Compensating for lousy miking

[Greg Miller]

Adam, I really feel for you! Whoever is issuing the edicts has lost his grip on reality.

Do they light the show by turning up the gas jets? Hell no, they're using lekos and fresnels. They're probably coloring with gels, probably dimming them for different lighting cues. Does anyone object to that anachronism?

When intermission comes, or the show ends, does the audience complain when the electric house lights come up? Hell no. Nobody misses the smell and oppressive humidity of the gas lights.

In summer, do they cool the theatre by having a celler under the auditorium filled with blocks of ice, and vent holes cut in the auditorium floor under the chairs? Hell no, they use electrically powered HVAC. Does anyone complain about that?

You have a psycho on your hands.

For future reference, I think Jay is right about problems mixing four boundary mics. Years ago, I tried micing a Broadway show, just for a tiny bit of reinforcement (it was a legitimate play, and all six actors could actually project, but we needed some fill under the balcony). I learned quickly that boundary mics weren't appropriate for that job. When the actors moved around, the phasing problems were quite obvious. And, as you've noted, footsteps and other stage action were quite noisy, too. I solved the problem by using a number of shotguns mounted on the railing of the orchestra pit (there were no musicians to worry about), aiming them at the places on the set where the action took place, then selectively fading from one mic to another when a scene change or stage direction so dictated. I never had more than one of those mics open at any given time. Unfortunately, that's not an option for you.

I suspect any major change in the future would best be implemented by multi track recording, with a mixdown later.

For action on stage, I wonder if you might get away hanging some choir mics on the upstage side of the first teaser (assuming you can avoid problems from dimmer buzz). Of course that's too close to the action to get a good stereo perspective from one pair, so you might need several mics, aimed roughly at action areas on the set. For overall mix, a pair of choir mics hung 15' or 20' back from the lip, at around the height of the proscenium, might work fairly well, and should be virtually invisible to the audience. None of this will sound like Annie or Phantom (thankfully, IMHO) but it might give you a decent archival recording.

{Alternatively, put five mannequins in the front row center seats, and rig the middle one for binaural. Now that's archival.}

Thinking about your present recording, I do have one suggestion. Create two different stereo mixes: one for musical scenes, and one for dialog.

Treat the dialog track fairly aggressively to get rid of room noise, boominess, etc. Your goal should be to have good intelligibility, even if there's a slight sacrifice in overall frequency response. Listen to the mono mixdown; if necessary make the dialog track mono, to avoid further phasing problems if it's played back somewhere in mono. Ride gain as needed so all the dialog is clearly audible.

Treat the music track for the best possible sound quality... which probably means less treatment, less level adjustment, etc. Then let your final mixdown be a combination of the two, fading from the dialog track to the music track, and vice versa, as the musical numbers start and end. Of course you can't over-process the dialog track; you want "Jimmy" singing to sound like the same person as "Jimmy speaking." Yet performers use a different voice to speak and to sing, so the audience will expect and tolerate a certain amount of difference.

When you think about it, most traditional musical films have dialog recorded on the set with one batch of gear, and the musical numbers recorded in the scoring studio with an entirely different setup. There is bound to be some mis-match between dialog scenes and musical scenes, but the audience doesn't stomp out and demand refunds. So if your final track is a collage of different tracks, it can be quite acceptable if it's carefully and tastefully done.

[Greg Miller]

Originally Posted by Colin McDonald

In one case I ended up with an on-screen apology and subtitles

Great idea... an entire high school musical with no audio, just subtitles. Or perhaps just key in an ASL interpreter in the corner of the screen, and ask the audience to learn to sign.

[Adam Gold]

Greg, thanks for the supportive post and great ideas. I'm not sure I'd classify them as psychos but it's certainly been a challenge. I just think there is a certain degree of territoriality in academia and they're not crazy about a parent who happens to have had a career in this biz intruding into their domain, despite the fact they're getting about fifty grand worth per year of pro bono stuff. Some people give money and we do this. I spent 25 years as a network TV executive and this is my way of giving back and atoning for my sins.

