I need a solution for live PA and recording

[Greg Miller]

Originally Posted by Marcus Marchesseault

Greg, I was not expecting such a detailed understanding of our situation from what I wrote.

That's why I said it sounds impossible. ;-)

Originally Posted by Marcus Marchesseault

The biggest problem I've had so far is handling noise from the lavs as one presenter switches over to another. If both lavs are on one track, it makes it impossible to remove the handling noise and other unwanted strays from the unused mic.

How do you address that issue with the PA mix? I'd hope your PA operator rides gain and mutes the presenter mics as needed. If so, why not just feed the finished PA mix to the camera. Of course if the PA operator screws up, you would not have the opportunity to fix it in post.

If the operator is reliably mixing the presenter mics, then ideally you would feed the PA mix plus the gain-ridden room mics to the camera.

You'd handle this in the broadcast world by using the "mix minus" feed (which is normally sent out to a to remote talent through a telephone hybrid) and sending that to the PA amps. Then you'd connect the room mics to the channels normally used for the return (talent audio) from the telephone hybrid. The operator would ride gain on the presenter mics in normal fashion, and also ride gain on the room mics when they are needed in the streaming/camera mix (but of course they would not appear in the "mix minus" which feeds the PA amps).

[Rick Reineke]

As was previously stated, mic splitters should be the most reliable, either multi-tracking the inputs or having an experienced operator mix it 'on-the-fly'. Other than the mic splitters, wire custom plugs for the house mixer's insert jacks. ( I would concur inserting standard plugs half-way is not recommended)

[Marcus Marchesseault]

Rick, I think I'm going to start by getting a better mixer with direct outs and sub groups. I also want something quieter and Mackie claims the VLZ3 mixers have better mic inputs. It seems that for everything they do that a proper mixer is key.

Greg, I can't count on a good operator running the board. Heck, in a 4-hour lecture, I don't think I can trust myself to be on the money all the time. This is why it would be ideal to have a multi-channel recording in case the camera mix is too bad to use. Things could be re-mixed in post when timing and operator experience are not so critical. Of course, mixing multi-track audio in post from a long lecture would be overly time-consuming on a budget, so an acceptable camera mix is most important. In the case of this location, the audience mix is not so critical for the video and mostly serves for ambience over the live stream. What is most important is the lav and handhelds on the presenter. This is why I think I may have luck with a camera mix to two channels that separates the lavs and excludes the room mics.

Another point I should bring up is that the presenters are somewhat accustomed to all this stuff and they can often be depended on to mute their mics when they are not speaking. I'd rather have one person there all the time who really knows their stuff and have the board run by a pro, but this place does not work like that. They spread out responsibility into the community and I'm hoping I will be able to adjust to that and figure out how it can be advantageous.

[Greg Miller]

Marcus,

First, let me confirm that I agree with Rick and you: the best audio solution is to multi-track, then mix in post.

But I guess I'm a bit confused by your priorities. I thought you said earlier that the long-form material makes mixing in post prohibitively time- and labor-intensive. But now you seem to be saying that you do want to mix in post.

Also, I thought you said earlier that you could handle the live audio mix yourself, because the camera-op requirements are minimal; but now you seem to be saying that you can not handle the mix yourself, no do you have a reliable operator.

Anyway, with this confusion, I don't know what to say, except that you seem to have a grasp of the problems and pitfalls involved, so... good luck with whatever you decide to try.

[Paul R Johnson]

The idea to have a multitrack doesn't really remove the need for an operator who knows what they're doing? You could use limiters on each channel then leave the mixer to itself, but surely if the project is worth doing, then doing it properly is by far the best bet - then you could mix to stereo or multitrack, whatever you fancy, but it's a bit like not using a cameraman, and putting the camera on wide and turning auto everything on.

[Greg Miller]

Paul:

I tend to agree with you. Ideally we want to do everything in the best possible way. A good recording operator, and then a mix in post, would be ideal. But based on the OP's comments, I don't think that's going to happen.

