Change order of Eq and Compress

Qu Qu Feature Suggestions
61

@MikeC

True but I would like to see a bigger picture of what he’s dealing and working with, system, speaker placement, the room, ect. Those factors all add up into the amount of available or lack of gain before feedback.

moderate size room ~~90x60x60 + stage and balcony
podium mike fixed position
one speaker - they are the problem
plenty of gain before feedback
upwards compressing will not lower GFB

The details are not relevant because they only change the possible gain value by some decibels. But any value of gain before feedback is the border you are not allow to pass. It is like the comparsion of a car with or without ABS. No matter wich car you use, there is one speed which will cause that you hit the wall. You cannot beat the physic. Hmm

unless they changed the laws of physics there is no more possibility of feedback problems when the loudest signal is lowered
same thing as lowering the faders to stop the problem

Unfortunately we know nothing about the rest of the system in question, make, models, how it’s been set up and aligned/rung out…if at all or even a general picture to get an idea of what he is working with.
I have asked many times.

Described above. Room is essentially flat. MD rung out problems with EQ on the mikes. Not my first choice but he runs things.

Use what you already have…….
Bring up the mic channel gain a fader level to the level needed for the soft speaking
people, give yourself a little room on the channel gain for the loud talkers.

For the loud talkers find the compressor threshold that clamps them down and yet does not touch the soft spoken people, try a 4 or 5 to 1 ratio and a longer release time so the comp does not pump.

If we do that then we do risk feedback howling. We have no control over the speakers or how loud they suddenly choose to yell.
Nor how soft they want to whisper. The ONLY answer is limiting the DR. We have compression on now but that does not help the low levels intelligibility. This is for an audience hearing spoken word and many of them are old with hearing problems. We need to control the DR and levels so they can understand the words.

I would say that statement is wrong!

It’s used more in the studio world that live, the SQ can do parallel out of the box.

The SQ may do it but some of the people here clearly do not use and and did not understand how UPwards compression works.

62

@MarkPAman

You’ve obviously spent a lot of time reading up on compression techniques, but you need to pay attention to which of them are best used in live sound, and which are only really suitable useful in a recording/post production situation.

I really think you need to read up on acoustic feedback too, as you seem to have got some strange ideas about what causes it. As mentioned above, it’s to do with the overall system gain, not the volume – though a system that’s just on the edge in a quiet room, can often be tipped over by a loud noise.

As you don’t seem to want to believe the folk on here, try this for starters (retype the link):
geoffthegreygeek DOT COM /understanding-acoustic-feedback/
paying particular attention to the Microphone Gain section.

I really like your posts.

But I have to disagree that upwards compression is not useful for live sound.

When we have speakers with extreme DR. [And I noted before that getting guest speakers to talk as we want is not possible]
And papers say for maximum intelligibility we need to set the SPL level and DR based on our background noise (HVAC and to those who say fix that I noted that is beyond our control) then my limited experience (50 years and former AES member, licensed engineer) says we need to limit DR with upwards compression (plus downwards compression too) and then set the level at the answer the intelligibility formula gave [assuming that is less than what could cause feedback problems which it would be].

If you or anyone has a feasible solution that is actually doable then I would love to hear it.

I am pretty sure I understand feedback causes, GBF, gain staging for best SNR, and whatever else folks think is important.
And I understand how specific frequencies can ring while others are not.

As to systems gain, let me replace SPL loudness we want with the amplification levels on the mikes with the Qu.
We can get the target SPL levels we need way below where feedback happens.
We want to lower (or not raise) the peaks of the speakers voltage to the overhead speakers while we increase the low levels.
Increasing those low levels can never cause feedback. Compressing the high levels can never cause feedback unless somehow that compression somehow causes one frequency to be greatly increased (as we are well below the threshold of feedback now).
No one has suggested that possibility. From what I see that does not happen.

