Have PA, Will Travel: A weekend working for a PA hire company

February 27th, 2011 No comments
PA speakers in the van

PA speakers in the van

Recently I spent a weekend driving around and setting up a small PA system for a friend of mine. This friend, who’s name is Jon, owns the equipment and rents it out for functions, parties, conferences and the like.

That weekend the rig was booked out first for a wedding and then for a garden party the following day. Jon was doing some live sound work elsewhere, so he asked me to step in and take care of the jobs.

Having set this kind of thing up before, both in college and at open mic nights, I felt relatively confident and accepted the offer.

On Tuesday afternoon that week I headed over to Jon’s so he could show me the gear and check I wasn’t going to have any problems. The main concern was that the garden party client had also booked a rack of graphic equalizers with which to “ring out” the PA.

Ringing Out

Ringing out is a process in which an engineer uses an equalizer to notch out troublesome frequencies in an acoustic environment which cause the system to feedback as the gain is pushed up. This is a particularly common practice with stage monitors, which are quite susceptible to feeding back due to the microphones onstage. The process can also be applied to FOH speakers though, which can suffer feedback issues, especially in smaller venues.

The EQ rack was comprised of a pair of dbx 1231 graphic equalizer units. Each of these units provides 2 channels of 31 band equalization, with the bands on each channel switchable between +/-6dB and +/-15dB of boost/cut.

The front panel of a dbx 1231 graphic EQ rack unit

The front panel of a dbx 1231 graphic EQ rack unit

The rack box also included a small 4-in/4-out XLR loom which meant the units could be easily connected up without wasting time poking around in the back of the rack.

Having 4 channels of EQ to play with, I was able to route the left and right main outputs (the FOH mix) from the desk through a channel each on the first dbx 1231 unit, and then run the auxiliary 1 output (which I was using for monitoring) through the first channel on the second dbx 1231 unit. This was great, as I hoped to make far less drastic adjustments for the FOH speakers than I expected to for the monitors, if indeed I needed to make any at all.

The Process

Jon took me through the basic procedure for ringing out the speakers, and then I went through it a few times by myself, as the best way to learn is to have a go. For this application we used the more aggressive +/- 15dB option on the units, to increase our ability to iron out troublesome frequencies.

Essentially, the process is simply to start turning up the gain until the system starts producing some feedback, at which point you locate the offending frequency and reign it back a few dB using the relevant EQ band. You can then continue pushing up the gain until feedback occurs again, and so on.

You can keep doing this for a while, but you should reach a point fairly soon where you should stop. But when is this point? How do you know when you’ve reached it? Well, here’s a few hints:

  • The most obvious time to stop is when you have the system up as loud as you’re going to need it. If you’re using the PA to amplify a quiet lounge band for an evening party, you aren’t going to need to push it to it’s limit, and so you don’t need to cut out more frequencies than you need to. Remember that the more frequencies you cut, the more the original sound is affected. Which brings us to our next point…
  • You should stop before the sound becomes unnatural or unmusical. This is really down to the engineer’s own discretion, but it’s very important not to ruin the original sound by cutting too many bands too harshly. Compromise between ringing out and sound fidelity is key here.
  • Finally, at times you’ll find that you logically can’t continue once you reach the maximum cut for a certain frequency band, because the system is continuing to feedback at that frequency. You could potentially use another EQ unit to cut yet more, but you’d be wiser to seek an alternative solution here, such as alternative microphone and speaker placement, and maybe addressing the venue acoustics themselves.

Tip: Harmonics – feedback caused by standing waves in the room have harmonics, which occur at a multiple of the original feedback frequency. So if you have a standing wave at around 100Hz, you would find harmonics at around 200Hz, 300Hz, 400Hz and so on.

Tip: The frequency of a note in any given octave is double that of the same note in the octave below. For example, the note A above middle C on a piano has a frequency of 440Hz (orchestras call this “concert pitch”). The note A below middle C has a frequency of 220Hz – exactly half that of the same note in the octave above. The note of A above concert pitch has a frequency of 880Hz, and so on.

This can help us in pinpointing frequencies when ringing out a PA. If you are searching for a troublesome frequency by raising EQ bands and listening for a match, you may hear some feedback of the same note, but at a higher pitch. What you’ve found is a harmonic of the frequency you’re searching for. If this happens, return the band you just tested to it’s previous setting, and then half that band’s frequency, and try the EQ band that corresponds to that new frequency. So, if you raised the 1.25KHz band (1,250Hz) and found it to be a harmonic of the frequency you were looking for, you should then try finding it at 625Hz (1,250Hz / 2 = 625Hz). On the dbx 1231 unit, this would correspond to the 630Hz band.