Unfortunately, I've been involved in several church projects in the past, and my assessment is that they (the church folks involved in these specific projects) felt it was OK to break all the "best way" rules, because somehow God would magically fix everything "in the mix." Sometimes these projects got lucky with their volunteer help; sometimes the results seemed pretty bad to me but the church folks seemed to be either very unaware of the quality issues, or very forgiving of what their fellow parishioners/volunteers had done.

So I've been trying to temper my comments here: a bit of realistic advice, and also a bit of advice that might change a "worst-case" outcome to a "somewhat acceptable" outcome, without implying that "acceptable" is as good as "the right way."

[Marcus Marchesseault]

Greg, I do want to mix in post when I have time and I also would like the multi-track mix there in case of disaster with the camera mix. Of course, if there are problems with one, the other is probably also trashed. I think I'm going to first try to get the PA working as good as possible and bring a sub mix to the camera. I can do a fair bit of magic in post quickly if I get the two lav mics that battle each other on separate channels.

You guys do have the right idea, and I do have difficulty ahead. The moderate camera operation needs will allow me to supervise the mix, but that isn't the same as having a trained set of hands on the camera and another pro on the mixer. Even recording all the channels and mixing in post doesn't solve things, but it could be useful in certain situations. For now, I am going to get a more capable mixer and better lav mics to deal with the worst issues. I also plan to add compression/limiting capability to hopefully get things to acceptable levels. I have also gone to the room and figured out a decent eq setup that gets rid of most of the ring. This also helped the clarity over the PA as those near-feedback tones muddy the sound.

On the personnel side, I can start putting out some training on how to use mics and can find helpers that can hopefully grasp the basic aspects of mixing. I will be close by, so a person that has some technical aptitude can be helpful in fixing problems (even if not as fast as I would like). This is still a huge challenge and I appreciate the help.

[Greg Miller]

Originally Posted by Marcus Marchesseault

I also plan to add compression/limiting capability to hopefully get things to acceptable levels. I have also gone to the room and figured out a decent eq setup that gets rid of most of the ring.

You might want some peak limiting for the times when the presenter screams before someone can grab the pot. Otherwise it should not be doing anything.

I would recommend against compression, as that will only bring up the gain at times, which will exacerbate the feedback (and noise) problems.

Get yourself a Sabine automatic feedback eliminator and you'll never have to worry about that problem again. But be sure it's installed outside of the recording chain... it should be just ahead of the PA amplifier.

[Marcus Marchesseault]

Righto on the limiter. I only want to keep things in check. Strong peaks contribute to reverberations in the room that contribute to feedback. I want a fairly level signal and will try to get that by better mic choice and a proper mix. I'll look into that Sabine feedback controller. I've been reading about feedback controllers a bit and internet scuttlebut is mixed on their performance with a number of people saying they decrease sound quality too much but others saying they do a great job clamping down on feedback.

[Greg Miller]

I installed Sabine FBEs in several large ballrooms that had overhead reinforcement speakers... a nightmare because the mic was invariably directly below the speakers! The Sabines clamped down on the feedback and ended my problems. (And talk about unskilled operators... that job was a nightmare.)

Of course you want to keep the FBEs out of the recording chain, but ahead of the PA power amps. If you are feeding much recorded music through your mixer, then you might want to create a sub-mix of just the live mics, and feed only that sub-mix through the Sabines.

And of course a pink noise sweep and EQ of the room can get rid of some feedback peaks ahead of time, so the Sabines aren't working as hard. But with people moving around wearing lavs, the distances are constantly changing, so the critical frequencies are changing. That's where the Sabines can work miracles that fixed EQ cannot approach.

I still remember being impressed. With the room empty, I'd slowly bring up the gain until I heard things start to ring, then I'd hear that one hot tone go down in level. I'd bring up the gain some more, and hear things start to ring at a different frequency, then I'd hear that tone go down in level. Etc. Etc. The Sabines are multi-notch devices so they can clamp down on more than one resonance at a time. Miraculous.

