Can someone explain this feedback problem?

[Dirk Diggler]

First subs don't usually cause feedback.

Feedback is simple, a microphone is the cause, the mic is picking up signal from a speaker/monitor. This causes an ever building feedback loop. Stop that interaction and you'll solve your issues.

Also the advice to turn things up to avoid feedback is the complete opposite of what you want.

Good luck,
Dirk

 

[Wasnex]

To align speakers, the delay needs to be on the outputs.

I checked the Ui16 user guide and could not find any reference that indicated it had delay on the outputs. It does have a delay effect. The best you can do is ensure you sub and mains are in the same polarity.

 

[Ostie]

Lots of good info here. I learned a lot.

 

[golfnsubie]

FWIW I have been playing bass and running sound from stage with a soundcraft ui24r for prob 140~ shows....I set channel gains so that they peak at -6db max and never actually clip, typically running between -12 and -6....soundcraft built in a soft clipping type or effect so that the it doesnt sound awful when you clip a channel, however it does help induce feedback by having the gains to hot....so I would start there by getting the gains in check...
Next once all mics and speakers are set....start with each aux send you use and then the mains....open the afs section of the eq page and first switch to fixed mode....go though your mics and bring the levels up enough to induce feedback...you will se these points checked by the afs...I then normally notch these manually, clear the fixed frequencies and bring it up again to see if there are any more frequencies, then switch it to live mode to deal with any changes that may pop up....do this for any monitor sends and the mains...doesnt take that long and you should be solid....I believe the EQ section on the ui16 is very similar if not exact....this is something you can absolutely practice at home to gain familiarity and comfort with the process...oh and I also always aux feed the sub....just my two cents

 

[FenderB]

Flipping the polarity can be helpful. A related idea is to change the distance between the mic and the speaker(s). Since the problem is frequency dependent, it is also wavelength dependent and therefore phase dependent. The feedback occurs where the sound from the speaker enters the mic close enough to "in phase" that a feedback loop is formed. By changing the propagation distance from the speaker to the mic, the sound enters the mic at a different phase angle. This may shift the problem frequency or eliminate the problem altogether.

Incredible, I wish I understood a fraction of your last three post. Amazing. Now back to our regular scheduled program.

 

[Rip Van Dan]

With the glass wall beside you and a wall behind you and the ceiling and floor it is acting like a big parabolic reflector. With the drummers vocal at the back where a lot of energy is being focused. You do get a lot of low mid energy of the back of most PA cabinets. Rolling right down the glass wall into the drummer vocal. 185hz must have been the resonant standing wave of the space you were in. Either notch that frequency or sometimes you can reverse the polarity on that channel to get rid of it.

Quick question for you: Wouldn't reversing the polarity change the phase on that channel? Wouldn't that cause a huge amount of phase cancellation from that mic?

Also, dropping the 185Hz frequency if you can isolate it should stop the problem in its tracks shouldn't it. The frequency feedback loop is caused by the natural boosting of that frequency by the environs, so cutting that frequency should both stop the feedback AND even out the sound because cutting it is pulling it closer to flat. Does that make sense?

 

[musicman7722]

Occasionaly when the band stops a tune I will hear a lo frequency feedback going on. if I leave it alone it just escalates. Not that I would. It always turns out to be the kick. Not sure why but my ear is more attuned to it. I don't feel I have it cranked and it only happens on rare occasions so keep that in mind to.

 

[rocu]

Soundcraft has some good tutorials on their YouTube channel. They mostly cover things mentioned in this thread but it can help to see it done on your own machine. The ui24r is a different animal from the ui16 but the interface is similar.

 

[trevcda]

It always turns out to be the kick.