“Aha” Moment: You’ll be able to appreciate now that the seemingly random frequency numbers on the dbx 1231′s 31 EQ bands are actually all harmonics of each other (well, nearly). There are three harmonic sets, as follows:

First Set

20Hz – 40Hz – 80Hz – 160Hz – 315Hz – 630Hz – 1.25KHz – 2.5KHz – 5KHz – 10KHz – 20KHz

Second Set

25Hz – 50Hz – 100Hz – 200Hz – 400Hz – 800Hz – 1.6KHz – 3.15KHz – 6.3KHz – 12.5KHz

Third Set

31.5Hz – 63Hz – 125Hz – 250Hz – 500Hz – 1KHz – 2KHz – 4KHz – 8KHz – 16KHz

You can clearly see that in each set, each frequency is approximately twice that of the one before it.

On The Day

The wedding job on the Saturday was pretty straightforward. My friend Luke helped me out setting the gear up, and the biggest cause for concern was making sure not to drop a speaker on the wedding cake.

Tip: Do not drop PA speakers into wedding cakes at weddings.

Luckily we managed to get everything set up without any major cake-related disasters, and the staff manager at the reception venue was very helpful, finding us a small table to put the 12 track mixing console on, and listening to the details of how to use it to pass on to the wedding party when they arrived. When we came back to pick up the gear at 11pm, everyone seemed drunk and happy and left us in peace to pack up the gear, so I assume that they didn’t have any problems.

The following morning, after a small satellite navigation related mishap involving a recently moved motorway junction, we arrived to set up the gear at the garden party. The family’s two daughters (one of whose 21st birthday the party was for) were going to be providing entertainment, singing along with instrumentals from the musical ‘Cabaret’. Everyone was very friendly.

We set up the microphones, monitors and FOH speakers (on stands) in front of a set of french doors, with the FOH speakers facing out into the garden, and the mics and monitors facing the house. Because the FOH faced a large open space without many surfaces to create reflections, there were pretty minimal issues when it came to ringing out. I spent some time on the monitors though as these faced the walls of the house, as did the microphones. Eventually I reached the maximum cut on the 1.25KHz band and couldn’t continue, at which point I pulled the monitor aux-send back down a couple of dB and we did a soundcheck. The girls’ vocals sounded great, even at high volumes, without any of the jarring resonances that often seem to appear when the gain is pushed a bit. In the end the clients asked for a little less volume than we’d originally soundchecked with, but I left the settings on the EQ rack as they were in case they decided to push it up again later once all the guests arrived and started chattering over the music.

When we came to pick up the PA gear up later on, the guests all seemed to have left again, and the girls were nowhere to be seen (unfortunately). However, everyone we’d met earlier on seemed very happy with us, and even helped us carrying the gear back over to the van. As we were preparing to leave, a man who I guessed was the girls’ grandfather stopped to speak to us. He said that at all the other parties and events he’d been to in the past, feedback had always been a problem as soon as anyone so much as approached a microphone, but that they hadn’t heard any all day.

“I know it was all down to the extra time you spent while setting up, and I really appreciate it,” he said. “So have a drink on me.” With that, he handed us a £20 note and waved us off.

So that was good.

Categories: Live Sound Tags: , ,

Vocal doubling: First takes aren’t always best

February 26th, 2011 No comments

Vocal doubling is a recording technique in which the vocalist is recorded singing the same take twice, and the two takes are layered on top of one another. This can help create a richer, more complex sound, and be used to add width to the stereo image, among other things.

However, many vocalists struggle to exactly replicate their initial performance on a track when trying to do vocal doubling. We’ve all been there – no matter how many takes you record, certain syllables in the phrase just aren’t sitting right. A particular plosive always seems to be a tiny way out, or a certain syllable seems to start just a little too early.

This can be incredibly frustrating, both for you and your vocalist, and can bring a session to a miserable standstill. Before that happens though, here’s something you can try.

Instead of trying endlessly to have your vocalist replicate exactly what’s on the lead vocal track, try having them do a few takes of the phrase, and then see if you can match these takes themselves with each other, while muting the original vocal. I find, more often than not, that the vocalist’s delivery in later takes tends to be very consistent, and perfect for doubling, while takes done earlier in the session have often suffered from imperceptible differentiations (which could be put down to all sorts of things – warming up the voice, getting comfortable with breathing, phrasing, the vocalist becoming more relaxed, and so on).

It’s often hard to hear any difference between an earlier take and a later take when you’re just listening to it on it’s own, but when you start directly comparing takes, as with vocal doubling, the slight differences start to become apparent.

Once you’re found two takes that work together, you can then just cut one into the original vocal, and use the second as the double.