SMAART Learning: Magnitude traces from room tuning - 1

P

Phil Graham

Honorary PhD
Mar 10, 2011
651
1
18
Atlanta, GA
Hello all,

I've been extremely busy as of late, and that has resulted in several things around SFN falling by the wayside. I haven't been updating the "how to mix blog" and didn't finish the how to interpret measurement" thread. I also didn't really finish my thoughts on non "flat" tunings.

One that that was mentioned previously was asking for some before and after traces of some measurements in rooms. I recently had a client who doesn't care about having such things shared, and whose desired voicing I also enjoyed, so I thought I would post the measurements here. I don't have any time at the moment to elaborate on these plots, but hopefully they will stand alone for the time being, and be instructive for some. If you post questions about these plots, I eventually will respond, but it may be in 2012!

Please don't think that there is anything magical about this particular tuning curve. This is simply the result of measurements guiding the tuning for each zone, and then 45min of listening to various playback sources in the space and tweaking things by ear to mine and the clients' satisfaction. Also, for a volume reference, "loud" for this church is 90dBA slow, so the tuning was balanced with this playback level in mind.

Most of my (primarily church) tuning clients are either unsure about their integrator's prowess in system commissioning, or are unhappy with their current system optimization. I find that the period of listening with the client's relevant people in attendance is key to everyone's happiness. Sometimes that results in curves that deviate a great deal from what I would prefer, but in this case I felt the client and I were completely on the same page with the system voicing. The music director of this facility is an experienced studio engineer with a good ear, and I think that certainly that helped on our agreement.

Measurement details:
  1. This room is a church. We re-aimed and re-tuned their entire system with the new DSP they purchased.
  2. The PA is configured in 4 zones L+R, L, R, and L+R. So, stereo in the middle and mono on the extreme side wings.
  3. The measurements below are a 12 location, coherence weighted average of the R-R zone. That means the house right speakers on the house right side of the center church aisle. All other zones of the PA were muted for these measurements.
  4. The measurements were taking in SMAART 7.3.x using MTW for windowing and two second averaging.
  5. I've not included the phase measurements, as they have little meaning for these kinds of averages.
  6. Turquoise trace is the "before" measurements, and olive trace is the "after."
  7. Before trace has a 125Hz 4th order BW high pass engaged, so that explains the LF definciency. Subs are muted.
  8. After trace has a 2nd order Bessel high pass at 125Hz. Subs are muted.
  9. Room has all chairs, but no people, in it for these measurements.
Before and After Magnitude: Twelve locations each, Coherence Weighted, 1/12 Octave Smoothing, MTW windowing

Trinty Avg Tops Only_small.png
 
Last edited:
Re: SMAART Learning: Magnitude traces from room tuning - 2

The plot below is the same as the previous with the subwoofers engaged. This room has subwoofers on the floor, too far apart, on either side of the stage. The sub response is not great, but an unavoidable consequence of the location compromises. The music here is not subwoofer driven, so it will serve their purposes well.

Before and After Magnitude: Twelve locations each, Coherence Weighted, 1/12 Octave Smoothing, MTW windowing, Subwoofers on

Trinty Avg With Subs_small.png
 
Last edited:
Re: SMAART Learning: Magnitude traces from room tuning - 2

Could I ask a question about the sub energy shown in that last trace? How common is that large bump in the sub magnitude? From the trace it looks excessive but in tuning my own venue I tended to set the subs (on their own bus on the console) output gain up about +4 to 6dB. At least that's what the trace tells me.
 
Last edited:
Re: SMAART Learning: Magnitude traces from room tuning - 1

Phil,

How purposeful are the slight dips at 150 to 300 and at 1.25k and the boost centered at 2.5k? Are those voicing choices or are they just variations that are inconsequential and were just part of the tuning outcome that wasn't purposely implemented?

Thanks,
Loren
 
Last edited:
Re: SMAART Learning: Magnitude traces from room tuning - 1

Loren Jones said:
Phil,

How purposeful are the slight dips at 150 to 300 and at 1.25k and the boost centered at 2.5k? Are those voicing choices or are they just variations that represent are inconsequential and were just part of the tuning outcome that wasn't purposely implemented?

Thanks,
Loren
Click to expand...

Loren,

This is an average of a lot of measurements, and I don't remember all the gory details. Certainly there are no boosted eq bands in use.

The octave spaced dips were most likely remnants of the floor bounce that showed up across many measurements. The 1.2k dip was the consequence of the lobing of the horizontally hung boxes near the XO point. I'm pretty sure the slight 2k5 boost was a voicing decision that was come upon by not cutting as much 2k5 as I might have otherwise.

-Phil
 
Re: SMAART Learning: Magnitude traces from room tuning - 2

Marsellus Fariss said:
Could I ask a question about the sub energy shown in that last trace? How common is that large bump in the sub magnitude? From the trace it looks excessive but in tuning my own venue I tended to set the subs (on their own bus on the console) output gain up about +4 to 6dB. At least that's what the trace tells me.
Click to expand...

