Feedback Management

[twotoedsloth]

Hello,

Many thanks for taking the time to read this and hopefully offer some advice.

I am looking for a feedback suppression device, for a small hall, seats about 500. The speakers attached will be Meyer Sound Ultra X40s.

I was thinking of two boxes, the dbx AFS2, or the dbx Venu 360. Is the Venu worth the extra $500 USD?

Please let me know what you think,

Peter

[AlecSp]

Save your money and learn how to optimise your system (mic choice, mic usage, positioning, appropriate levels, EQ, acoustic treatment)

No feedback suppressor is a magic bullet, as so many have discovered over the years.

In the end, the factors I mentioned will have far more impact than any processor driven dumb. Though you'll never be able to beat simple physics.

And I'd hope that, with that level of FOH speakers and venue, you'd have the budget, people and kit to deal with engineering appropriately.

[resistorman]

^^
second this.

[Wonks]

If this is for music, rather than theatre, then you are very unlikely to get feedback from the FOH speakers unless people are using mics in front of the speakers.

99.9% of feedback is from monitors.

Keep the stage instrument levels under control then the monitors don’t need to be as loud and you can seriously reduce the risk of feedback.

In-ear monitoring reduces the risk to almost nothing.

Inexperienced stage acts are going to be the main problem.

If you are using stage monitors, then 31-band graphics on each one notching out the most prone to feedback frequencies will help, but if the mics change each time, then the problem frequencies are likely to change with them. So it’s sorting things out at the soundcheck and learning to tell what frequencies are feeding back so you can adjust the graphics accordingly.

But far better to have everything set up so there’s no likelihood of feedback to start with.

[Mike Stranks]

Yup!

Feedback suppresion devices IME are never that useful. As others have said, far better to spend time looking at configurations, levels and channel and various 'group' EQs - especially FoH.

Might be a good idea to find an experienced live-sound buddy to sit-in on some gigs and help sort out actual/potential feedback issues.

(I was at a large funeral on Monday. Sophisticated digital sound system. Pulpit and lectern mics on the verge of ringing the whole time. I yearned to nip back and tell and show the operator how to apply some judicious lo-mid cut to ameliorate or eliminate the problem. Obviously I didn't! )

[Luke W]

Agree with all that's been said so far. There have been more than a few times that I've solved these sorts of problems simply by pointing an amp in a slightly different direction, moving a monitor to the side (or, heaven forbid, just turning something down a bit!) and so on.

(I was at a large funeral on Monday. Sophisticated digital sound system. Pulpit and lectern mics on the verge of ringing the whole time. I yearned to nip back and tell and show the operator how to apply some judicious lo-mid cut to ameliorate or eliminate the problem. Obviously I didn't! )

There's definitely a time and a place isn't there...

[MOF]

(I was at a large funeral on Monday. Sophisticated digital sound system. Pulpit and lectern mics on the verge of ringing the whole time. I yearned to nip back and tell and show the operator how to apply some judicious lo-mid cut to ameliorate or eliminate the problem. Obviously I didn't! )

A related topic, sadly last year two relatives died months apart and their funerals were streamed on the internet, one audio channel was from the lectern microphone, the other was from the on-board camera microphone at the back of the room. The latter overwhelmed the former and made it, at times, impossible to hear what was being said.
I was able to record just the lectern microphone and send it to family members, they appreciated being able to hear the service clearly.
I had to repeat the process for the second funeral because nothing was done to route the lectern microphone to both channels, despite my comments being forwarded to the company providing the system.
Also I agree with the replies, best to use the laws of physics to your advantage than a feedback eliminator, once it can’t cope there will be horrendous feedback.

[MarkOne]

Pulpit and lectern mics on the verge of ringing the whole time. I yearned to nip back and tell and show the operator how to apply some judicious lo-mid cut to ameliorate or eliminate the problem. Obviously I didn't! )

Probably the operator wouldn’t have known what EQ was anyway. Midweek funerals are always most likely not to get the experienced PA operator. (Or even a regular production team member at all)

[twotoedsloth]

Hello,

Thanks for your kind responses. I was hoping that the feedback suppressor would make life easier, and potentially save time.

I work at a university, and I was hoping to not work for every concert, masterclass, lecturer or lecture/recital. I would like to pass this off to some enthusiastic students. The lecture recitals are the most difficult to accommodate.