I will check your link and read it but I guarantee that our microphone gain is fine for normal speakers.
It is the few that are hams and emote for emphasis which cause the big problem.
The main pastor starts low then gradually increases in loudness as he keeps talking, but without the sudden highs&lows of others. We can and do ride the fader to keep up with him but it would still be nicer to automate that leveling.

If only we had a LUFs option on the compressor:)

63

@Mfk0815

Hmm, they arrive and then tell them „we have some diffucult audio situation. just some simple rules. Use normal voice and speak clearly. If you soften your voice reduce the distance to the microphone, if you soften your voice too much you won‘t be heard. If you speak louder than normal, increase the distance to the microphone. Don‘t point the microphone to your stomach, point it to the mouth, where the voice comes out. Do not mumble. If you obey those rules, everything will be fine, if not your message gets lost. And i believe that god gave you the intelligence to do it right.“
Not even in five minutes per speaker the „training“ is done. It worls for primary school children and it should for sure be possible for grown speakers.

On the mixing console EQ you can lower the lower frequency round about 200-300 Hz and use a high pass and feature a little bit around 2.5 kHz. Use a brickwall limiter to dame to high levels of the speaker but do not use any kind of makeup gain, hidden or not. Explain those simple rules to your volunteers.

Great in theory but totally impossible. We do not talk to the speakers. The audio people are in the balcony. We see them when they come on stage.

Even if we told them they would revert to doing what they always do.

We have HPF EQ and other factors set for normal operation. Expecting volounteers to make RT changes is a bit much even if the MD would allow them to try it.

What is wrong with the simple solution using upwards compression plus the normal down compression we already have on.

maybe we have to explain you the world of live audio again. Feedback has nothing to do with SPL or how loud the speaker is talking. It only is related to the microphone, the loudspeakers, their position relation to each other and the gain of the signal (not the signal strength or SPL). You know for sure that if you increase the overall gain of the mic it will cause feedback also if nobody is talking into it. There is a certain amount of gain you are not allow to reach, the gain before feedback. Otherwise you will cause that feedback.
So what increases the overall gain? first of all the input gain, then also, of course, the channel level and the main level (positions of the related fader). Also the EQ can increase the gain on some bands, the the make up gain of a compressor can increase the overall gain. and, now it comes, you say

Upward compression, works from the opposite end of the spectrum: when a quiet signal falls below a predetermined threshold, it’s brought up in level. …

it’s brought up in level is just another way to in increase the overall gain of the signal. So any, A N Y, process which increases the overall gain can cause feedback, no matter how much SPL the signal has.
And that’s also the reason why we train the speakers to speak as loud as possible and not to speak inn a great range of loudness. if the input signal is very loud, you do not need to increase the overall gain so much to make that signal good hearable. If the signal is soft you increase the overall gain, and that cause feedback.

Let me make it clearer. orators level → mike out → mike gain gain → preamp → A/D → Qu32 process → D/A → voltage —> amp → hanging speakers → SPL levels → also some back into mike → may feedback or not depending .

replace what I said about SPL with gain if that makes you understand what I say better.

The level of the orator does matter. When they yell we get more fed back into their mike.
But that is NOT the problem.

The real problem is hearing them when they are too soft.
We have the high end compressed and gain set quite well.

How does increasing a low signal to a medium signal that is way below the highest signal going to cause feedback?

We are NOT increasing the overall gain!! We are increasing gain where it can NOT cause feedback.
The overall gain will likely be less to hit the optimum SPL level for intelligibility.
64

Sorry but I am out. My understanding of a community is that each member should to the best to serve the community. And also if someone is here to tell others his thoughts will normally do whatever is needed to do the job successfully. Anf if that means to obey some simple rules, so what. Are they the pope or higher in the hierarchy of your church? I do not understand such mindset, really.
And if you are not able to understand that „it‘s brought up in level“ is exactly the same as „raising the gain“ no one here and in the rest of the world will be able to tell you the mistake in your thoughts.
And as long as the upward compressor, your glorified remedy for your problem, is not available for the QU-Series you will stay in that situation because it seems that you do not even allow other solutions to be discussed in your community. Or not even you will seriously think about that solutions.
So stay as you are and maybe you find some minutes to read a small little book called „The Situation Is Hopeless, But Not Serious: The Pursuit of Unhappiness“ written by the great Paul Watzlawick;-)

65
We have HPF EQ and other factors set for normal operation. Expecting volounteers to make RT changes is a bit much even if the MD would allow them to try it.