Not surprisingly, it's the newest hire/youngest/least skilled faculty member who is the most problematic. On our very first show, I happened to mention that it was a bit dark onstage and we couldn't promise much, picture wise (this was well before my lovely Z5s) and the dressing down she gave me, including the phrase "creating a theatrical experience," would make your hair curl. Now I just slink in invisibly (with four sizable rigs) and shoot silently.

I grew up in, and spent most of my career time, in LA, where everyone is pretty TV savvy and lots of people want to be in the business and will bend over backwards to help you. Up here in the boondocks, they just couldn't care less.

I like your idea of treating music and dialogue differently, and as it turns out what I did sort of worked out that way. As the dialog tends to be quite low on the lesser-treated track, the more treated one tends to dominate, while the music comes up quite nicely and fills in the shortcomings of the highly processed track.

Steve, as I think I noted in the first line of my first post, I know you're right and we can't expect miracles. But there are dozens of ways to process sound in all the software packages out there, and they all do something, and presumably at least some of them can affect audio in ways that make it sound better. I was just hunting for some ideas. As it turns out, the final mix isn't too bad and I'm the only one who considers it sub-par. Well, you guys would too, but the civilians who get our DVDs realize they are several orders of magnitude better than the ones they used to get before I started doing this, and are all pretty nice about it.

My youngest is there until 2018. Maybe by then they'll let us mic up the place properly.

[Colin McDonald]

Originally Posted by Greg Miller

Great idea... an entire high school musical with no audio, just subtitles. Or perhaps just key in an ASL interpreter in the corner of the screen, and ask the audience to learn to sign.

Ha ha! No, that's not what I did, and it's not what the Beeb do either. The idea is to give the audio postproduction your best shot (some good ideas here now like your suggestion of music and dialogue mixes) and only where the intelligibility of dialogue is in question add the subtitles. It wasn't a whole school show that I had to do this for, it was for much shorter extracts. A lateral "solution" obviously.

For the small theatre I do most of my school recordings in now, I have a trick which works well, at east for the stage area. There is a very conveniently placed and electrically raised/lowered projector screen mounted over the front of the stage. I attach Sennheiser G2 radiomic trnsmitters by taping to the bottom rail and hanging MKE 2-ew Gold mics down from them over the front of the stage. These are invisible to the audience and do a pretty good job when properly mic placement can't be done - usually in my case because its a last minute request to record something (don't get me started on that topic).

I think we all know here that there is no substitute for proper mic choice and placement.

[Greg Miller]

Originally Posted by Colin McDonald

and only where the intelligibility of dialogue is in question add the subtitles.

No, really, you should subtitle the whole play. In fact, during the musical numbers, instead of static subtitles, you could use the "bouncing ball" effect and the viewers could sing along. It it weren't so time consuming, I'd at least produce one copy like that, and give it to the young genius who doesn't want any mics used.

[Steve House]

Originally Posted by Adam Gold

....Steve, as I think I noted in the first line of my first post, I know you're right and we can't expect miracles. But there are dozens of ways to process sound in all the software packages out there, and they all do something, and presumably at least some of them can affect audio in ways that make it sound better. I was just hunting for some ideas. As it turns out, the final mix isn't too bad and I'm the only one who considers it sub-par. Well, you guys would too, but the civilians who get our DVDs realize they are several orders of magnitude better than the ones they used to get before I started doing this, and are all pretty nice about it.

....

Oh, I don't doubt that audio processing in post can make reduce some of the problems, just that the handicaps you're saddeled with gives you a starting point that doesn't have much to work with, making it hard to get it very far up the ladder.

I'd wager the teacher that reamed you about wanting to diminish the "theatrical experience" has never HAD a theatrical experience in a major theatre town like New York, London, or Toronto.