[Paul R Johnson]

Feedback suppressors have one major drawback. They need to hear feedback to be able to remove it. In live sound, you can hear the ringing that happens before feedback starts and back off. A good feedback suppressor hears a sudden howl, analyses what frequency it is and slaps in a big high Q filter - but it needs a decent burst of howl to work. The multi filter devices then lock on to the next howl and zap that, so for emergency use - where perhaps people may accidentally move too close to a speaker, they can be useful - but for everyday use, they're not that good - because after all those notches are in, the sound isn't exactly smooth.

[Marcus Marchesseault]

That's another nail you hit on the head - overhead speakers. I know there was much improvement after ringing out the room, but I suspected that the roving lavs would cause problems too dynamic for an eq to fix. That's a good idea putting the mics on a submix so any effects don't interfere with recorded music. Do mixers with sub channels usually have an effects return just for the sub channels?

[Greg Miller]

Paul:

My impression of our Sabines was that, in real use, they could notch out any building feedback faster than an operator could have done. And this happened long before the feedback developed to a "howl"... it happened when the ringing was starting to be noticeable.

Besides, an operator could only pull down the fader, which would certainly create audible issues. The FBE was smart enough to pull down only the offending frequency(ies) without lowering the broadband gain setting, so it wasn't especially noticeable unless you were an audio guru listening for it.

I'm honestly not sure whether the FBE keeps the notches forever, or whether they eventually return each filter to unity gain. But even the units I worked with (over ten years ago) could be operator-set as to the maximum number of notches. If I wanted to, I could set them to have just one or two notches. Presumably, when a new FB frequency started to ring, one of the old notches releases back to unity gain, and is then re-tuned to the new ringing frequency.

Given that the presenters were constantly moving around, and the frequencies were constantly changing, I would take one of those Sabine FBEs over an operator any day. Well, OK... maybe a very experienced Broadway-level operator could do almost as well. But still, I think notching was greatly preferable to wide-band gain reduction.

Marcus:

I don't know what most mixers "usually" have... there's such a very broad range of mixers in the world. But you clearly have enough understanding of the situation, and of equipment specifications, to look for this specific capability and to specify something appropriate.

[Paul R Johnson]

In live sound they're treated with distain by most ops - I've got 6 in my foh rack and never use them. If you have static mics, then you can push the gain and let the notches handle it, but when people move, as with radio mics, the problem frequencies move, and when you run out of filters, you're a bit stuck, and releasing them often makes things worse. The soft lock doesn't work on ringing until a distinct note starts, which an op normally never lets happen - worst still is that in a band situation, spill from instruments can be mistaken for feedback, so the guitarist holds a note, and the vocal mic mistakes it for feedback and notches it out!

Some people leave them in circuit and just let them work. I'm never convinced this is a suitable way to work and looking at the eq curve produced they do very odd things to frequency response. Some people swear by them, in the same way that people love limiters, but these are just excuses for lazy mixing - and with more than a couple of mics on the go, a sound op is not a luxury, it's essential.

[Marcus Marchesseault]

Paul, I definitely understand what you are saying and I know I'll have to be careful not to abuse a feedback controller. Our situation seems to be ideal for their use as we rarely have any sustained notes. We only have talking. We also have a room that is a nightmare for live sound. It has almost all hard surfaces and some traffic noise. I need to get something audible from the lav mics out to everyone at a level they won't have to strain. Currently, with the room "rung out", it is just barely workable and the recording is crap. If I can get the live mix to a point where it can be self-sufficient most of the time, I can deal with it and get someone to help ride levels. With a bit of limiting and feedback control, I think I'm in business. I think I can learn how to deal with the feedback control and reset it during breaks. I definitely heed your warning. I will also still use some eq and try my best to get a good clean mix overall. I don't want a bunch of crutches, but I have a tough scenario and absolutely can not have a giant feedback howl interrupt what is mostly a quiet ceremony.