Help wanted using room measurement to setup dual subwoofers

[napuli]

Hi All,

I am experimenting with using room measurements to integrate dual subwoofers into my stereo setup. I am using REW, Dayton UMM-6, and a microphone stand. The UMM-6 was set at the listening position pointing down. My main speakers are Vandersteen 2ce Signature II couple to the floor with spikes with tilt angle adjusted to my ear height as recommenced by Vandersteen. My room is long and narrow with no back wall. It leads into an open floor dinning room, kitchen, and family room. The front wall has a big window and one side wall has a big glass sliding door. I have 4 GIK acoustics corner bass traps in the front and 2 GIK acoustics bass traps in the side walls flanking the main speakers. The subs are dual REL T9i connected to the speaker terminals of a Cary SLI 80 tube amp. Room is hard to define because back wall opens up Ito the rest of the house but it is about 23 x 11 x 8 ft.

I have experimented with various sub positions in the room. I am not fix on their current position but for now I would like to leave this variable out of the equation. The subs are place diagonally in opposite corners of the room. My goal is to learn how to use room measurements to help me adjust the level, crossover, and phase with the subs in their current location. Once I get them dial in as good as I can I will try other locations and repeat the same process.

Through out this process I also want to pay attention to see if I can actually hear differences when making acoustic corrections using room measurements. To me, this means listening to music for several days and playing different tracks.

Ok...sorry for the long intro. Please see the attached files. The blue traces show the main speaker frequency response without the subs. Red is the main speakers and both subs with rear and front sub phases set to 0. Purple is the same but with rear sub phase set to 180 and front sub phase set to 0. What I like about these results is that I see a drastic change when I switch the phase in the rear sub from 0 to 180. I am trying to decide which is better? I think 180 shows both subs in phase which makes sense considering their physical orientation. Also, the downward slope from 30 to 10 Hz seems to better mach the specs of the subs. When out of phase, the frequency response looks a little more flat, and a possible resonance peak at 15 Hz is reduce by about 5 dB.

What is disappointing, is that when the subs are in phase they don't seem to add anything new to the mains other than an increase in volume. Looking forward to your feedback.

 

[detlev24]

Hello!

This is a very complex topic... So, let us start from the beginning.

I guess you know how to take accurate measurements with REW, using your UMM-6 microphone? Did you load the specific calibration file for your UMM-6 in REW and are you aware, for use in which orientation it was calibrated [horizontal, or vertical]? // It should be a similar procedure to the "UMIK-1 setup with REW".

What are you trying to accomplish with those two 'T/9i' subwoofers? Their low frequency extension, by manufacturer specifications, seems to be very similar to what the 'Model 2Ce Signature II' shows: -6 dB at 28 Hz [REL] vs. ±3 dB at 29 Hz [Vandersteen]. However, this in not what to expect in-room, as "room gain" [= modal peaks] will typically allow for a lower audible frequency response.

Do you aim for a crossover of the subwoofers with your mains - or wish to run all (4) loudspeakers in parallel and at their respective full-range?

 

[napuli]

Thanks for the reply!

I did load the calibration file for the UMM-6 but there is no info about horizontal or vertical orientation for the calibration file. I don't really have a target for the REL subwoofers. My hope was to integrate them with the mains to fill some of the frequency nulls at the listening position . I agree there is not much difference in extension between the REL and Vandersteen. However, you can hear a significant difference with the subs on vs off. Obviously, I am boosting the lower frequencies by about 5 db or higher. In addition, the room sounds more spacious with the subs on. I am guessing this maybe a result of the subs adding more room reflections but I don't really know.

 

[detlev24]

Great! This information helps a lot.

Regarding the UMM-6: I checked one calibration file and the corrections between 10 Hz and 300 Hz, which I would consider the range of interest, are not too big for whatever plane the calibration has been done [all are within roughly +0.3 dB and +0.7 dB] - so, no significant impact on accuracy should be given.

Yes, total output will be higher, due to multiple sound sources emitting the same frequencies simultaneously and in a relatively close proximity to each other [considering the wavelengths at low frequencies; to keep it simple]. For instance, on the upper graph's blue line, a 20 Hz tone (most probably) cannot be heard, as it is roughly at or below the audible threshold (see link in previous post) - and further masking during music playback might occur. On the red line instead, you have significantly more output from 20 Hz to ~75 Hz, which then certainly gets audible down to 20 Hz.

One of the best options you have 'to fill some of the frequency nulls at the listening position' is with a DSP that accepts "biquad" data. You simply have too many degrees of freedom to cover [gains/delays/EQs etc., for each channel]; and additionally, not enough settings on your gear to try out at least a few combinations or "wild guesses". I would suggest something like the miniDSP 2x4 series; also see "Appendix A. Notes on using an external 2x4 for the subs". [A software upgrade of the '2x4 HD' is available to add Dirac Live room correction on top of MSO, if desired at a later stage. Else, the DDRC-24 would be readily upgraded.] Just to provide an idea: This diagram shows a possible wiring.

[Of course, there also exist limitations of multiple subwoofers and EQ.]

Multi-Sub Optimizer (MSO) would calculate all filters according to the measurements you provided; and output them as biquad TXT-files, which then would be loaded in such a DSP [usually*, all subwoofers in AVR or LFE-configuration, to bypass their internal signal processing]. One example of application can be found here. Just note, that in this specific case I did correct for an LFE-channel exclusively [which is a little less complex to set-up]. In your case, you would want to integrate the subwoofers with your main loudspeakers, which is possible with MSO likewise. Either, by running your mains at their full-range or, by letting MSO calculate the correct crossovers with your subwoofers.

In any case, a proper subwoofer placement vs. the arrangement of your listening position(s) - all specific to your room - is the basis for success.

Have fun!


*The REL Ti series has the standard "120 Hz fourth-order filter" applied to the LFE-input; which according to the manual cannot be bypassed.