In the ongoing effort to balance intensity with volume levels I made a change to our system processing this week.
Last weekend we got the usual complaints but this time I was told that people said “it sounds great, it’s just too loud.” I’ve never heard “sounds great” as part of a volume complaint so I perked up. It might be because people who complain generally like choirs and we had one last week. Just a guess.
It was interesting to me because I featured the choir and lead vocal prominently and let the band take a little of the back seat. Ironic after all my talk about vocal placement but it just felt right for the morning. To give an idea, measured in dB-A Slow the band even with guitar solo ranged from 85 to 89 at most. At the 9am service I let the choir run up to peaks of 93 with no complaints. At the 11am service I only ran the choir up to 91 and 92 with complaints. In each case I just gave the room what felt right at any moment.
On Monday I gave a critical listen to the raw CD over the system. Truthfully, it’s a little in-your-face in the high-mids especially when it breaks 89dB-A. I can see how it would be uncomfortable for some.
After the jump: Smaart screen shots and audio clips of adjustments made to the system if you’re interested. If you don’t feel like wandering off into the technical weeds, just jump down to the audio clips.
Different Levels are Not Equal
Sounds obvious but it’s not intuitive. Most people think you can just turn it up or down with no more to consider than a volume change. Speakers and rooms don’t sound the same at every volume level. We hear and perceive sound differently at varying levels too. That’s why you dial in your system and your mix at actual listening levels. I believe part of our problem is that we’re riding the line where the sound of our boxes starts to shift slightly and enough to matter – especially in the stadium seating in the direct fire of a side line array.
2.7k
2.7k is a real knife blade of a frequency. It’s the jangle in a Tele/Vox guitar tone that fights with vocals. When prominent in a loud, saturated mix it forces involuntary squints and grimaces from humans who haven’t abused their ear drums. You don’t really want to slice it out of your system because at lower volumes it’s a critical frequency for clarity and vocal intelligibility. With our Nexo line arrays (which I like a lot by the way) I hear a tipping point around 92dB where if I were to do anything I’d make a slight cut centered on that 2.7k. Makes sense because that’s the level where some say it sounds “piercing.”
Multiband Compression
We have a TC Electronics Finalizer in the rack and use it for light multiband compression. It adds a slight polish and provides a margin of grace when a song starts getting up there in volume. There’s a fine balance to strike mixing live with multiband compression. Heavy compression can make you sound slick like radio but it also can suck the life out of your drums and make it all sound like a shiny plastic toy. It keeps your mix honest though because anything out of balance takes over.
I’ve set crossover points before at 1.6k (the passive crossover freq of the Geo boxes we have) and 4k to tame the high-mids but I’ve never been completely stoked on the results. Multiband compression is better suited for sweetening than problem solving.
By the way, you’ll find a decent multiband compressor in the Yamaha M7 internal rack as we’ve discovered at our remote campus.
Dynamic EQ
The Finalizer has a Dynamic EQ insert I finally played with and I really like the results I’m getting. It functions like one band of multiband compression but the concept is slightly different. Rather than set crossover points you set the center frequency and Q like a parametric EQ filter then the threshold, ratio, attack and release like a compressor. As the level rises above the threshold the spectral balance wraps around the curve of the EQ filter relative to its threshold and ratio. The result sounds smoother and more natural than anything I’ve gotten by problem-solving with the multiband compressor. The bite is reduced to a bark and the drums still breathe.
Smaart Examples
I hooked up the Finalizer output to Smaart to demonstrate what the dynamic EQ does. Here are screen shots of what pink noise does at output levels that would create the corresponding dB level in the room. The spectrograph displays about the last 5 seconds of sound. There’s an inherent dip at 4k but that’s something else. Follow the 2.7k area.
85dB is as low as I go for rock and roll. I definitely want linear system response (what comes out of the board is what you hear in the room) at this level because speech and video playback generally happen south of 85. The dynamic EQ doesn’t kick in yet.