[Adam Gold]

Originally Posted by Steve House

... the handicaps you're saddeled with gives you a starting point that doesn't have much to work with, making it hard to get it very far up the ladder.

Granted and stipulated, your honor. But I guess I was hoping you could give your thoughts on... what would you do if you found yourself in this situation, just to get a couple of rungs up the ladder. I have a hunch you wouldn't throw up your hands and say nothing could be done. I'm guessing you're too much of a perfectionist to let it stay as is.

But anyway, I'm happy with what I was able to achieve. Just to close the loop on this, here's the post by Jon Fairhurst in another thread with an idea for solving a different problem, which was the idea I used to help out my tracks:

http://www.dvinfo.net/forum/all-thin...ml#post1636115

So thanks, Jon. You helped me out of a jam without even knowing it.

[Steve House]

Have you tried this? ... You said you have miced the stage with 4 boundary layer mics and you're mixing them down to 2 stereo tracks while recording, right? You might consider working out your recording devices so you can record each mic on its own isolated track and forgo altogether mixing on the fly. Because of the differing distances from a given source to each mic, each particular sound arrives at each mic at a slightly different time. Mixing them directly means you're mixing waveforms that are slightly out of phase with each other, inducing comb filtering that makes the already bad situation worse. Rather than mixing all 4 down to 2, treat the sound from the stage as mono, while in post panning from one mic to another as the sound moves around, only using the track from the one mic that has the best rendering at any point in time without mixing in the other channels.

[Greg Miller]

I agree with that, Steve. That's why I suggested, for future events, that he record multi-track and mix later.

However, that doesn't help with the present mixing problem. Actually, I get the impression that Adam has reached some level of satisfaction for now. After all this discussion, I'm rather curious to hear how bad or good it actually sounds...

For the future, I am torn about making the dialog mono vs. stereo. On the one hand, "real" films have mono dialog, which avoids problems with phase cancellation etc.; but that assumes a clean track to begin with. On the other hand, recording and playing in stereo can sometimes help intelligibility, because it sometimes lets the brain decode the stereo information and "tune in" on the desired sound (the dialog). This "brain filtering" works best with binaural recordings played in headphones, and it somewhat sacrifices mono compatibility. So I hope Adam can prevail and get a better miking situation next time around.

I wonder whether his high school has an old set of footlights, which might be unused presently. If so, that would be a good place to hide a few mics.

Now if I owned that school, I'd drill a bunch of 3/8" holes along the lip of the stage, and stuff them with Panasonic electret capsules. Of course I'd need to get down under the stage to run the wires... minor detail. However, there would be massive problems with impact noise. In reality, I think multi-tracking with hidden choir mics is the best solution given his constraints.

[Adam Gold]

As Greg very correctly noted, all these recording tips are interesting and well-intentioned, but I already know all that stuff -- I've been well-schooled by Steve and others in the many other threads I've posted about miking techniques -- and if I could have done any of it, I would have. But I couldn't then and likely never will be allowed to. Believe me, on my list of the five best ways to mic this sucker, the way we did it isn't even in the top ten. And phase issues were the very least of our problems. In the future we can put mics on stands as we've done in the past -- it just wasn't an option this time. Greg, which specific "choir mics" would you recommend?

Ironically, in replicating an old-time theatre, they actually built some very nice footlights for this production, but they were actually being used as footlights so no room for a mic. The nerve.

In a way, by tweaking the audio we got to be at least minimally acceptable, I may have undercut our argument that we need permission to properly record in the first place. Oh, well.

Greg, if you're really a glutton for punishment, shoot me your email and I can send you some samples. The one-minute wavs I pulled are a little too big to post here.

[Jay West]

Adam:

I just stumbled onto this thread this morning. Had I found it earlier, I would have recommended the the techniques described in the Jon Fairhurst thread. I have been using variations on those for for more than a decade of shooting in circumstances like those your describe. From that experience, I offer some additional suggestions that may or may not help.

