Reason causing feedback during live vocal processing

[Bjørn Felle]

Hi all,

I recently performed live for the first time at an open mic event. The gig was great and my material went down really well. The only problem was that I couldn't use any live vocal effects processing, which detracted from the performance from my point of view (though the sounds was still great).

I'll outline the setup I attempted to use and I'm hoping someone will be able to help me to work out what went wrong.

My mic went into a Behringer mixer to function as a preamp. The output of that went to the input of a Behringer USB audio interface. Reason was set to use this device as both input and output. On one audio track in Reason I had a mixdown of the whole track without the vocal. On another audio track, I had all of the vocal effects and automation events to fade effects in and out and control formant shifting etc. This track had monitoring enabled. When I rehearsed at home with the output of the Behringer connected to my studio speakers, it sounded great. But at the venue, there was so much feedback that I couldn't use the system. The output of the Behringer USB interface was connected to the venue's mixing desk which then went out to their PA speakers.

In the end I had to connect my mic to the venue's mixer, and just use Reason on the laptop to play the backing with no input enabled. It worked, but of course I had no effects on the vocal.

I'm wondering whether the problem was caused by compression, limiting and output gain boosting going on on the vocal tracks. I did this of course to keep the vocal at a high enough level to sit right in the mix, but because it was a monitored input, all of the backing being played out of the speakers would have been picked up by the mic, and then compressed/limited/boosted too. I'm thinking of changing the setup as follows:

My mic into new Behringer USB interface with a mic input. Reason set to use this device as both input and output. One audio track in Reason with all of the vocal effects and automation events to fade effects in and out and control formant shifting etc., monitoring enabled, but attenuated to around line level (can be figured out during sound check). Behringer USB interface output to one channel on venue mixer. Backing track being played on iPad, synced to Reason so they play at the same time, and output connected to second channel on venue mixer.

Any input would be really gratefully received. Thanks in advance

[BertKoor]

Hi Bjo/rn,

Congrats with the succesful gig! Too bad you suffered from feedback though...

I've done FOH (FrontOfHouse) & monitoring mixing for some years, maybe I can shed some light on the subject.

<tl;dr> / management summary: it depends on the mic, speakers, acoustics, effects (esp. comp/limit/distortion) and probably a bunch of other factors as well, like skills of the engineer.

Now the long answer...

To start with the last point: as soon as your mic was plugged directly into the mixer of the venue, all feedback issue were gone, right? That's because now the sound engineer of the venue had control over the gain & eq of the mic, which are essential in killing feedback. With just the total mix he could hardly do anything about it.

So, given a simple setup: mic, mixer, amp, speakers. No more than that. Will you get feedback? Not with the right gear and right engineer.

Microphones... What's yours?
For vocals on stage I prefer the Shure SM-58. It never has let me down. My rock band in the first years we used all cheap stuff, sounding bad and very feedback sensitive. Once we switched to proper mics with flatter frequency response, it sounded better and we could have monitoring much much louder without any feedback.
One time I had a girl on stage that brought her own fancy expensive phantom-powered AKG. It sounded OK, but gave me a headache with feedback all the time. So it's not always cheap mics that give feedback. All you need is a good hump in the frequency response to make it extra sensitive to feedback.

Mixer...
I usually worked with analog A&H, Mackie or Soundcraft desks with relatively simple 4-band EQ on each channel of which two are sweepable parametric. The other two shelve low & high. That is enough for both shaping the sound so it resembles the original again, and also remedy any occurring feedback. Brand of mixer doesn't really matter, number of EQ bands does! And you can do without limiter, compression etc for live sound. Keep your finger at the fader to control volume. I do add reverb & delay to taste. My taste. Might not be your taste.

What does an engineer at the desk do about feedback?
1. Turn the offending channel down asap! Can be a challenge finding it, so turning mixer outputs down also is an option.
2. Dial in the feedback frequency on the parametric EQ and turn it down. Will compromise sound quality, but needs to be done.
3. Tell the vocalist to not point the mic directly into their monitor, for fecks sake! Hands too close to the capsule can have adverse effects as well.
Oh, an older engineer taught me this trick: insert a 32-band graphical EQ inbetween mixer output and amp for the monitors. Dip the bands 2k, 4k and 8k by 6dB. Then chance on feedback is reduced by 90%.
Behringer sells cheap 32-bit graphical EQ's including automatic feedback killing cirquit. It works reasonably well.

