Room Acoustics Guide.

Posted by: 911gt3r on 15 December 2018

Dear Forum.

I have used the Search Function here to learn about room acoustics, but found that previous posts on the subject are both scarce and with very limited interest strangely. Having also read ongoing posts and forumers replies to various issues people suffer with their speaker/ room interaction has made me realise, that there are clearly people here that know, what they are talking about. I am hoping that this thread could serve as a comprehensive guide to Naimees, who are in the need of help with their listening room issues and who can be bothered about using the search- function in the future.

I myself was lucky with speaker placement in my previous house, as it literally turned out to be plonk and play. I lived with my relatively big listening room ( 4.7x 7.8) very happily with Naim SL2 speakers which were ultimately replaced by PMC Fact 12 speakers, as I found the SL2s sounding a bit to lean simply because the space was too big for them firing the long way down the room. The PMCs very much have the qualities of the SL2s ( transparency, speed) , but additionally can produce some earth shakingly deep and clean bass being a front ported transmission line design.

As I recently retired my wife and I decided to move to a smaller new build house, where I was lucky enough to have a didicated listening room ( for clarification my avater should give you an idea of layout).

Now this is where my problems started ????. The room is 3.5Wx6.0Lx2.4H. Speakers are placed 1.6 mtrs in front of bifold glass doors. Listening position is 3.2 mtrs from the Facts. Walls is a particular low density plasterboard drylined ie a 20mm gab to solid wall behind them enabling some bass absorption. Floor is suspended concrete and although feeling very solid it has got a slightly boomy sound to it when you jump up and down ( and that I have been in sheer frustration and Interspersed with foul language). Due to being a low energy modern construction having to comply with modern building regs the plaster board ceiling has 200 mm of rockwool battens placed between the wooden joists above it. This means on the old clapping test, that the room is far from bright. However on the flip side the room has proven not to be able to absorb the pressurisation/ energy induced by my much loved Fact 12s this resulting in too much bass/ bass boom.

Beyond head scratching and throwing all my the toys out of the pram the last couple of weeks has given me an opportunity to read up on room acoustics and watching a lot of Dennis Foley’s videos @acousticfields.com. These very instructive short videos explain the fundamentals of room and speaker interactions and have helped me a lot to in simple terms understand, what is at play in there. Essentially you don’t have to have an engineer’s grey matter to understand, what he is talking about ( suits my little brain to a T).

Initially the sound was literally so bad I could have cried and ‘ The baby nearly got flushed out with the bath water’. Reassurance was on tap from other Forum members and Foottapper’s suggestion as to using Gaia III feet elimated 90% of the vibrational influency from the floor.

An intesting observation was the influence ( or not as case ended up) of the bifolds behind my speakers. I essentially opened them right up to work out at what point the didn’t interfere with the sound from my speakers. This probably also helped by the Facts being front ported turned out to be 1.5 metres in front ( measured from front baffle) and at that point the sound evened out to also produce a fine tonal balance.

I initially used Dynaudio’s rule of thumb namely dividing the room square up into 1/5th as a good starting point. Ie neither listening position nor speaker placement baffle to rear wall and side wall within the 1/5th distance. This will ofcourse mostly apply to a floorstanding design I guess. It certainly worked well. My speakers drive unit- centre is now an equal 85 centimetres from the side walls and with a 20mm toe in to establish a good soundstage but also to reduce the impact of first reflection. In addition I am using 2 GIK Acoustics panels to dampen first reflection. These particular panels both diffuse and absorb at the same time to prevent mid/ highs loosing too much energy and actually even look nice.