One trick I've used when the recording levels were low is to double or triple the tracks. In PPro, I've found that this can make quiet recordings seem louder without getting the noise or distortion that might otherwise result if you simply boosted volume levels. I'm not sure why this works but I have found that it can help a lot.

If you had any centrally located mikes, PPro and Soundbooth (dynamically linked) can sometimes help things along by doubling the track, applying the fill-right effect to one and fill left to the other, and adjusting the panning in the audio-mixer pan controls.

You mentioned that the room sound seemed very good during the performance. You also said something about having a couple of cameras to the sides of the room. Have you checked the audio tracks from your cameras?

Maybe they got audio that can help fix up the sound when properly panned and mixed in. If you mix the room sound with the stage sound recorded by your H4, and judiciously adjust the panning (left cams to the left, right cams to the right), you may be able to fill holes.

I think I recall you using Sony Z5s or FX1000s. If there was house amplification, maybe you had a camera close enough to a loudspeaker to get usable audio without a lot of audience noise.. Since these cameras have have audio limiters (as opposed to the more common audio AGC) you may able to use that audio to patch around or paper over overloads and clipping in the stage audio recorded to the H4.

Again, you will want to work volume levels, pan controls, and and fill-right/fill left audio effects. For example, if you had a camera on the left side of the room. and you were recording with the on-board stereo mike, one channel may be better than the other. Maybe the left side of the camera was too close to the wall and picked up the wrong reverbs, but maybe the right channel was getting pretty good room sound. Could be vice versa, too. Maybe the right channel got all the audience paper noise, babies, talking etc. while the left channel got cleaner sound.

Obviously, there will be some phase differences. Depending on the distances from the stage, you might be able to adjust for this in PPro by sliding the camera tracks a frame one way or the other relative to the on-stage tracks. You can go into soundbooth and make finer adjustments and produce a tereo mix to import back into PPro.

Although it does not help you fix you last project, for the future, I suggest taking advantage of your multi-cam shooting set-up and use it to improvise "multi-track" recording by dividing the mike and audio feeds between your cameras and devices. This gives you more to work with. Of course, there are times when having more to work with is simply more to work with and a pain. At other times --- in the kinds of situation you described ---- it gives you the material needed to deal with the vagaries of the staging.

A final observation: This kind of event shooting can never be as as good as the kind of PBS/BBC broadcast production we have in our minds when we go to do these things. The reality of shooting momentos for the parents and grandparents is that audio which we think of as just above terrible is often far superior to what they would otherwise get with their iPhones, flip cams, etc. that it is more than good enough. When the parents (and grandparents etc.) play the DVDs, they want to be able to see and recognize their children's faces Pleasing that audience is the object of the production. We have to keep reminding ourselves of that.

[Adam Gold]

Hey Jay! Great to have you in on this discussion.

Appreciate the tips. We always use all the audio from the cams as well as from the Zoom and mix them to get the best sound, presence, audience and the like. We usually have two cams up front and off to the sides and the on-cam mics on the Z5s are actually pretty good. We often mix this audio in with that from the Zoom and the stage mics to get a really nice mix. But for this show, those two cams were in the back corners of the theatre and so thoroughly useless.

Our two center-rear cams helped a bit, especially when the performers were up in the aisles where there were no mics at all. They're good for applause but not much else, usually.

I like the idea of doubling or tripling the tracks. I'm also wondering if making one or more of the tracks a frame or two out of sync might help make the sound fuller. Will have to try that next time.

Thanks again for the advice.