Marsellus,

I would take the subwoofer response of these plots with a grain of salt. First, there were two separate subwoofers spaced far apart, and these measurements were taken at multiple locations that were assuredly in and out of lobes. My summary report to the venue shows 2 out of band eqs cuts on the sub, 1 in band cut, a 4th order lowpass, a 1st order allpass, and a 4th order highpass.

I would say that in these church environments where the levels are typically 90dBA or less, 6-12dB more sub than tops is typical. I have certainly tuned systems with more LF haystack than this, and with less.
 
Re: SMAART Learning: Magnitude traces from room tuning - 2

Phil Graham said:
I would say that in these church environments where the levels are typically 90dBA or less, 6-12dB more sub than tops is typical. I have certainly tuned systems with more LF haystack than this, and with less.
Click to expand...

I would say that 6dB of sub over the energy at 200Hz is the least I have done, and what most people would probably consider "flat". 9dB is more typical, 12dB for more sub heavy genres and many modern rock acts.
 
Re: SMAART Learning: Magnitude traces from room tuning - 1

Thanks Phil,

That makes sense. I was just curious if the slight dip in the midbass was on purpose as that is a frequency range that I have sometimes thought of as making things sound muddy if there is too much energy there.

Take care,
Loren
 
Re: SMAART Learning: Magnitude traces from room tuning - 2

Bennett Prescott said:
I would say that 6dB of sub over the energy at 200Hz is the least I have done, and what most people would probably consider "flat". 9dB is more typical, 12dB for more sub heavy genres and many modern rock acts.
Click to expand...


Bennett, your saying that these +6-12dB humps of sub over the relative energy at 2000hz in the transfer is actually close to flat, or that people simply prefer a rig tuned this way?
 
Re: SMAART Learning: Magnitude traces from room tuning - 1

Bennett Prescott said:
I am saying people prefer a rig tuned that way, and will call a system with as much as 6dB additional sub energy "flat".
Click to expand...


Thanks. I had come to the same conclusion myself but wanted another opinion. I had my sub energy tuned "flat" relative but always had myself and every other engineer push the sub bus up 5dB every show.
 
Re: SMAART Learning: Magnitude traces from room tuning - 1

Marsellus Fariss said:
Thanks. I had come to the same conclusion myself but wanted another opinion. I had my sub energy tuned "flat" relative but always had myself and every other engineer push the sub bus up 5dB every show.
Click to expand...

Marsellus,

Its important to remember that "flat" for most of general populace is a pair of stereo speakers in a smallish room. If these speakers were balanced to a flat line in an anechoic or time-windowed environment, then the response in said room will generally have an upward tilt in the lows and low mids. This might be a part of the 4-5dB haystack sounding "flat" to a broad range of people and mixers.
 
Re: SMAART Learning: Magnitude traces from room tuning - 2

Marsellus Fariss said:
Bennett, your saying that these +6-12dB humps of sub over the relative energy at 2000hz in the transfer is actually close to flat, or that people simply prefer a rig tuned this way?
Click to expand...

There is another thread where I posted a live average. You can find it searching my posts,...there aren't that many. That's pretty close to how I walk away from my tuning job with everything on. The haystack sounds cool but it's got lots of issues. If you are mixing a band and have a real desk you don't need it. That's what those EQ knobs on each channel are for. I do prefer the subs on a separate send but I do not use the send as an EQ knob. Once it's tuned it's set and forget.

Now when I am doing corpie gigs and I only have playback sources going to subs, and likely do not have a decent desk, I haystack the subs depending on the situation (room, pa). I will say some of those guys that haystack the subs (and if ((when)) I did it, it would be 12-24dB) do not have any more energy coming out of the PA at those frequencies than I do, they just did it at a different place in the path between mic and speaker. My board recordings sound better. My console reflects my mix. If I have a neutral PA day after day my mix will translate from one brand to the next. I have made my living for many years now by making pleasant sounding PAs. Since the subs are on an AUX, any engineer may choose to goose them and haystack the rig. Part of the optimization is to make it harder for them to mess the crossover up when they do that. They are happy so I keep working. Now back to mixing.....
 
Re: SMAART Learning: Magnitude traces from room tuning - 2

In general we're all used to a LF boost of sorts, either from being in a small room with home stereos or with the usual buildup from any kind of speaker array. I generally find it hard to tune subs directly from SMAART because all sorts of other weirdness shows up at sub frequencies. Even with the subs muted the magnitude trace will show a lot of junk rumble from air handling or whatever.

I'm just amazed at that red coherency trace, though. That must have been a nice room to work in.
 
Re: SMAART Learning: Magnitude traces from room tuning - 2

Scott Helmke said:
...magnitude trace will show a lot of junk rumble from air handling or whatever.

I'm just amazed at that red coherency trace, though. That must have been a nice room to work in.
Click to expand...

Scott, if you see the original post, this was a coherence weighted average of 12 measurements in one area of the audience. There's nothing particularly good about this room, in fact it is rather unremarkable.

The good coherence here is the result of careful measurements, then averaged in a useful way inside of smaart7.
 
Top Bottom