quite well.

The red is normal configuration, the black line is with the extreme toe in and the green line is no toe in.  It seems these movements fix one area to the detriment of another.  It's very frustrating. 

 

 

 

 

 

[Robin.perry]

If you can't fix the room issues maybe think of getting a Room EQ and speaker correction unit like the DEQX. 
Fantastic units, they helped me fix my room issues. Also DEQX will be able to time and phase align the Subs with the maggies. 
 

As I mentioned I have an interest in the maggies which I listened to at the Melbourne show. They did an A-B comparison with and without the DEQX. Have to say the DEQX was the most interesting thing at the show for me. Very convincing but at 6k it's a bit much for my current system. Might be the best solution for you though. Imagery improvements were immense which is all to do with timing according to the web site, and no doubt a flat response curve too that you seek Bilbo?
I'd have to agree with Black Orange


Sent from my iPad using Tapatalk

[Bilbo]

Going down the DEQX path requires money I don't have. I'm still recovering financially from purchasing the 20.7s!

[HdB]

I hesitate to be a 'bringer of bad news' but it seems you have exhausted most position variations with the speakers at that end of the room and it hasn't changed the fundamental problem of both the 'twin dips' and low bass 'rolloff'' so, it's time to consider something that mayn't be WAF and that's reverse the room so the speakers (amp and rack) is at the other 'fat' end of the room - maybe there's more chance of getting a better a result - it's a lot of trouble, I know, but it just might be what's required.

 

Those DEQX units are marvels of technology indeed, but there are limits to what they can do and correcting response cancellations (troughs) are at their limit before other side effects appear (to me, anyway) - and you'll always be conscious of the sound deficiencies of the room with the relatively new purchase (plus, as you say, another outlay isn't within the present budget

 

Just looking at the change due to the toe-in tests reminds me of something but just a niggle - the drops in the 'twin dips to a more reasonable response there for dramatic increase of the bass response trough means ....  damn, can't quite catch the thought - will see if it comes to me ....

 

Not all is lost, just a hiccup in progress, frustrating as it is

 

Cross post from the other thread -.

My B2 fractal diffusers have been finally cut (the 'leanfuser' ones with the extra fine 'fins' on top of the basic units) - not quite following on to the design exactly, but close enough - 10 pieces x 1200 high and 4 x 600 high for $300, which includes the $100 setup fee = $25 ea - the next lot will come down to about the $16 ea for the 1200mm ones - they suck up paint like a sponge, even the thick paint that's used to render the outside of houses -  applying the 3rd coat at present and it maybe enough to create a hard, tough sealed surface - it'll be interesting to see if the painting actually makes much of a difference in practice ...

[aechmea]

I haven't given up hope for you yet Bill.

 

Forget the 150 and 180 for the time being; lets look at the big suck at 70Hz

 

Time to get the subs out of the far corners and trial them along the side and back walls.  Behind the chair worked for me as did one sub out in the room away from walls, but that proved to be an annoyance for walking.

 

This little gem can give some ideas.  http://www.hunecke.de/en/calculators/loudspeakers.html

It won't work for planars due to the reverse polarity backwave but will for the subs.  So just use it to trial some possible sub positions.

 

I have this ATM, which has doorways, furniture etc. taken into account. Not really to scale, but close enough. [Remarkably similar rooms, by the way.]  Yes, the left sub is in the back right of the room.

 

 

 

-------------------------------

 

When I measure close to and with my (leather) chair in situ then I get all sorts of funnies, mainly in the mid range.  Might be worth trying without the chairs to see whether some lumps and bumps disappear.  I doubt it, but one never knows.

 

-------------------------------

 

Hmmmmm. A speaker (sub) 1.2m away from a wall will have a null at 70Hz when the direct wave and primary reflection combine.  Since both subs are the same distance from the front wall and about 1m-ish away, both are producing a 70Hz null.

 

The Maggies look to be 2m-ish from the front wall.  The backwave has to travel an extra 5m-ish, a wavelength which is equivalent to a freq of 70Hz.  Because the backwave and directwave are reversed polarity, when they combine, this gives another null. 

 

So all 4 speakers (2 maggies and 2 subs) are producing a null at about 70ishHz.  Worth investigating.  Time to get the subs on the move.

 

Edited June 20, 2017 by aechmea

[Nigel]

19 minutes ago, aechmea said:

 

So all 4 speakers (2 maggies and 2 subs) are producing a null at about 70ishHz.  Worth investigating.  Time to get the subs on the move.

 

 

That's cause it is a primary room dimension null, no matter where you put the speakers, that spot is a room mode.

Move the listening / mic position. 

[Bilbo]

@aechmea

I had a good listen last night with the extreme toe in position that lifts the 120 - 200Hz band and I was quite impressed.  The BOING in the bass came back and there was more substance and better timbre in voices and cello etc.  The downside is the sounstage narrowed and the extra roll off in the HF was noticeable.  (Note I am not running the subs at the moment so I can get the best spot for the 20.7's first).

So there is hope that with the right toe in I can address (at least partially) the missing 120 - 200Hz band.  The dip at 68Hz is another matter.  Here are all the plots within a 1.5m (front wall)  x 1m (sidewall) grid of the current position.  The toe in used for these measurements focuses the inside edge crossing at the listening position therefore the ribbons would focus about 1m behind the listening position.  So no matter where I move the 20.7's that dip is persistent.  This indicates to me that these are more likely room modes interacting rather than furniture in the room. Is that a reasonable assumption to make?

 

 

Interestingly here are the measurements with the listening seat up (green) and reclined (orange).  It appears sitting at attention is clearly better.

 

 

 

 

 

[Satanica]

Have you tried Parametric EQ with a computer software player? It might at least help you debug. I use JRiver MC which has extensive DSP options

[Bilbo]

34 minutes ago, Satanica said:

Have you tried Parametric EQ with a computer software player? It might at least help you debug. I use JRiver MC which has extensive DSP options

 

Yes but EQ doesn't work well with dips like mine.  Trimming peaks is another matter and does work but I can hear the filter working as it slightly compresses the dynamics around the treated frequency.

[Snoopy8]

If you have a dsp, suggest trying Multi-sub optimizer.