At that frequency, I was think drums as well. If you're using a Ui16, you have adjustable high pass filters which will get rid of a lot of that mud and tighten up the low end. But you said you've isolated it to the drummers vocal mic. Again adjust the high pass on that channel, but also have a look at your main out EQ. Perhaps there is a lot of unnecessary boost around 185 Hz? Also check your channel EQ. If you're using the EQ presets (possibly in addition to main EQ issues) you may not find them to be as useful as one would hope they would be.

 

[JoeWPgh]

Were your PA speakers under the overhang?

 

[golfnsubie]

At that frequency, I was think drums as well. If you're using a Ui16, you have adjustable high pass filters which will get rid of a lot of that mud and tighten up the low end. But you said you've isolated it to the drummers vocal mic. Again adjust the high pass on that channel, but also have a look at your main out EQ. Perhaps there is a lot of unnecessary boost around 185 Hz? Also check your channel EQ. If you're using the EQ presets (possibly in addition to main EQ issues) you may not find them to be as useful as one would hope they would be.

....this is other thing I was curious about....what does your EQ page look like for your kick? Using subtractive eq'ing rather than adding gain to the lower frequencies to achieve your goal sound is much more "feedback friendly "

 

[ktedrow]

I set channel gains so that they peak at -6db max and never actually clip, typically running between -12 and -6

I've been running the preamps a lot hotter than this, averaging around 0dB, peaking much higher. This is something I need to look into.

soundcraft built in a soft clipping type or effect so that the it doesnt sound awful when you clip a channel, however it does help induce feedback by having the gains to hot

This explains a lot. I was wondering why clip light is on but I don't hear any clipping. I assumed that there's some sort of built in compression happening.

 

[ktedrow]

Hey guys, a lot of good stuff here that I need to sort through. Really good stuff that I think is gonna help a lot.

Right now I'm off to a gig, with house sound (luckily). Will read through all of it this evening.

Thanks again.

 

[HolmeBass]

TL; DR, so someone may have suggested this, or explained it away, but here is my wild guess:

Low-mids from your tops had enough omni-directional energy to reflect off of the glass and the wooden wall behind the drummer. If he runs his vocals hot then his mic would close enough to that back wall to pick up the stuff and feedback. The glass added enough additional echo to cause problems.

If a material resonates at a certain frequency, it is absorbing acoustic energy at that frequency. You may hear something buzzing or ringing in the room, but that doesn’t cause feedback in the PA.

It’s the non-resonant, reflective surfaces that make feed back so much harder to avoid. Glass reflects for the most part, but the transfer function that determines the fate of acoustic energy at that boundary is almost certainly dependent on frequency, as well as material and surface roughness. My guess, from personal experience, is that glass tends to reflect high frequencies, almost entirely pass low freqs but for a bit it absorbs, and a bit of all 3 (reflect, pass, and absorb) in the mids.

Wood has much better damping, but can still reflect a lot, especially lower frequencies. So given that completely anecdotal explanation I think the wall close behind the drummer was the main contributing factor.

 

[Wasnex]

Quick question for you: Wouldn't reversing the polarity change the phase on that channel? Wouldn't that cause a huge amount of phase cancellation from that mic?

Also, dropping the 185Hz frequency if you can isolate it should stop the problem in its tracks shouldn't it. The frequency feedback loop is caused by the natural boosting of that frequency by the environs, so cutting that frequency should both stop the feedback AND even out the sound because cutting it is pulling it closer to flat. Does that make sense?

Feedback occurs when sound comes back to the mic in phase and has enough gain to cause a feedback loop. Sometimes you can control the feedback by flipping polarity, but sometimes this just shifts the feedback to a different frequency.

Flipping polarity causes a 180 degree phase shift at all frequencies. Mics are normally intended for one sound source, so the polarity of an individual mic doesn't typically cause any major problems. In fact there's a trick when mic'ing horn sections in a big band, to flip the polarity of every other mic.

