feedback problems

[jamlucas]

hi guys, wondering if you could help,

 

I'm a lighting tech, and somehow got a job doing the sound as well in a venue. I'm having trouble with a bit of feedback when using headmics, well a lot of feedback. hand held mics are fine, but with headmics I'm getting all sorts of feedback all over the place. I have a DBX driverack installed, and reading the manual it says that there is a feedback destroyer in it but I really don't know how to make it work properly. the only way I have got round this is taking bottom end off the mics and running it quieter than I would like.

[Wilflet]

feedback suppressors arn't generally the best idea. They require it to feedback before they can do anything, and often take out too much of a frequency band, then have limited bands so say the 6th time it feeds let go of band 1 to catch the 6th, at which point the first frequency is a problem again.

 

Is there a graphic EQ over the System?

Between that and the channel EQ you can get a a flatter response that will give you the best gain before feedback.

 

Other than that you need to look at things like proximity to the speakers and mic pick up patterns.

[Shez]

Other than that you need to look at things like proximity to the speakers and mic pick up patterns.

 

The first thing to look at is where the speakers are in relation to the mics in my opinion. Feedback destroyers can help a little if all else fails but if you fix any positioning issues, you'll often find you don't even need them. It's perfectly possible to get omni headset mics very loud but it relies on both them and speakers being of reasonable quality, suitably EQed and speaker dispersion patterns made optimum use of.

[Wilflet]

Other than that you need to look at things like proximity to the speakers and mic pick up patterns.

 

The first thing to look at is where the speakers are in relation to the mics in my opinion.

 

Agreed, I'd only put it in my first reply as something of an after thought to EQing due to the OP giving me the impression that it might be the sort of nightclub set up where its a hell of a battle with management to alter anything 'installed'.

[johndenim]

Another useful point, would be to make sure that the mics are rung out properly.

 

Feedback destroyers are okay within limits, they do tend to notch out many important frequencies though, you're better off as suggested with a good EQ unit.

[Stan Hope-Streeter]

Another useful point, would be to make sure that the mics are rung out properly.

Feedback destroyers are okay within limits, they do tend to notch out many important frequencies though, you're better off as suggested with a good EQ unit.

 

Where did you get that idea?

 

If correctly used, a FBX only notches out the feedback frequencies, and does it far more precisely than a graphic, and far more quickly than a manually adjusted parametric.

 

 

To the OP:

 

You really shouldn't be using head mics unless you have a good quality insertable equalizer - either a third octave graphic, a 5 (or more) band parametric, or a FBX.

 

I've never used a dBx Driverack, but I wouldn't want an FBX over the main system outputs in this application. The anti-feedback EQ ought to be applied solely to the headmic channels, via a subgroup insert.

However it might be better than nothing. Does the user manual have any instructions in setting it up?

 

The main point is to use it to ring out the mics pre-show, then lock it.

NOT to allow it to search for feedback during the performance.

 

HTH

[Joe90]

Another useful point, would be to make sure that the mics are rung out properly.

Feedback destroyers are okay within limits, they do tend to notch out many important frequencies though, you're better off as suggested with a good EQ unit.

 

Where did you get that idea?

 

If correctly used, a FBX only notches out the feedback frequencies, and does it far more precisely than a graphic, and far more quickly than a manually adjusted parametric.

 

 

To the OP:

 

You really shouldn't be using head mics unless you have a good quality insertable equalizer - either a third octave graphic, a 5 (or more) band parametric, or a FBX.

 

I've never used a dBx Driverack, but I wouldn't want an FBX over the main system outputs in this application. The anti-feedback EQ ought to be applied solely to the headmic channels, via a subgroup insert.

However it might be better than nothing. Does the user manual have any instructions in setting it up?

 

The main point is to use it to ring out the mics pre-show, then lock it.

NOT to allow it to search for feedback during the performance.

 

HTH

 

 

 

 

 

 

 

 

I must ammit I haven't done any theatre work with head worn mics but recently I managed to pick up a DBX feedback suppressor and looking through the manual the 'notches' are extremely narrow and each channel can have upto 24 notches applied. This can be split into some fixed and some roaming, at the present I've got 12 set up for fixed only and no roaming and this seems to work very nicely. I get more volume from monitors than I ever did with a GEQ.