It sounds like there is no trust between anyone on your AV team and the MD…maybe with good reason!

Like Mfk0815 I’m out as well.

I do encourage you though to join up at Prosoundweb.com and enlighten everyone there with your vast knowledge.

66

@MikeC

the MD management style is cautious.

Not going to enlighten anyone that way unless they pay my consulting fees.

67

@Mfk0815

What part of bringing up low levels is not raising any gain that matters for feedback is not clear to you?

Everybody seems to stuck on what they know about downwards compression which is the usual way of doing that and not actualy grokking what/how UPwards compression works nor where it works at and what it actually affects wrt other factors like GBF.

We are not raising the gain on the down compressed signal as you and others normally do as ‘make up’ gain.
The problem is not the loud signals.
The problem is the low signals caused by the orator and his personal ‘style’.
Upwards compression is the best answer I have found to fix that problem.

I would love other solutions if they were feasible.
Training the speaker is not an option.
Telling guest speakers how to talk would be good until they forgot. But we do not get to tell them anything.

I would like to think that if we were in a room and I had a blackboard to draw pictures on that you might understand where your confusion is and see you are not talking about upwards compression but other compression problems causing feedback.

68
I think you are confused about something.

that could be, but not on compression

You will need to do more than assert your claims

Wich claims? I tried to point you to some knowledge…

The only way to get feedback after regular down compression is to add makeup gain

I was talking about upward compression only, as stated many times in the above posts

Upward compression, works from the opposite end of the spectrum: when a quiet signal falls below a predetermined threshold, it’s brought up in level.

That’s what we talking about, all the time and this will cause feedback and some other unwanted side effects

mathematics with SPL are the wrong way to describe feedback or compression
a turn on the volume knob will change your complete calculation
you have to use electrical values… dB not dB SPL to describe compression

This is a case for upwards compression—use it to bring the quieter bits of the vocal UP in level.

and what happens to the ambient noise floor?
this is normally lower than the lowest signals,
an upward compressor will raise the noise floor as well (because it is under the threshold)
this will drive your loop reinforcement over 1 and then you have feedback caused from noise floor…

to compress the loud signals to prevent the system from feedback
gives you an unusable signal since your ambience signal is driving your upward compressor to full work and the limiter is bringing this down…
if you change the order, first peak limiting and then upward compression it will lead to the same problem as above without limiting

the problem is that the upward compressor is doing things that are not managable in a live sound scenario
the lower the signal goes the more gain is introduced with upward compression

@ Ratio 1:2 6db under the thershold will introduce 12db of gain, 12dB under threshold will introduce 24dB of gain, 18dB under thershold will introduce 36db of gain

The Fabfilter guy says it is mandatory to limit the gain compensation in an upward compressor for this case.

but if you are close to the point of feedback this makes the upward compression still a higher risk then regular compression

The problem lays in the point to show you this because no live sound mixing desk is implementing this type of compression
because it is dangerous in live situations
so there is no chance to show you that on a real physical device

and I know only one device that implements it as hardware at all… the already mentioned OmniPressor

EDIT: I was writing this post in the last hour
and saw that Mr. Adams answered some more topics with unqualified comments

I’m out here as well

69

@williamadams

What part of bringing up low levels is not raising any gain that matters for feedback is not clear to you?

this is unbelievable… bringing up low levels is close enough to raising gain
the signal is getting louder

you have to stop this

70
But I have to disagree that upwards compression is not useful for live sound.

you’re the only one, ask yourself why? and the audio industrie did what in the last 75years?