Does anyone know if the feedback suppression in the Venu360 is the same as the AFS2? It might be handy to have the RTA functionality at some point in the future, or is this a red herring? If I do go with the Venu360 what mic should I be purchasing for the RTA?

Sorry if you feel I am not observing your advice. I do understand that it might not do all I am hoping for, but I have found it is relatively easy to find funds for equipment, but next to impossible to find money for an assistant.

Again, sincere regrets and thanks for your advice.

Peter

[Wonks]

It might help if you could describe the setup, where the speakers are in relation to the microphone(s) etc and if there's any monitoring, and what sort of performances you expect to host? Is there a proper stage?

[Wonks]

The problem with automatic feedback destroyers is that unless you set them so that each available notch filter works only once and doesn't keep reacting after that, then you end up with the width of each notch getting wider (and if allowed to, deeper).

It's very easy to end up with a terrible sounding and much quieter system.

And the once only mode is pretty much like ringing out and using a 31-band graphic EQ, but at least you've got control of that.

You'd probably need a combination of a 31-band graphic for basic system set-up and anti-feedback duties, and an auto feedback suppression unit, set for single use operation, which will need to be reset each time the system is used (if the kit isn't powered off between uses and doesn't auto reset).

[twotoedsloth]

Hello,

Thanks for your patience, I had to figure out how to post an image here again.

http://walterhallrecordings.music.utoro ... forSOS.jpg

I'm sorry for appearing to be recalcitrant.

This is the hall I will be installing the feedback eliminator in. A big part of why we have feedback issues is because the speakers are hanging on the walls behind the performers. I can usually manage by manually adjusting the EQ to remove the offending frequencies, but as I said, it would be a tall ask to put that on a student with little to no experience.

Please accept my sincere apologies for appearing to spurn your advice.

Thanks,

Peter

[Drew Stephenson]

Nice hall but I can see why feedback is a problem. And, frankly, is probably always going to be a problem.
Is there any scope to get the speakers moved? Perhaps a flying installation above the stage so that they can be at least in line with the microphones?

[Bob Bickerton]

The problem is clearly the speaker positioning and as has been said feedback eliminators have very limited value if you’re breaking the laws of physics.

You say your main issue is with recital lectures. Are you talking about a lecturer speaking whilst musicians play acoustically? Average lecturn mics can be feedback prone and maybe an alternative solution could be found there.

If the problem is from amplified musicians then you really need to deal with speaker placement.

Having said that, I found the QSC TouchMix desks pretty handy at eliminating feedback. Check out https://training.qsc.com/mod/book/view.php?id=595 if I was in a hurry I’d use their eliminator to identify and notch out rogue frequencies, but never used the roving function.

You could/should train your students in basic feedback management before letting them anywhere near a sound system anyway

Bob

[Exalted Wombat]

I hope we can assume the microphones we see in the picture are a recording rig, not for pa?

[twotoedsloth]

Hello All,

Thanks for your helpful posts.

First, Exalted Wombat, yes the ORTF rig in the center of the stage is for recording only.

Next, blinddrew, we're not likely to have budget to move the speakers, but I will continue to ask for this.

Bob Bickerton, the reason lecture recitals are problematic is that you have to alternate between settings suitable for speech, and settings suitable for classical chamber music, or small jazz ensembles. I am not likely to buy a new mixer, but if I do, I'll look closely at the QSC Touchmix. Is the touchmix that much better than the dbx AFS2 when it comes to notching out offending frequencies?

[Bob Bickerton]

Bob Bickerton, the reason lecture recitals are problematic is that you have to alternate between settings suitable for speech, and settings suitable for classical chamber music, or small jazz ensembles. I am not likely to buy a new mixer, but if I do, I'll look closely at the QSC Touchmix. Is the touchmix that much better than the dbx AFS2 when it comes to notching out offending frequencies?

I can't really comment as I don't know the DBX.

I'm a bit confused about the input source of your feedback (cognisant that the speakers are in a poor position).

Are you saying that feedback is a problem for the three scenarios you mentioned? Is the 'speech' scenario from just one mic, multiple? Lavalier? Headset? Lecturn?

I don't understand why you'd get feedback for classical chamber music - is it from a presenter or lecturer mic?

With Jazz Ensembles, is the feedback from miked instruments, soloists?

Not being critical here, just trying to understand the problem more fully - and not sure I can help.

And another thought - are the speakers shown in shot the only FOH speakers or are others flown? They would be working hard to fill that size of auditorium and pretty much impossible to do without feedback. Can you move them?

