Considering Room Anomalies in Speaker Design

Hari Iyer

Well-Known Member
Joined
Feb 8, 2010
Messages
3,822
Points
113
Location
Mumbai
Over the past six months I have been working with different drivers with my Transmission Line floor standing speakers and have observed certain parameters to be of important consideration when designing the speakers. Last month I had the motivation (thanks to FM Bhagwan) and the opportunity to measure my speakers in my room and observe various parameters and anomalies of working only with the crossover simulator. No doubt the simulator is the starting point in your design, but IMO it displays more of the electrical response of the speaker system rather than the acoustic response. The acoustic response involve primarily the room and the baffle and they together constitute the speaker system in addition to the drivers and the crossover.
Below is the simulated response of my speaker system without considering the room and the baffle step (in my case also the wave guide for the tweeter).
2z8viox.png


As you can see the perfect straight line as one would like to expect from any good sounding speaker.
But this is what I got when I measured them.
20rkcua.jpg


After due consideration I concluded that the issue lies with many parameters of my room, the most significant being the room resonance, room reflections, the baffle step and the wave guide for the tweeter.
This is what I designed to compensate for these anomalies,
2nulmo4.png


As you again observe that this curve looks a bit weird and is no more a straight line but has several dips and unbalance in the mid and high-frequencies.
But this is how it measures in my room
mlpp46.jpg


As you can see its more straight line with all the high energies of the drivers disappearing. Also I get a very good RT60 indicating a very good dispersion, and room decay. Also the RT60 indicates even energy in the rooms and no sweet spot in particular. The imaging too is razor sharp in the centre when I play pink noise.
2w70efs.jpg


Below is the impulse and step response of the speaker system.

2wci1j5.jpg


vqty76.jpg



Waterfall Plot:
2nl5x8z.jpg


Note: The ridges in the mid-frequency is because of either the room anamolies or the microphone factor as I get them even without the speakers been tested.

Subjective listening test reveal a very well controlled mids with less clutter and overhang in the mids. Highs are extremely mellow, with no energy storage or ringing making you sometime wonder if the tweeter is playing or not. This setup gave me more soulful play of music than the one which I had earlier.

This makes me wonder how does commercial speakers take care of the room factor or does speakers intended to play in smaller rooms play well in larger rooms or vice-versa. I can recollect FM mpw Merlin speakers how they were a bit suffering in low frequency energy and sometimes highs too even though the speakers themselves were quite good. In this sense DIY has an upper edge as you can factor that part in the cross-over design itself and built the speaker to your room. FMs who have heard my speakers last month may have found the extra energy build up in my speakers particularly in the mid and high frequencies and actually will sound very harsh if not corrected either with room treatment or in the speaker design.

With this new observation will definitely have a paradigm shift in my future speaker designs where the most important question I will ask is where will the loudspeaker be placed with some images of the room itself. A 20% to 30% variation in room size will not matter much but not more. Ie. The speaker designed for a room size of 200 sq.ft can work well in 150 sq.ft to 250sq.ft but will suffer with energy build up if placed in a 100sq.ft room or can face loss of energy if placed in 400 sq.ft room.
 
Last edited:
After due consideration I concluded that the issue lies with many parameters of my room, the most significant being the room resonance, room reflections, the baffle step and the wave guide for the tweeter.
This is what I designed to compensate for these anomalies

Hi Hari,

Can you please elaborate on how you went about fixing the Anomalies?

Thanks,
Venki
 
@Venki
I usually design with a crossover simulator and implement them. I use measurements only to check there are no major deviation from the simulated values. The impedance and phase measurements helps in doing this check rather than the FR measurements. Since FR measurements are both electrical and acoustic related they cannot be simulated and measured accurately as FR is also a function of the room.

IMO for the final tweak and getting the perfect balance between low, mids and high its still the human ear which is the best judge. Unfortunately not many of us can be very good at doing this judgement and you will need to depend on others to do this voicing for you and those who are experienced in hearing live music and many high-end speakers can be the right candidates. You can call FMs to your home to help out with this final tweak after you are satisfied with your simulations and measurements. Will post later my final tweak simulated and measured results for my TL FS.

Cheers,
 
And this is what i finally settled for my speakers and room considerations. Baffle step is around 2dB and room contour around 6dB.

Response with correct phasing

15gzhna.png


Tweeter Reverse phasing

2ag9kls.png


Woofer and Tweeter Reverse phasing

1zcobag.png


Measured Impedance and Phase

192uki.jpg


The critical crossover mod was the accurate tuning of the filter phase along with the baffle step and room contour consideration. The result is clearly obvious in the final outcome with improved dynamics and uncluttered mid-range and highs. Finally satisfied with the outcome.
 
Last edited:
For excellent sound that won't break the bank, the 5 Star Award Winning Wharfedale Diamond 12.1 Bookshelf Speakers is the one to consider!
Back
Top