I have managed to even out most of the excessive bass by literally standing in each corner of the room and listen ( YEP, bonkers hobby this ????). I found that keeping absorption at the listening end to a minimum served the mids and highs not to loose energy- otherwise the overall sound became flat and lifeless without sparkle. This could ofcourse in most cases be achieved by soft furnishings/ curtains etc. In my case this wasn’t sufficient, so I installed 2 GIK Acoustics tri trap absorbers on top of each other in the corner behind my listening position, but strangely only had to do one corner- use your ears for where you feel the need to reduce bass lift/ boom. ( even get down on your knees and listen- at that point my dear wifey was beginning to fear for my sanity admittedly) 

The speaker end ( dead end rule literally ruled in my case) proved to be the biggest challenge beyond moving the speakers round first in big chunks ( a foot) and then in the end literally millimetres at a time. Once you find the best compromise in the first place for placement, make sure to mark that point say with decorators tape on the floor. It is actually difficult to remember how something sounded half an hour ago. I found that I always managed to end up at that same place again, so there must be something to be said for going with your first instinct

After having achieved a very good bass reponse for both overall resolution and decay a 100 hz mode decided to pop its ugly face up. This is a distinct lift from the deepest string on an acoustic guitar or say a double bass being played. To this effect I have experimented with Hofa bass traps from Thomann.de. These are specifically tuned to be most efficient at around a 100 Hz and after having realised , that they need to fill in the corner full height floor to ceiling stacked on top of each other to work properly, I am now receiving 2 more on Monday. They are very heavy and slightly cumbersome, but really work in the speaker end corners ( Check my profile for pics). 

I shall report back next week, but after the last couple of weeks desperately hoping that I would achieve the best compromise with my somewhat tricky new listening space , I really hope it will be mission accomplished . It would great if you guys would chip in, so maybe we can tap into each others experiences with the for most of us tricky speaker/ room interaction and viable solutions. In my case I never imagined it would so difficult to eke out my issues, but on the flip side I am chuffed to have persevered instead of of chucking my speakers in the skip.

Thanks Fellas ???????? ATB Peter

Posted on: 15 December 2018 by tonym

Great to see you back on the forum old bean!

I can't add much to what is undoubtedly a very interesting subject, and you've obviously gone about this problem in your usual careful way. The only comment I can make is regarding the effectiveness of good speaker isolation with decent speaker/ floor interfaces. My room's pretty benign accoustically, and the floor's pretty inert, but I found supporting my DBLs on decent stands made a really surprising improvement to bass and clarity. Something I think folk should try, even if they believe their speakers won't benefit.

Posted on: 15 December 2018 by wenger2015

It sounds like you have an made a lot of progress.....

another Forum member ‘Finkfan ‘ has sorted out his office/listening room extremely well, using GIK bass traps and side reflection panels...I have benefitted greatly from his experience. 

GIK offer great advice on ones rooms acoustics.... and it’s a free service which is a bonus 

I am in the process of sorting out my own room.....so far, all I can say is the panels work extremely well.....

ps you have a great looking room 

Posted on: 15 December 2018 by Bart

Our listening room is of very similar dimensions.  I bought a little calibrated microphone and a long cable and plan on trying out the (free) REW analysis software. I dont know what I'll learn - or if I'll even be able to interpret what I learn - but all-in it was very little money for the mic and cable.

Have you thought about playing with room analytics software?

Posted on: 15 December 2018 by 911gt3r

Hi Bart, like yourself I would not trust myself interpreting the read outs sufficiently and maybe more to the point once you have had the speakers playing in all possible locations in your room, you still need to apply the correct remedial solution. I do think it takes a lot of experimentation and foremost to trust your own ears.

ATB Peter

PS. Am following your journey with The Magicos with great interest

Posted on: 15 December 2018 by Bart
911gt3r posted:

Hi Bart, like yourself I would not trust myself interpreting the read outs sufficiently and maybe more to the point once you have had the speakers playing in all possible locations in your room, you still need to apply the correct remedial solution. I do think it takes a lot of experimentation and foremost to trust your own ears.

ATB Peter

PS. Am following your journey with The Magicos with great interest

We're pretty sold on the Magicos . . . but I want to give Kudos 606's a fair shot.

Posted on: 15 December 2018 by 911gt3r

Hi Tony ????????

What type of stands did you use? Townends platforms produced in a ship yard ? Tehe ATB Peter

Posted on: 15 December 2018 by 911gt3r

Undiminished bass across  the full spectrum dictates that the speakers be large - whether standmont (e.g. PMC MB2) or a floorstander. Small speakers inherently are more limited, though they can give the illusion of doing bass because you hear the harmonics - but that is not the same.