71
We are not raising the gain on the down compressed signal as you and others normally do as ‘make up’ gain.

you will not compress it down… you will compress it up and this will change the level upward
that is the entire problem

low levels get raised… I don’t know how anyone can ignore his own words with that… don’t know a word for… I’m stunned (learnd a new word :-))

72

Arghhhh,
downward compression lowers the output for higher inputs signals. Nothing more.
As you always say, upward compressor bring up the signal for lower levels. Nothing more.
And now tell me which compressor raises the gain based on the level of the input signal.
No makeup gain, no extra gain on preamp, no changes on the channel fader or any bus fader.

Training the speaker is not an option. Telling guest speakers how to talk would be good until they forgot. But we do not get to tell them anything.
Everything is possible, nothing is unchangeable. At least in the interaction between human beings with a reasonable amount of IQ and social skills. And your speakers are not god-like. They should know that they can fail and that someone knows something better. You said that there is a fixed position for the speaker. Put a sign with the simple rules on this position so that it always remember the actual speaker what they have to do. There are a lot of options to change the current situation which are easier to implement than the upward compressor for the QU-series.
73

If I can train a DJ not to send a distorted “all the red lights” signal to my PA, and to trust me to make it loud, then anything’s possible!

???

74

@#SteffenR

I will agree that upwards compression is not useful for the live sound you do.
But can you understand that for our situation it would improve the intelligibility of the speakers for our audience?

Tell me how this is wrong:

We tune the room EQ the mikes and do all that good stuff you like
then we raise the mike gain in the preamp until it just sings gently.
Then we back it off as many dB as you like for a safety margin.

Now we back it off another ten dB.

So our mike gain is set at a point where the speaker can speak until the mike puts out 10dB over the existing noise level with no compression of any kind.
And we still have our safety margin.

The speaker can speak from anywhere at nothing to a loudness causing 10dB signal.
If the speaker is whispering and causing 1dB of signal then could raise the gain in the preamp 9dB and still no feedback as we are under our safety margin. But that would be risky because as soon as he talks louder the system howls.

Or we could use an upwards compressor that raises that signal till it shows 3dB and still no feedback.
Note that we did NOT increase the gain in the Qu. The mike is louder but well below his louder talking that causes no feedback.

Now if the speaker is talking loudly and putting out a 10 dB signal and we compress that down to only 7 dB then there is still no feedback.

What we have done is change his wide range from causing 1dB of signal to 10 dB signal to now be constrained at 3dB to 7dB.
Not only no feedback but, even more room in gain before feedback in case he yells.

I guess if you were paranoid you could add another limiter to ensure nothing was ever too loud to trigger the feedback problem.

Try drawing a signal level diagram at every point in the chain and you would see that upwards compression will not cause feedback. It is not different from the speaker speaking up instead of whispering.

Then see that regular compression will not cause feedback if you do not add make up gain .
If you are paranoid you could EQ after all the compression to be extra sure it wont – then add a limiter at the end too:)

75

@SteffenR

you will not compress it down… you will compress it up and this will change the level upward
that is the entire problem
low levels get raised… I don’t know how anyone can ignore his own words with that… don’t know a word for… I’m stunned (learnd a new word :-))

It only changes the low level upwards, not all of them.
We lower the high levels with regular compression.

Have you looked at the diagrams or any of the links that described how this works to reduce the DR?

But I have to disagree that upwards compression is not useful for live sound.
you’re the only one, ask yourself why? and the audio industrie did what in the last 75years?

I did not say it was good for everything. I said our sound is live and it would work for us and our audience.

as to your long post
yes you do have to consider noise both internal and external but upwards compression does not mean you will have feedback, although it will limit the amount of DR reduction you can achieve.
76

@Mfk0815

If you looked at the diagrams with the links I posted andor read the description you would know that the compressors are set
so that low signals are increased and high signals are made lower and that it is normally done by heavily compressing the signal and then adding in the right amount of the un (or very lightly) compressed original signal often done in parallel with two channels then summed.

The net affect is that the highs of the signal get pushed down while the lows get raised up. And the net is in the middle of where it was originally. The DR is reduced. There is no risk of feedback.

