Botched audio for an interview.

[70Deep]

While doing everything for this interview from lighting to sounds to prep and moving things onto location I managed to set my zoom h5 (shotgun module) recorder wrong and botched the audio by setting the gains way too low. Like whispering on max laptop volume low.

Everything else went perfectly with both cameras and i managed to get a shy women who had never been interviewed to open up swimmingly only to realize it night all be for nothing

I have pro tools 10 and figured I might be able to compress the audio louder but before I waste time trying anything I figured I'd take a shot seeing if anyone had made this mistake before and would know a good solution.

 

[DPStewart]

While doing everything for this interview from lighting to sounds to prep and moving things onto location I managed to set my zoom h5 (shotgun module) recorder wrong and botched the audio by setting the gains way too low. Like whispering on max laptop volume low.

Everything else went perfectly with both cameras and i managed to get a shy women who had never been interviewed to open up swimmingly only to realize it night all be for nothing

I have pro tools 10 and figured I might be able to compress the audio louder but before I waste time trying anything I figured I'd take a shot seeing if anyone had made this mistake before and would know a good solution.

Compressing ALWAYS makes the noise floor louder. ALWAYS.
Don't do that. Just raise the fader all the way up. And maybe normalize the gain for an additional 5-10db if putting the fader all the way up is still not enough.
Then your issue will simply be noise reduction, and there are many widely documented ways to do that.

 

[Alex H.]

My first question, and this ismto be filed under "Lessons Learned", is... "Where were your headphones?" If you aren't listening, you aren't recording. You may as well turn off the camera monitor, too.

Just raise the fader all the way up. And maybe normalize the gain for an additional 5-10db if putting the fader all the way up is still not enough.
Then your issue will simply be noise reduction, and there are many widely documented ways to do that.

There's a catch to this: your record levels are so low that you've only used a very few of the available bits in your audio's bit rate, so there's only so much you can do without your audio falling apart and starting to sound like a bad Internet download ca. 1998. You can probably add a gain plug-in to the track in ProTools, and possibly also use the gain tools in EQ and other plug-ins, in order to bring the levels up. But start making large changes with EQ or messing with the bits through something like NR and your sound is going to suffer. You may just have to live with noise floor in order to get usable mix levels.

 

[70Deep]

Compressing ALWAYS makes the noise floor louder. ALWAYS.
Don't do that. Just raise the fader all the way up. And maybe normalize the gain for an additional 5-10db if putting the fader all the way up is still not enough.
Then your issue will simply be noise reduction, and there are many widely documented ways to do that.

Fair enough fair enough. Sounds plausible then. And alex, my headphone were right where they should be (on my head), but with the way zooms are set up you could always hear much more then one would believe. For instance, I could hear the creaking of the wooden floor we were on and the slight thud when the woman's hand would come to rest on the leather couch when she would put it down too quickly.

Luckily the project isn't due in for a short while so I have time to breathe while I finish working throughout the weekend.

Thanks again for the posts guys, it's my first thead and probably one in the wrong section at that haha. Happy Easter to all those it applies to.
Checking the levels on screen would have immediately told me my settings were off and that is what I clearly missed

 

[Greg_E]

Do the zooms set record gain after the monitor circuit? Or were the headphones turned all the way up? Otherwise, if you can hear it, you can probably save it to a certain level.

Suggested workflow:

Connect audio recorder output to camera and set level to correct, don't touch it again. Then if the meters on the camera are low, you know your audio is low in the recorder. The Tascam DR-60d has a camera output with level adjustment and tone generator to use to set the level. Also has a separate headphone out with level control so you can monitor quality and content (not level).

 

[Bruce Watson]

My first question, and this ismto be filed under "Lessons Learned", is... "Where were your headphones?" If you aren't listening, you aren't recording.

There is a lesson to be learned here, and the one you state is a good one. However, I'm thinking it's not neccessarily the correct one. I think the correct lesson to be learned is that audio is a full time job. Just like video. And when you try to do two full time jobs at the same time, stuff falls between the cracks. And most often, it's the audio that gets screwed. Just like what happened to the OP.

