What levels (dialogue and music) to use in post of a documentary

Fernando Lino

Fernando Lino

Active member
Hi guys,

I'm editing the audio of something I shot for the first time in Final Cut Pro 7 (and Soundtrack Pro to remove the noise). I recorded the sound in a Zoon h4n with the dialogues ranging from -20dbs to -12dbs and I did the same with the music. I was wondering if you could tell me what's the standard levels for dialogues and music in documentaries?

Here's a sample of the video:

http://vimeo.com/31598131

Password: audio

What do you think of the levels like this? Should I use some filter for the dialogues? There's something a bit odd in them and I don't know for sure what.

Really appreciated for your help
 
Last edited:
I am not an expert, but Where was the mic in this shot?

In this case, you have a poor original voice capture from your talent, this makes your job SOO much more difficult. You are getting the -12dbs on your H4n, but it was probably a lot of reflection noise adding to that -12 number, so when you clean it up or add any gain in post, you get a louder voice, but you also get louder reflections (echo) because your mic was far from your talents voice. Or you might have been using the wrong mic, such as an omni or cardioid when you should have used a super cardioid or hyper cardioid or shotgun (worst case with shotgun).

What is the piece for? Student work? Promotional? Paid work (I hope not)?

As for music levels, I don't think they are that bad (because you have a clear signal with no reflections in your music) at -20.
Ian
 
Hi Ian, thanks a lot for your post!
It was like 1 meter away from the subject on the left (from the point of view of the viewer).
It's a paid job but very poorly paid. I had a lot of limitations with the gear because of that, I only used the material that I've and some borrowed from friends - the only thing that I could afford to rent was a slider for the camera.

To record the sound I only had the Zoom h4n and that's with what I recorded the dialogue and the music also... I know it's not the best option but I had to make it work with that. We really didn't had the money to borrow a mic or friends that could lend me one...

What should I do in this case?
 
I don't know what you 'should' do, but in all honesty, next time you should prepare better. If you must shoot an indoor scene without the proper mics (and hey, we can't all have the proper equipment from day 1) then you need the mic as CLOSE as possible to your talent. If I only had the Zoom H4n, with it's internal mics, I would have made sure that the zoom was taped or secured to a wooden pole or mic stand, and hung above the frame of the interviewee. This would have put it about 12" to her mouth, FAR better than the placement you had.

Also, I would have changed the scene, knowing about the reflections to show the interviewee outside. (because outside, there are fewer things for sound to bounce off of, and create echo).

What you should do right now, is think about your story, and what you are trying to accomplish with this paid work. What is it she wants to highlight (assuming she is the client) and focus everything on that, if she was communicating some important piece of information about an event, I might just voice-over the entire thing. If that is possible, I would probably cut the audio while she is talking, and boost it (echo and all) just to make sure her information does come accross clear.

If you CAN go back and re-shoot (and why not, just tell the client you were editing and found more inspiration for something 'MORE') and have her do the interview outside. Maybe outside on a picnic table with the H4n hidden inside or behind a small flower arrangement or something, and make sure the internal mics are turned to 90 degree's and the body of the H4n is pointed at her mouth.

Those are the things I would try to do. Remember that paid work should be as best as possible. No one cares that you needed to go back 3 or 4 times if the final product is awesome, the only care when they are disappointed with the final result/product.
 
I showed the entire clip - the entire video is 15 min. long and it's kind of a promotional documentary of a 5 star hotel in a small village in Portugal (although it's a 5 star hotel they really didn't had the budget to spend on the video) - to them and they truly loved it. But I know that they only loved it because there's a emotional connection to what was shot and because they knew of my limitations...

It was my first time doing sound all by myself but you're right... I should have prepared better...
So now I really don't have time to go back and re-shoot everything, the edit is done and it would take me a lot of time to finish the video.

So I'll have to work with what I have in post...

If that is possible, I would probably cut the audio while she is talking, and boost it (echo and all) just to make sure her information does come accross clear
Click to expand...

So far the only filter that I used was gain to boost the sound (and soundtrack pro to remove the noise). Do you know of any filters in Final Cut that you think would diminish the echo?

Many thanks for your help on this
 
The balance seems OK. You do have a lot of room bounce. There is not a lot you can do about that. An expander gate can sometimes help a bit if set really low.

Even if you shoot again with only the Zoom you need to get the mic closer. Also with the zoom pick one mic to use and make sure that is the one pointed at the talent and is closest. In post throw out the other track and only use the one you aimed. That will help with the bounce.
 
Hi Noiz2, thanks a lot!

I used the expander gate throughout all the video and it made it work a lot better.

There are the settings that I used -
Expander/Noise Gate:
Threshold(db): 0
Ratio: 1,3
Attack Time: 1
Release Time: 100

Here's a sample

password: audio

http://vimeo.com/31614064

What do you think?

Sometimes I hear small *bips* or small flutuations in the sound. Do you think that changing the attack or release time would improve that?

Thanks a lot guys
 
Fernando Lino said:
What do you think?

Sometimes I hear small *bips* or small flutuations in the sound. Do you think that changing the attack or release time would improve that?
Click to expand...

If you hear it, others will too. So just trust your own ears when trying various adjustments.

Before you're done, be sure to try to play on (regular) speakers instead of only through headphones. I listened to your clips and I don't think there is an obvious improvement. It still sounds like the woman was speaking in a big empty room.

BTW, because I don't understand what she was saying, my feeling was not swayed by her words (their meanings) at all. I felt the background music was more annoying and intrusive than helpful. Obviously, it's not as echoing as the voice, so it feels unrelated to the voice or the big empty room - as if you were playing an FM radio or watching a music TV while recording a woman talking. It's more like a continuous loud ambient sound (a.k.a., noise). Well, this is just my feeling. Others may feel differently.
 
Fernando Lino said:
Hi Noiz2, thanks a lot!

I used the expander gate throughout all the video and it made it work a lot better.

There are the settings that I used -
Expander/Noise Gate:
Threshold(db): 0
Ratio: 1,3
Attack Time: 1
Release Time: 100

Here's a sample

password: audio

http://vimeo.com/31614064

What do you think?

Sometimes I hear small *bips* or small flutuations in the sound. Do you think that changing the attack or release time would improve that?

Thanks a lot guys
Click to expand...

Yea this is a lot better. Sometimes I find it better to be light on the expander and just manually write in volume automation and really smoothen out everything using automation points. You're not the only one that hears small fluctuations; I ALWAYS hear it at the movie theaters, but maybe that's because post sound has ruined my ability to enjoy a movie ;)

I've had a lot of success using multi-band expansion with a combination of EQ. At least with multi-band expanders, you can at least attenuate a certain frequency range of noise instead of the entire spectrum.
 
Thanks a lot! I'm going to try the multi band expansion very smoothly, at this point I don't have enough time to be completely thorough...

Amazing resumé by the way...
Thank you guys!
 
You must log in or register to reply here.
Back
Top