However, the construction cost of a big speaker capable of doing low bass well is considerably more than that of a small speaker not going so low, therefore it can be harder to find affordable good full range speakers than standmounts with limited bass. 

A common way to reduce adverse bass effects in smaller rooms is to curtail the bass output by using smaller speakers - yes, it can avoid a resonance by not ‘exciting’ it - that particular note might even be at about the right level because if the boost of that frequency by the room - but the notes either side, and particularly lower, would be much diminished, which is acceptable to some people but not to others. A better option is to use the speaker and listening positions to minimise adverse room interactions, and use room treatments to reduce room problems. But both of these can be tricky in real life room situations where the room is not a dedicated listening room. At the end of the day it is a matter of finding the best compromise that best suits your expectation of musical performance, budget and domestic acceotability.

In my view, if the full range of bass undiminished is important to you for best enjoyment of music then floirstanders could be better - but be prepared to have to pay more, and to expend more effort in optimising the room set up. I would only consider secondhand, possibly including ex-dem,  as it gives far greater buying power,  and for me transmission line designs have been the only ones that I have found pleasing in the way they present the bass.

Allow me too add Innocent Bystanders insightful post from another thread. ATB Peter

Posted on: 15 December 2018 by 911gt3r

Unfortunately if you use full range speakers that generate significant output at 34Hz (>-6dB) firing down the length of a 5mx3m room,you willget a significant bass peak at that frequency (and you'll also get a somewhat smaller bass peak at 57Hz).  The only way to stop this occurring with full range speakers would be to position the speakers right in the middle of the room - this situation is dictated by simple physics and is unavoidable.  A substantial amount of bass trapping will reduce the peak considerably, but it won't eliminate it.


In smaller rooms there are only two viable way to get an even bass response.

The first is a complete DSP based 'room correction' system, and then you need to accept the sound quality alterations that are also imposed by the wideband processing of such systems.

The second is to use main speakers with curtailed bass response and then use a correctly aligned sub to fill in the bass peaks, using notch filters to prevent the output form excessively exciting the resonance modes.  This approach also requires that substantial bass response damping is achieved in the room - usually this requires loads of bass traps.

Equally a clever comment from Huge from the same thread “ Should I discount Floorstanders”. ATB Peter

May I add that even with a different speaker design it invariably will never sound better, than the room allows it to. Ie room treatment will be the way forward , if you want to enjoy both the highs and the very lows to get the full musical experience as it was intended.

ATB Peter

 

Posted on: 15 December 2018 by benjy

Hate to say it, but as "dedicated" as many of us are about ideal sound, even with the inclusion of a dedicated room, many of us do not have the "luxury" of adapting a room to ideal sound. Compromises must be made , to fit life-style, to fit furniture, to fit decor..... etc. While my own layout works reasonably well, I have no doubt it could be made to work better, with judicious room treatment. Not an option in my case, so I ignore the subject - sort of like probing at a cavity. 

Posted on: 15 December 2018 by 911gt3r

Hi Benjy.

As a dentist I would always suggest to get a cavity looked at tbh. ????

ATB Peter

Posted on: 15 December 2018 by Huge

A good place to start when planning a listening room layout is to use the 'Room Simulation' dialogue of REW.

For those people who have invested in a calibrated mike (e.g. the miniDSP UMIK-1) and connected this to REW (@Bart ?), I do understand the various graphs it generates and I can usually comment on the results if people post them on the forum.

This is the approach I used to achieve a +/-4dB response in a fairly difficult room: it's almost square at 3.75m x 4.25m.

Posted on: 15 December 2018 by Loki

Is this the best room? When we move House the first room to be earmarked is the listening room, and it's acoustic properties are the starting point for negotiation.

Posted on: 15 December 2018 by Bart
Huge posted:

A good place to start when planning a listening room layout is to use the 'Room Simulation' dialogue of REW.