Amp / Speakers...
Cheap speakers or badly positioned ones will give issues. Same as with mics.
Not much problems with amps themselves, but the total volume matters. You had no problems at home, good thing you tested it! But how loud is that? Recommended volume safe to endure all day long is 85dB SPL. About as loud as in a cinema. Producers tend to have monitors somewhat louder, like 95dB. In a live venue it can easily get to 105dB. That is much louder.

So... what can you do?

I blame your effects. A compressor is a box containing a fader controlled by a gremlin. With no input he turns the volume up, as soon as you sing it gets turned down. So it adds extra gain, which for feedback is very bad.

Hope this helps...

[Forgotten]

+1 on everything Bertkoor says.

Personally I would focus on the processed signal before it goes to the house desk to manage the feedback. If you push the house PA to a level higher than you expect to use during the gig and then run through a few tracks, you should be able to to figure out which frequencies are causing the feedback. If you put a multi band EQ between your output and the house desk, you can identify frequencies that cause issues, and edge down each band until you get rid of the feedback. Running at a lower volume than you EQ at means you have less chance of the probems occurring during a gig.

Also, don't forget that every venue is different, so don't assume that your mix is where the issues occur - it's a combination of your mix and the venue's acoustics: you need to EQ for every venue you play.

[Bjørn Felle]

Thank you so much guys, I really appreciate the support and feedback (if you'll pardon the pun ). Just to reassure you, we switched to using their mic directly through their PA with a bit of reverb added, so there was no feedback during the actual performance. The feedback was picked up during sound check, so I wasn't struggling with it during the performance.

I'll try to address each point individually but won't quote for the sake of brevity.

• Mic: we switched to using their own mic and cable with my setup, and it made no difference, which rules out the mic as an issue.

• Mixer: I don't know what theirs was specifically, but it looked like a Behringer judging visually (I was a bit too discombobulated to pay close attention). It was at least 8-track, and definitely had EQ and built-in effects per channel. For open-mic events, however, I don't think there is time to play with EQs for each performer just to do two tracks. The engineer/organiser was performing too so had only limited time to help me set up which is understandable.

• Amp/speakers: again I'm restricted by what's available at the venue, and I don't know what they were using. But while it's possible this was a factor, it didn't cause a problem once the mic was connected directly to their mixer.

• Effects: I'm really glad you said that you think the compressor is contributing to the problem. It makes so much sense that compression and limiting would be contributing to feedback, especially the way I use them not only as tools but also as effects in themselves (e.g. sidechaining the vocals). I can do without all that in a live performance, however. Much more important is the formant shifting and finely tuned reverbs and delays. I have tracks which I simply cannot perform without that.

• EQ: I can use multi-band EQ in Reason, before the output to the sound card. But again sadly I don't think I'll have time to get this set up before actually performing. I think in a situation where I'm playing a whole set, I'll be able to do a more thorough sound check earlier in the day and really nail the sound. But for open-mic it isn't really possible due to the number of performers on the night. I totally get what you're saying about needing a different EQ for each venue due to the varying resonance in each room.

I think what I'll have to do is try bypassing all of the compression and limiting in the Reason project file. I have found that I can also send each audio track to a different hardware interface, so I can have one going to my USB Behringer interface, and one to the built-in interface. That means the vocal can be levelled independently of the backing without me having to worry about syncing two separate pieces of software. I'll see how I get on with that after having disabled all the compression and limiting and perhaps reducing the overall level. I'm performing in the same venue on the 8th of August so I'll report back after that.

Thanks so much again guys, I really really appreciate you both taking the time to help me

[BertKoor]

Hmmm, maybe if you adjust the compressors settings you can get away with it. Especially don't set threshold too low, let it compress only the loud sections.

Having vocal output on a seperate channel also is a good idea. You could even do that with a simple stereo interface: vocals on left output and all the rest on the right. You lose stereo ofcourse, but the mixing guy regains control again. He can pan the two channels in the middle on the mixer.

[Bjørn Felle]

Hmmm, maybe if you adjust the compressors settings you can get away with it. Especially don't set threshold too low, let it compress only the loud sections.

I could tweak the threshold during sound check to see if I can get away with it. Failing that I'll just disable the device and let the engineer tweak the level throughout the performance.

Having vocal output on a seperate channel also is a good idea. You could even do that with a simple stereo interface: vocals on left output and all the rest on the right. You lose stereo ofcourse, but the mixing guy regains control again. He can pan the two channels in the middle on the mixer.

That's a good suggestion, but panning is important in my stuff. Things are getting panned all over the place and I don't really want to lose that aspect of the performance. I'm a Mac user and it's easy enough to create an aggregate device with outputs from multiple interfaces. So I can have the backing going to the output of one interface, and the vocals going to the output of another, which should allow the engineer more control.

Thanks again dude