There is always noise and other practical considerations but that limits how much you can do not mean you cannot do it at all as some seem to think.

77

You can repeat your stuff as often as you want, it will not become true that raising the gain in one or the other way do not increase the risk of feedback in live situation. Period, exclamation mark.
A standard (downward) compressor cannot cause feedback by itself, because it lowers only a signal. But if you combine it with a second processing step (call it makeup gain) the risk of feedback increases.
Of course you can use all these processing options, but you have to be aware of the potential risks. But if you say „no this upward compression stuff will never increase the risk of feedback“ you are completely wrong. You have to learn that truth the one or the other way. Afaik the QU32 is equipped with sub groups, so you can do some experiments with parallel compression in live situation. Do it an learn your lessons.
Period.

78

@MarkPAman

If I can train a DJ not to send a distorted “all the red lights” signal to my PA, and to trust me to make it loud, then anything’s possible!

I suspect you had access to the person and they actually listened to you and tried to change.

We do not have the luxury of talking to the orator nor would they listen to us if we could.

79

@Volounteer - This thread’s got very long, and some of the info’s in other threads, so forgive me if I’ve missed or forgotten something.

You said:

Tell me how this is wrong:

We tune the room EQ the mikes and do all that good stuff you like
then we raise the mike gain in the preamp until it just sings gently.
Then we back it off as many dB as you like for a safety margin.

Now we back it off another ten dB.

So our mike gain is set at a point where the speaker can speak until the mike puts out 10dB over the existing noise level with no compression of any kind.
And we still have our safety margin.

The speaker can speak from anywhere at nothing to a loudness causing 10dB signal.
If the speaker is whispering and causing 1dB of signal then could raise the gain in the preamp 9dB and still no feedback as we are under our safety margin. But that would be risky because as soon as he talks louder the system howls.

That’s not normal - it is wrong, but why?.

Putting aside compressors of any sort for now as none would cause this.

There’s something not right here, a louder input from your speaker does not, by itself, increase the system gain, as the speaker is not part of the loop. You should not be getting feedback in this situation. But evidently, something must be increasing the gain if you get feedback.

So, I wonder what is causing this?

I think you’ve said that mic is in one place and not being moved or cupped?

I’ve a couple of guesses, but that’s all they are!

I think you’ve mentioned somewhere in the past that you think the AMM is used for some mics on the desk? Is this mic one of those?
If, as your speaker gets louder, the AMM is turning down other mics while turning up this one (keeping their combined gain the same), that could be causing this. Especially if the other mics are further away or back from the PA, so less likely to contribute to feedback. So check the AMM settings, (if you can access them?) and if this mic is in the AMM, take it out.

I’ve had something like this happen when the room’s installed induction loop system has somehow been picked up by something and fed back into the desk*. May well have been the desk itself (which was old & analogue). I don’t know why it only became a problem at higher volumes, maybe the loop had some sort of gate it?
This is easily tested by turning off the loop, though that’s rarely an acceptable long term solution, you will at least know why you have this problem.

  • Some guitars seem to be particularly good at picking up induction loops, though any unbalanced part of the system can contribute.

There could well be another cause, but it’s difficult to know what else to look at without more info.

80

I suspect you had access to the person and they actually listened to you and tried to change.

We do not have the luxury of talking to the orator nor would they listen to us if we could.

Yes, of course I was able to speak to them, and I’ll admit that it took time - several events. Take time to get to know them, be their friend. We both want the same thing - the best sound possible for the audience.

Show them that you’re not there to be awkward or spoil their fun, but that in fact it will sound better if if the equipment is used as it was designed to be.

You speakers will want their message to be clearly heard, soo show them that that’s what you want too.

When I do PA for large meetings with inexperienced speakers, I tell them to talk to the first few rows as if there’s no mic there, and I’ll take care of getting them heard at the back. Most people can gauge how loud the need to be, and an appropriate DR when there’s no PA, and if they can ignore the PA and aim for that level at the front, the PA job becomes much easier.