Okay, so some of the things I've worked on recently have had terrible audio. After hours of setting up lights, shots, etc. it really sucks to have audio tear down your production and make it amateurish crap.

I'm currently working with my Sennheiser ME66, the Canon MA100, and the Canon XL1s. These are my audio acquisition tools at the moment. What I need to do is ensure I get the best quality stuff I can out of this setup that I can.

During acquisition, looking at the audio meter on the XL1s, what do I not want to go past? -12? -6? 0? I thought -12 was the digital equivalent to 100% on an analogue VU meter. Is this right?

On the ME66, what are recommendations for settings as far as the bass rolloff switch goes, for dialogue and other sound (foley work, for example)? Seems having the bass rolloff turned on can make voices a little tinny, while off the seems to pick up all sorts of extra crap.

On the XL1s, what are the recommendations for settings as far as using the MIC or MIC ATT setting for the MA100 inputs, for dialogue recording and other stuff (like foley, again). I was told MIC ATT might be more appropriate for dialogue, as the camera adds gain and some more technical stuff I don't remember.

When mastering for final output (yes, for broadcast), what do I not want the audio to go past (this is judging from Vegas 4's audio meter. It IS accurate, isn't it?) This guy told me something I'd done recently was way too low--I know most of it was past -12 on the meter. His way of telling whether stuff is the right level is to plug his camera into the TV, put a tape with the output on it, and then play the tape, see how it sounds, then change the TV to regular channels and see how loud/soft they are by comparison. Good method? He says it works for him.

Think that's all of them.

I don't know much of audio so I'm not going to have answers to
you. But you say it sounds terrible. What is wrong with it? What
exactly does sound terrible? Did you use headphones to monitor
the level and quality of sound?

I myself did not monitor it, what with having to spend my energies keeping my camera level and the subject in frame. The guy monitoring it didn't complain, so we didn't find out how crappy it was till afterward. The scene we shot had regular speaking levels, and then people screaming at each other. Not wanting to try to ride the levels, I set audio so the screaming wouldn't clip, and that of course made the regular speaking very low. furthermore, because of something I previously didn't know about the mic, the audio levels distorted anyway, even though on the meter they weren't above 0.

Hmmm.. any audio specialist can answer this?

Josh, -12db is = to 0 in analog, so if your audio peaked at -12db you should be OK on that end BUT if your normal speaking audio is less than -20 you might have a little challenge bringing it up to an acceptable level without brining in other ambient noises.
If the screaming audio isn't clipped then go to the normal tone stuff in your audio editor and try not only raising the gain but also play with the EQ a little to keep the noise level down. Write down every move you make so if you have to go back to 1 you'll know what it was.

MicAtt will lower the levels by about 20db, so for general speaking stuff I usually don't use that. For music yes and I keep my gain at no more than 50% but for speech I keep it on MIC and keep my levels at 50% or less, depending on the speaker. Please also keep in mind that I use PD150's and not the XL1(s) but audio is audio and I believe that all digital cams run about the same.

Also remember that the lower the number the louder the level so if you're peaking at -12db at the screaming part you should be fine. As for the normal speaking if you go past -20 (ie -25) it COULD be too low.

Remember ZERO in digital is way bad.

Like I said, I'm not an audio guy but learned the hard way so I hope something here will help.
Don

Thanks. Still looking for more answers--someone on this forum mentioned that 'cause of all the in camera gain add, and some other factors, that MIC ATT might be appropriate for dialogue, despite what you said.

So levels for speech should be between -20 and -12? I'll keep it in mind.