For those people who have invested in a calibrated mike (e.g. the miniDSP UMIK-1) and connected this to REW (@Bart ?), I do understand the various graphs it generates and I can usually comment on the results if people post them on the forum.

This is the approach I used to achieve a +/-4dB response in a fairly difficult room: it's almost square at 3.75m x 4.25m.

Thanks Huge -- I'll take you up on your offer once I run through it.  I did buy the miniDSP mic, and have downloaded REW for my Macbook.

Posted on: 16 December 2018 by 911gt3r

Hi Huge.

A 4+/- DB response is very impressive. What did you do room treatment wise? ATB Peter

Posted on: 16 December 2018 by Foot tapper

Hi Peter and welcome back to the forum; we missed you.

I have an XTZ Sound Room Analyser II Pro measurement kit which captures, analyses and explains a great deal about the acoustics of your room.  All it needs is a windows PC/Laptop to plug into.  If you would like to borrow it to have a play, let me know.

If you search for "XTZ Sound Room Analyzer II Pro Measurement Kit Review" on a well known internet search engine, you can read a review all about it.  I ran it several times, then emailed the results off to someone who understands these things and they came back with excellent, effective recommendations on how to improve the speaker-room interaction.

p.s. Like the yellow car.  It's almost as nice as an Evora 400 .

Best regards, FT

Posted on: 16 December 2018 by 911gt3r

Bless you FT. I will give you call. Yes I LOVE the GT4 and had to wait some time to find the colour I wanted. It is an amazing car to drive, however I still haven’t got under 9 minutes round the Green Hell- need to grow some bigger balls me thinks. Possibly didn’t do much for the confidence, when the group went out in June getting hit by another Porsche on my outlap???? The bugger undertook me on a sharp left hander- thankfully the car was driveable to complete the rest of the day just looked a bit sad. All fixed now thankfully ????????

Speak soon Peter

Posted on: 16 December 2018 by Huge
911gt3r posted:

Hi Huge.

A 4+/- DB response is very impressive. What did you do room treatment wise? ATB Peter

The room treatment consists of 12 blocks of mineral wool in acoustically sealed packs 1.2m x 0.45m x 0.45m with the mineral wool encapsulated in a lightweight membrane.  Their main absorption band is 30Hz to 300Hz, trailing off to a fairly low absorption at 500Hz and almost no absorption over 800Hz.  This approximately matches the LF rise in the measured RT60 time for my room when tested with no acoustic treatment.

Used alone they reduced a measured 'in room' +24dB @ 40-41Hz response peak to about +12dB.  The +12dB is then reduced to about +3dB using a notch filter in the DSP that controls the crossover to the sub.  Trying to reduce it below +3dB is counter productive as it causes too many compromises to other aspects of SQ or of practicality.

Posted on: 16 December 2018 by Sloop John B

“Home audio fidelity” get very positive reports over on the Roon forum. From my reading of such posts sorting major issues physically and then tweaking with DSP gives the best solutions. 

Basically you measure your room, send the results to HAF and they create a DSP curve which can be added to various software (including Roon). 

.sjb

Posted on: 16 December 2018 by 911gt3r

GULP Huge, that’s a lot of volume. Did you end up with 2 stacked in each corner and the rest behind the speakers? Out of curiosity what speakers do you have? ATB Peter

Posted on: 16 December 2018 by Simon-in-Suffolk

Of course eq and DSP filtering  might be the only way to achieve a satisfactory outcome for some setups, but bear in mind you have moved away from being bit perfect (in a big way), and are introducing digital artefacts and digital distortions and so you have deliberately interfered in  the digital / analogue reconstruction chain......  although this is only going to be very noticeable on very revealing DACs like the Hugo and ND555 connected to high end amplification.

From an engineering and SQ perspective if one was going to digitally process the audio.. one ideally should have a higher resolution as possible in the media.. I wouldn’t want to do this ideally on anything less than 94kHz 24 bit media... 

Posted on: 16 December 2018 by Sloop John B
Simon-in-Suffolk posted:

Of course eq and DSP filtering  might be the only way to achieve a satisfactory outcome for some setups, but bear in mind you have moved away from being bit perfect, and are introducing digital artefacts and digital noise and so you have deliberately distorted the digital / analogue reconstruction......  although this is only going to be noticeable on very revealing DACs like the Hugo and ND555 connected to high end amplification.