 

The only issue is setting the suppressor up as you need to let the suppressor hear the feedback for a few seconds for it to recognise the frequency and notch it out, this can be quite a loud painful experience but if you insert a compressor across the mic channel you are ringing out set at the highest ratio will stop the feedback running away with you. With this method I just keep increasing the aux pot and just let the suppressor do it's job, once I have enough gain before feedback I lock the notch filters in place and 'Bob's yer uncle'...

 

The only con's are, once its set up in sound check thats it as with a GEQ you could do some adjustment mid show if a frequency starts to ring, a compressor is needed so you don't deafen the venue's staff with howling feedback when your setting up.

 

The pro's are, 2 channels of suppression in 1U where as a decent 31 band GEQ will be atleast 2U and more than likeley 3U, narrow notches and plenty of them, easy to set up...

 

I've always used GEQ's in the past and have been sceptical about feedback suppressors but I have very much changed my opinion. I would only ever use them on monitors and never FoH but that also applies to a GEQ. I can't comment on the Behringer unit which seems popular but the DBX unit I would recommend.

 

Just looked at manual again and the highest notch Q is 1/80 octave!

 

Cheers,

 

Joe.

[johndenim]

Another useful point, would be to make sure that the mics are rung out properly.

Feedback destroyers are okay within limits, they do tend to notch out many important frequencies though, you're better off as suggested with a good EQ unit.

 

Where did you get that idea?

 

If correctly used, a FBX only notches out the feedback frequencies, and does it far more precisely than a graphic, and far more quickly than a manually adjusted parametric.

 

Have you any experience with feedback suppressors at all, or is just a vain attempt to mock me?

 

A feedback suppressor reduces problem frequencies, or attenuates the similar frequencies that are related to feedback.

This attenuation is not instant, and the whole audible range is affected.

An example to anyone never hearing this attenuation, would be similar to an OP fading a specific frequency up and down in a rapid motion, or a heavy compressor.

I didn't even mention parametric EQ's, Do we not have manually adjustable 31 bands then?

Are you saying that an experienced engineer is not able to compete with a feedback rejection unit?? :)

[lightsource]

Have you any experience with feedback suppressors at all, or is just a vain attempt to mock me?

 

A feedback suppressor reduces problem frequencies, or attenuates the similar frequencies that are related to feedback.

This attenuation is not instant, and the whole audible range is affected.

 

Totally agree here JD, but if you do use a feedback destroyer, then where you place it determines how the 'whole audible range' is affected.

 

If you wire it as a L+R insert on the Master out, then yes, it will affect the whole EQ of the show, however if it is only placed on a single vocal mic, or a vocal mic subgroup, then it will only affect the channels it's assigned to., rather than the whole L+R mix.

[Bobbsy]

Feedback is a perennial problem and curing it is a step by step process. There's no "magic bullet" to eliminate feedback and, frankly, a "feedback destroyer" in the Driverack would be low on my list of options for the reason Lightsource mentions.

 

Things to play with in roughly the order you should attack them:

 

1. Play with the position and aiming of the speakers (both FOH and any monitors you're using). For FOH, the goal is to point them so the maximum audio hits the audience with the least possible getting back to the stage, either direct or bouncing off hard walls and ceiling. Small changes can make a big difference to feedback. Similarly, get the monitors, as much as possible, in the nulls for the mics you're using.

 

2. Play with the position of the "headmics" that are causing you issues. Our old friend the "inverse square law" means that getting the mic even slightly closer to the sound source (i.e. the mouth) can make a huge difference. Many headset mics are omni pattern (better for sound, less good for feedback rejection) so positioning is key. If they happen to be directional mics, then make sure they're pointing the right way.

 

3. EQ the mic channels. I don't what mixer you have or how good your channel EQ is--but the human voice is largely in mid-range frequencies. Assuming decent EQ, you can get rid of a lot of the lows and highs without hurting the voice at all--but for a reasonable improvement in gain before feedback.