The reason flipping the polarity may eliminate feedback is because wavelength is related to frequency. Lower frequencies have longer wavelengths and higher frequencies have shorter wavelengths. The sound emanates from the speaker and travels to the mic and the distance can be measured in wavelengths. But remember wavelength varies by frequency. So the phase angle the signal arrives at after traveling the distance will vary by frequency. Perhaps the part you are missing here is there may be different propagation paths. So the 185hz may take a direct path to the mic that is in phase, and it may also take a couple of indirect paths that are also in phase. The arrival times are not the same, but they do arrive in phase so they sum at the mic. Consider 200hz taking the same three paths and it should be clear that they won't arrive perfectly in phase because of the wavelength to distance relationship.

 

[EastCoastThump]

Drum gating was a breakthrough for us about 6 months ago. Before gating we were having a horrible problem with low freq feedback through resonating kick drum head. Gating worked like magic.

Gawd how I hated it when that low freq feedback would start, and get really loud really fast.

Hmm. Lots of potential culprits that are hard to internet-diagnose, but...

[Habitual low-end feedback at patio-band volumes] + [frequent feedback problems at several venues] = sparking my curiosity about your mixer's compression settings.

Over-aggressive compression can summon the shrieking feedback demons, tout de suite.

 

[tadawson]

IMHO, tuning the system for flat response does very little to eliminate feedback. Some programs do give you graphic display so you can quickly identify problem frequencies and notch them out. But when you apply a deep notch to control feedback, you are intentionally modifying system response so it is not flat.

I tend to disagree. One venue Imix is a church with a glass ceiling . . . crazy amput of reflections, and the buildup in the corners is unreal . . . When I got there, feedback management was a challenge. I took my SMAART rig in and tuned and time aligned the system, and simply addressing the peaks in the system to give a flat response has drastically increased gain before feedback, and has pretty much eliminated the problem. Peaks in the system response in the space are where feedback typically starts first . . .

The other thing is to high pass/low pass/band pass channels to suit what they are bringing in. IE, no need for much high freq in the kick, low in the hat, and snare/top can roll off bothnhigh and low somewhat . . . The less the mic hears that isn't 'useful' for that input, the cleaner the overall sound will be, and will sacrifice nothing.

 

[sawzalot]

What kind of feedback? High frequency, mids, lows?

 

[Wasnex]

I tend to disagree. One venue Imix is a church with a glass ceiling . . . crazy amput of reflections, and the buildup in the corners is unreal . . . When I got there, feedback management was a challenge. I took my SMAART rig in and tuned and time aligned the system, and simply addressing the peaks in the system to give a flat response has drastically increased gain before feedback, and has pretty much eliminated the problem. Peaks in the system response in the space are where feedback typically starts first . . .

The other thing is to high pass/low pass/band pass channels to suit what they are bringing in. IE, no need for much high freq in the kick, low in the hat, and snare/top can roll off bothnhigh and low somewhat . . . The less the mic hears that isn't 'useful' for that input, the cleaner the overall sound will be, and will sacrifice nothing.

I'll concede that a well tuned and aligned system is less likely to feedback.

But, even a perfectly flat system will feedback if it is pushed loud enough. When the system feeds back, the response is no longer flat because the feedback adds extra energy to the signal. To fix the problem, you typically begin notching problem frequencies. But you typically don't just dip a little, you dip a lot. So now there is deep, but narrow hole in the frequency response.

A big part of tuning the system for feedback has to do with the microphones' frequency response and polar pattern. So even though you tune the speakers for flat response, the system is not flat because the mics are not flat. Try tuning your system for feedback with all of the vocalist using one type of mic. Switch to a different type of mic and see what happens...typically lots of feedback.

 

[bumperbass]

Hmm. Lots of potential culprits that are hard to internet-diagnose, but...

[Habitual low-end feedback at patio-band volumes] + [frequent feedback problems at several venues] = sparking my curiosity about your mixer's compression settings.

Over-aggressive compression can summon the shrieking feedback demons, tout de suite.

This is what I was thinking.