Don’t most DACs including Hugo and ND555 apply their own DSP anyway?

is this not how Naim “voice” their units?

.sjb

Posted on: 16 December 2018 by Simon-in-Suffolk

DSP is simply a general term for digitally processing sample data. The Hugo (original and TT) and ND555 as part of their reconstruction process within their closed systems provide oversampling and a low pass filter (a key DAC antialiasing and reconstruction function).. but both carefully implement these two functions to be as simple and straightforward as possible for the required performance despite the increased internal resolution so as to avoid artefacts and robbing of detail.

The ND555 low pass filter implemented in the Analog Devices digital system processor (SHARC)  is just a few lines of code, and is the same few lines of code that started with the Naim DAC all those years ago... we know however the firmware is critically tuned to carefully control the execution timing of this code with respect to I/O and memory operations so as to reduce artefacts.. audible here as a change in ‘tonal character’ and ‘detail’

Naim don’t use DSP functions to voice their streamer devices (but do in Muso series) ... it’s the optimisation of digital noise through careful code execution timing in running the low pass filter and other non DSP routines that ‘voices’ the device.

Posted on: 16 December 2018 by Huge
911gt3r posted:

GULP Huge, that’s a lot of volume. Did you end up with 2 stacked in each corner and the rest behind the speakers? Out of curiosity what speakers do you have? ATB Peter

Because of the layout of the room I have 4 in each corner behind the speakers and 4 in one corner on the listening side of the room.

The speakers are Spendor SP2s with modified crossovers, modified internal damping arrangements and with the ports blocked.

Posted on: 16 December 2018 by Huge
Simon-in-Suffolk posted:

Of course eq and DSP filtering  might be the only way to achieve a satisfactory outcome for some setups, but bear in mind you have moved away from being bit perfect (in a big way), and are introducing digital artefacts and digital distortions and so you have deliberately interfered in  the digital / analogue reconstruction chain......  although this is only going to be very noticeable on very revealing DACs like the Hugo and ND555 connected to high end amplification.

From an engineering and SQ perspective if one was going to digitally process the audio.. one ideally should have a higher resolution as possible in the media.. I wouldn’t want to do this ideally on anything less than 94kHz 24 bit media... 

That's exactly why I only use the DSP in the audio chain to the sub-woofer, not in the audio chain the the main speakers (even using 24/192 for both processing and output).

Even with a ND5 XS + Nait XS2 I could clearly hear the degradation of SQ caused by having a DSP in the system, either reprocessing the data in the digital domain before the ND5 rendered it or processing in the analogue feed from the ND5 to the amp (which was even worse).  So the damage is revealed even with much less revealing equipment than 500 class boxes.

Incidentally the phase alignment of the crossover from the mains to the sub is adjusted to within <+/-30° minimum phase.

Posted on: 16 December 2018 by Sloop John B
Simon-in-Suffolk posted:

DSP is simply a general term for digitally processing sample data. The Hugo (original and TT) and ND555 as part of their reconstruction process within their closed systems provide oversampling and a low pass filter (a key DAC antialiasing and reconstruction function).. but both carefully implement these two functions to be as simple and straightforward as possible for the required performance despite the increased internal resolution so as to avoid artefacts and robbing of detail.

The ND555 low pass filter implemented in the Analog Devices digital system processor (SHARC)  is just a few lines of code, and is the same few lines of code that started with the Naim DAC all those years ago... we know however the firmware is critically tuned to carefully control the execution timing of this code with respect to I/O and memory operations so as to reduce artefacts.. audible here as a change in ‘tonal character’ and ‘detail’

Naim don’t use DSP functions to voice their streamer devices (but do in Muso series) ... it’s the optimisation of digital noise through careful code execution timing in running the low pass filter and other non DSP routines that ‘voices’ the device.

Does this mean there are mostly harmless digital artefacts and then more harmful ones?

.sjb