88dB is where some of our older folks start to feel uncomfortable but they’ll accept it if it sounds nice. The dynamic EQ kicks in slows down the high-mids. The difference in the room is only slightly noticeable.
91dB as an average is the threshold of many complaints but as a peak level we’re still ok. At this peak level the high-mid reduction is more noticeable but welcome because these are passing moments.
93dB – this is just an example of what happens when you crank it up. Honestly, it doesn’t sound as bad as it looks. It’s quite smooth and leans hi-fi.
Audio Examples
Here’s a section of one of the most moving moments during worship last Sunday at the loudest point.
Raw
This one is raw straight off the console. In my in-ear molds this sounds close enough to what happens in the room. You have to turn it up till it’s annoying to get the full effect and to compare it with the other clips.
Dynamic EQ
This one is the same clip run through the TC dynamic EQ at levels that would average 91dB in the room. It’s smoother. The 500Hz in the vocal seems to pop out a bit more but, again, these moments are temporal, not constant.
Dynamic EQ with Multiband Compression
This one adds the multiband compressor after the dynamic EQ. It evens out the boxiness of the 500Hz in the previous clip and leaves the top nice and shiny. It’s pretty heavy for my taste (it’s just an example) but I’d actually be good with this if it keeps me at acceptable levels. The intensity is all there.
An Artistic Filter
The thing to keep in mind when using this stuff is that it’s an artistic filter used to achieve an effect. It’s not much different than a photographer using a lens filter to manipulate a photograph. The photo is not as much the real thing as it is an artistic interpretation. If you forget this when mixing with dynamic EQ and multiband compression you’ll force tones to sound original again and end up with a really mangled sound. (Personal experience.) I like to get tones and check the band for a good bit with the TC bypassed then drop it in after the mix sounds solid.
Running at these volumes I am perfectly comfortable putting a filter on things. I don’t suppose I’m audio purist necessarily because I’m driven to achieve a larger-than-life sound within the parameters I’m given. At the end of the day part of what we do is create an illusion. It’s all smoke and mirrors.
Entries (RSS)
hey Sam its Caleb, I was wondering if one of these days you can show me how to use those compressors in the event center? I run sound sunday nights and one of the compressors is pretty weird and the last few weeks Ive been running vocals with no compression and just using to faders to control them. Any advice or help would be awesome!
Here is a resource related to the different levels are not equal comment.
http://en.wikipedia.org/wiki/Fletcher%E2%80%93Munson_curves
This is a really interesting effect, I wonder if its possible to build this
curve into a master level slider…. could be interesting.
Greeties from Bozeman! Set up a makeshift studio here with my KRK Rokit 5s and ABC’d the three clips, then FELT the wooden tabletop vibrating under the nearfields! Oops. Had to isolate them from the wood and then go back for a few more passes. The multiband compression does bring out the low end as noticeably as the high end info. One need only listen to a 57 on a snare before and after EQ to concede that yes, this is totally an illusion! (Except for a certain snare, a 10″ Pearl that my drummer, Steph, brought to Venezuela from England – I totally promise you that drum coudn’t be improved by any EQ I could do to it. It must have been having an out-of-the-shell experience.)
Yeah Caleb. For sure. The Event Center is one of the worst rooms on the planet.
The trick I use in there is inserting a compressor on a subgroup then slamming the vocals into the subgroup (not the L/R) to get that wall of sound. An advantage to using compression on subgroups instead of inserted on individual channels is that, since the monitor aux sends on the Midas are post-insert, post-eq, you’re not compressing the vocal sends to the wedges. Compressed vocals in monitor wedges is a lot of trouble especially in that room. As the vocalist, if you don’t hear your full dynamic range you’ll always ask for more, not feel confident, and then feedback issues ensue.
Awesome Sam thanks! Im definately going to try that tomorrow night! What do you say to coffee at some point tomorrow? short notice but if you have time, or another day works too!