[Chad Johnson]

I don't know if you're done with your project or not, but I think an app that would help you now and in every production, is iZotopr RxII. it's magical the way it can remove rumble. And if you recorded at -40db, there is probably some hiss present after bringing it up in level. RX removes hiss better than most. Any videographer who deals with sound should have this. You just open it up and throw the file on, and you can see the full spectrum of sound and easily see the sounds you don't like, select and delete. I'll drop a couple demo vids:

YouTube - iZotope RX 2: Advanced Explained
YouTube - iZotope RXII intro


Then after you've cleaned up the crud off the tracks, you bring them into your DAW. For the most part you set up an EX and a Compressor on every channel. Just try to sculpture a decent voice sound, and slap a 4:1 ratio on the compressor, and set the threshold to not trigger on the quietest dialogue, then pump the make-up gain. If you cleaned your tracks well the hiss shouldn't get any worse than it originally was when pumped, but then you can use a graphic EQ to make a little spike pointing down, then drag it around until you find the hiss. Just minimize it, don't try to totally remove it.

Then after you get a reasonably even mix through cutting clips and adjusting levels, you can either bounce out a "MIX" file, and master it in another app, or as I have been doing on projects is popping iZotope Ozone, a very sweet mastering plug, into the FX section of the Master bus. In Ozone you can further smash the entire mix (tastefully of course) with a maximizer to get a more even overall level, while applying EQ, Multiband Dynamics, and other mastering tools. But if you don't know what you're doing, they have a lot of presets. You find one that gets you close, then tweak it. Again, it's like magic. But it's best to make sure your mix is done. The Ozone is the icing on the cake. But when you put it in the master bus, rather than bouncing out a stereo file and mastering that (which is the more common approach to mastering), in the master bus you have the luxury of making small tweaks to the mix without rendering a long stereo file every time.

Maybe you don't have the scratch for all that software at the moment, but for future projects I'd suggest them both. I use Ozone with Cubase, and man what a great combo. I use RX stand alone. I can even just drop the video file on it (if my audio is in-cam) and RX makes a wav file, named the same as the video file, so they appear next to each other in the finder.

Even without the RX, I cave you some of my methods that you only need EQ and Compressor for. Then you could put a maximizer only (if you have one) in the master bus and you'll be able to boost the levels while not clipping. With proper compression and limiting(maximizing) you don't need to cut out as many clips to micro-tweak volumes.

Hope that helps.

Chadfish

PS
And if the company is not reasonable enough to even have a discussion about allowing mics, even with your assurance that you would hide the mics, and tape your LEDs, I would say "screw them" and not do them any favors. I'm sure another playhouse would appreciate you trading free video in exchange for you getting to hone your craft. Then you sell them on the idea of all the cast agreeing to buying a DVD for 20 bucks, which would pay for your services (in the future, after you got a work-flow down). With the cast and crew doing that for friends & family you can actually make it worth your while.

[Adam Gold]

Chad, this is great. Thanks for the suggestions. Sounds like a tool that should definitely be in my kit.

Appreciate the advice.

[John Isgren]

Adam,

I have dine sound for a theater group for a couple of years and this is what works for us when we record. First you need to think in terms of two completely separate sound systems - one for sound reenforcement, this is what is used so the audience can hear the actors- and then a second system that is the audio feed for the recording. If you think about it they have completely different goals. For sound reenforcment the audience hears both the actors/orch live plus what comes out of the sound system. When you record this signal you only get the sound system and it is not properly balanced.

The boundary micas work okay for the first case as they as just filling in, but sound terrible on their own. So my suggestion is to let the stage use whatever micas they need, and then set up your own for recording. We usuall put 2-4 shotguns across the stage, a couple of cardioids above the orch, and maybe another stereo pair out in the audience. We do this on a Thursday night show where attendance is lower. We also then take two channels out of Tuesday's board. One gets the omnis, choir mics, etc - any area type mics. The other carries only solos from those on wireless.

These are all fed into a separate board and mixed live while we record. We are just a local theater so budget is always tight. We have a hodge-hodge of equipment and wireless mics. My crowning glory was to get 28 wireless running at the same time with no feed back and relatively good service!