 

4. Ring out the system. Ideally, you'll do this on the vocal channels only via a subgroup rather than the main outs. I won't go into the details of how to do this here but searching on "ringing out a PA" will give you lots of instructions. However, don't take this too far--four to six problem frequencies reduced by a few dB each is the ticket.

 

5. If you have to (and have no other option) then applying EQ or a FB destroyer in the main outs is the last choice. If you're using the feedback destroyer, you're better off just using it as an automated "ring out" then fixing the frequencies rather than using it on full auto in a live situation.

 

Finally, be realistic in your expectations of how much level you can achieve in a small space without feedback. Sometimes you just have to accept that you're fighting the laws of physics!

 

Bob

[Stan Hope-Streeter]

Have you any experience with feedback suppressors at all, or is just a vain attempt to mock me?

I have been using the various Sabine units regularly for at least ten years.

A feedback suppressor reduces problem frequencies, or attenuates the similar frequencies that are related to feedback.

This attenuation is not instant, and the whole audible range is affected.

The whole point of an FBX is that it ONLY attenuates the feedback frequencies. A few narrow notch filters have less effect on the overall sound than the ubiquitous 3rd octave graphic EQ - far less of the audible range is affected.

An example to anyone never hearing this attenuation, would be similar to an OP fading a specific frequency up and down in a rapid motion, or a heavy compressor.

Can you explain what you mean by that?

An FBX should only be used during the ringing out process, as a self-tuning notch filter. Just like using a manual parametric but much faster. In either case you have to hear feedback or ringing before you can EQ it out.

 

I don't know about the dBx unit mentioned in another post, but the Sabines have a setup mode which effectively puts a compressor/limiter on the signal to hold the feedback at a low level whilst it finds and notches an offending frequency.

 

I didn't even mention parametric EQ's, Do we not have manually adjustable 31 bands then?

A parametric is better than a third octave graphic for ringing out feedback, as it can be set to a much higher Q and only take out the precise frequency ranges necessary. The disadvantage is that it is much slower in use.

Are you saying that an experienced engineer is not able to compete with a feedback rejection unit??

In terms of speed of setup, I certainly am.

 

Bobbsy's answer is comprehensive, pay special attention to part 4.

 

Too many people seem to think that an FBX is supposed to be used to find and stop feedback DURING THE SHOW. This is absolutely NOT how they should be used. If you have set up and rung your system out properly you should never, never get feedback during a performance.

 

Theatre and speech applications using headworn and lavalier mics are where FBX units are especially effecitve - if used properly.

[dbuckley]

Most sound engineers have nothing but disdain for any of the feedback reducing solutions.

 

Most of the folks who have this disdain have, however, never actually used them in any serious fashion, and are thus pontificating from a position of theory rather than from practical experience.

 

At the end of the day feedback reducing solutions are just another tool in the engineers armoury. I find these tools to be useful.

[Shez]

It's perhaps worth thinking about what causes feedback to occur in the first place. A system will feed back at whatever frequency has the highest loop gain. For example, if your mic happens to have a big presence peak at 6k and your speakers also happen to have a big peak around there, that's the frequency that will feed back first.

 

Better kit tends to have a more even frequency response and consequently will suffer a lot less from this problem that kit from the budget end of the market. In either case, spending a bit of time with an eq on the master bus to even out any unpleasantness will pay dividends.

 

I've mixed a fair bit of musical theatre and have thus far never found the need to use any kind of eq for feedback prevention purposes on headset mics. Which is why I like them. Even with cast members coming out in the the audience and walking right in front of the speakers, there's always been plenty of level. As Bobbsy points out, the capsule is so close to the actor's mouth that unless they're stood next to a rock band, their voice should always be by far the loudest thing being picked up. (One venue that had a reverb time the grand canyon would be jealous of was rather more problematic but no amount of electrickery would have helped much there. That's something else to consider actually - the more reverberant the space, the more trouble you're likely to have with feedback, not to mention speech intelligibility.)

 

I did used to use a feedback destroyer a few years ago but no longer see the need now. With something fixed like a float mic I might still consider it but for a moving target like a headset mic, the other approaches in this thread will, in my experience, be a better bet.