Bob

[twotoedsloth]

Hello again, and many thanks for all of your advice.

The problem with the classical lecture recitals is that to avoid feedback I have to notch 2.5kHz and 4kHz for speech to prevent feedback, but for recording I do not want those notches present, so I have to return to a flat frequency setting when the presenter is not speaking. Usually the presenter is speaking through a wired mic at the lectern (most of the time a Sennheiser e935, but we can do an SM58 if the presenter wants it), on request we can also do a lav mic.

With the Jazz recitals we get feedback on the double bass mic and the piano mics, the solo horns and guitar are usually fine without tweaks.

So, I guess you are all saying that a feedback eliminator is a waste of time and money... do you think it could have even a small incremental improvement in feedback rejection?

[Wonks]

To operate the system in the way you've described, someone is going to need to turn the feedback eliminator on and off or bypass it in the way you are currently doing with the graphic.

It sounds like you probably aren't setting up the FOH graphic EQ and the recording path correctly if the graphic is affecting the recordings from the desk. Possibly worth starting another thread if you want to describe the desk and how the FOH graphic and recording outputs are configured.

The different scenarios you've given mean they each will have its own problem range of feedback frequencies, and what a feedback eliminator learns for one set-up will be wrong for another.

There is normally a mode where a feedback eliminator will reset a notch filter if there hasn't been any feedback on that frequency for a given period of time. That's the sort of setting you'll need if you want to leave it to do its own thing and not have to manually switch the unit in and out. But it takes time, so after setting up for feedback suppression for speech, if the presenter then plays back some example music, that music will still be affected by the feedback filter notches until the filters have fully reset themselves.

But go on a pro PA forum and no-one likes leaving a feedback suppression filter in auto mode as the notches do tend to get deeper and wider and affect the sound more and more. With your double bass example, you could end up with almost all the bass end notched out. Or if you limit the filter notch width, then a filter may not be able to eliminate all the feedback.

Some eliminators allow you to select the number of fixed filters vs the number of filters that are automatic. The fixed filters are set for the system before any performance by turning up the system volume until feedback occurs and a filter kicks in, and then you turn the volume up again until the next filter kicks in, and repeat until all the fixed filters have been activated. You can then leave some filters as automatic to catch any feedback that occurs during the performance. But these would normally be limited in notch depth and width (if the system allowed) to minimise the damage these can do the sound.

Whilst the feedback detection algorithms have got better over the years, there is still the chance that they can mistake a musical (or non-musical) note for feedback (say a guitarist with an amp on stage creating feedback deliberately), then the filter would try and notch that out from the FOH sound. Which is why FOH sound people prefer to do things manually.

According to the dbx literature, the dbx AFS2 and the dbx Venu 360 have the same anti-feedback algorithms, the Venu just seems to have more inputs and outputs than the AFS2. Which, as you only appear to need two channels, the AFS2 should do the job.

The dbx Drive Rack series started out as a combined 1) crossover for passive component speaker systems with power amps, 2) a FOH graphic EQ auto setting device using pink noise and a real-time analyser, 3) a time delay function for delaying the signal to secondary speakers forward of the stage speakers and 4) some feedback elimination thrown in to use up the remaining available DSP.

They've got more sophisticated as time has gone by, but the crossover function is required less and less as people install active speaker systems. Most of the other functions offered are standard featured in digital mixing desks, so there's often now no need for one at all.

I'd imagine that it should be possible to hire one for a week or so, and it would be informative for you if you could install one and play with it if you have a week of mixed performance styles that covered all the normal scenarios. Hopefully these would be non-critical internal events where if feedback did occur, it would be known beforehand that tests were being carried out to avoid complaints.

It may do what you want, it may not. But you'll never really know until you try one out.

We've tried to highlight the issues with this type of device and that they have definite limitations if used in auto mode for a wide variety of amplification uses. Note that it only takes someone to say substitute an omni pattern mic for a cardioid one because they prefer the sound, before you get to a situation where a feedback suppressor can't cope and you need a knowledgeable FOH engineer to step in and fix things.

But it does sound like you may need some help in the recording vs FOH feed configuration on your desk.

[Murray B]

What wonks said, put an eq in the path of the lectern mic, You might need to purchase a separate one to use and a insert in the channel depending on the routing/ patch-bay options on your current mixer and/or if you are using a inbuilt graphic or still need to have the existing graphic in the path of the overall mix.