Just having the headphones on is not enough. You have to actually be listening to them. And you can't really pay attention to them if you're working a camera. I know, I've tried. It's exceedingly difficult. The clinical term for it is attention blindness. Humans don't multi-task well at all, even though each and every human thinks they personally are the exception that proves the rule. And none of them are. Not even me. Sigh...

So if the OP is to learn a lesson from this, that's the one I hope they learn.

 

[Michael Carter]

Bruce makes a good point, but in the real world, there are many gigs where hiring an audio guy would mean it's not worth doing financially. If I'm getting a grand to shoot and edit for a small business and I need to pay $350 for an audio guy, there's not much point unless I haven't worked in a week or something.

Doing audio and video at the same time is doable, you need a solid workflow and it needs to be 2nd nature. Often, for my interviews I'm not just the DP and the sound guy - I'm also the eye-line, which means I have a little monitor in my lap and a focus whip sticking out of the rig. If the subject is really animated, I'll try to sit him at a desk or a table and pin him in; usually they stay fairly static.

I usually do a basic setup where I sit in the chair and check framing with the monitor turned around, tweak lights, roll and replay - it's a pain but i can do it very quickly now. Then I do a final setup with the subject and do a final level check. I put on the cans, have them talk as loudly as they expect to, make sure I have a -6DB backup track rolling… and then I take the headphones off and sit down and roll camera and recorder.

So I'm not only not monitoring audio, I'm barely monitoring video - because losing eye contact jacks up the interview, and it feels rude and inattentive. I generally take a peek between statements. I keep my attention 100% focused on the subject, without headphones on, which distance me emotionally from the person.

I always get the best audio possible for the given room; are there times my focus isn't as sharp as i wished? Yeah, smaller rooms leaving me stuck with F4 or so to soften the BG when 5.6 or 8 would be safer. But my clients consistently range from "pleased" to "blown away" by what I can pull from a subject. I warn the client that their budget doesn't allow for an optimal shoot of course, but maybe I've been lucky - maybe I've been prepared - but I love love love interviewing, I love finding the passion behind a business or personality, and I usually get an "oh wow, I didn't think he had it in him" when they see the final cut. It can be done, even if it's hard. Be prepared and get as 2nd nature with your sound gear as you are with your camera. If you want a great interview, you simply can't be worrying about your gear.

 

[Alex H.]

And alex, my headphone were right where they should be (on my head), but with the way zooms are set up you could always hear much more then one would believe. For instance, I could hear the creaking of the wooden floor we were on and the slight thud when the woman's hand would come to rest on the leather couch when she would put it down too quickly.

...

Checking the levels on screen would have immediately told me my settings were off and that is what I clearly missed

This is less about the design of a Zoom and more about user setup. Always, always, always set your record levels (i.e. sound check) before you record. This means having the mic in the right place, and having the interview subject answer a few sample questions (even simple chit-chat) at their nominal speaking level for the interview. Make sure the meters are hitting the right place, but make sure also that the sound is clean (free of distortion and shorted cables and wind, voice is on-axis to the mic).

Set your headphone levels from there... this isn't often cited as a step in proper gain staging, but I believe there's an argument to be made there. Your ears are more important than your meters, and having the headphone level calibrated to a proper dialog recording level means you'll immediately know when something is too low.

Thanks again for the posts guys, it's my first thead and probably one in the wrong section at that haha.

Luckily, we have the technology to fix that (moved to the Sound forum).

 

[Noiz2]

As to "fixing", definitely don't compress. What I would try first is to gain the level to where it needs to be and then try some noise reduction and see what comes out.

If you are lucky the signal is strong enough that the NR will get you usable sound. If not you are kind of SOL.

A lesson that is just as true in post as it is in the field, is that judging things by ear only works IF your signal chain is calibrated. It is a really common mistake with inexperienced editors to change the speaker levels and not keep them consistent. Then come back and wonder why some sections are a lot louder than others.
So don't feel too bad having headphones cranked and not realizing you were recording way too low is something that has happened to almost everyone. Once. Learn and don't forget.

 

[Rick R]

If recorded in 24 bit, bringing up a low signal doesn't introduce as much noise and grit as 16. NR can reduce the hiss some. But its still gonna be lower res than normal. Hopefully the OP at least recorded in the PCM format.