How can i get the sound right?
 

Hey Folks,

ok, this might sound like a odd question but how do i get the sound 'right'? My Setup is this: Typically i use a sennheiser SENNHEISER EW 100 ENG G2 setup with one Lavelier. It goes into my Canon XH A1 XLR input.
Then +12db gain on camera is disabled and the Inputs are set to 'Mic'.

On my transmitter i've set the sensitivity to -10db, the AF-out on the reciever is set to -24.

The leveler on the camera is set to manual with a medium setting.

Well, my problem is: i have noise within the sound. The Lavelier Mic produces a wide range of loud and quiet parts (as there's obviously no limiter in place) and when i apply tons of dynamic filters to raise the audio to a steady level for all parts i get *lots* of noise within the spoken audio.

It's hard to explain for me as i am from germany but i've attached you a edited soundfile to listen to it.

It would be great if someone could advise me on where exactly it is going the wrong way here. ;)

Thanks alot!

The link: http://www.treknews.de/noisysample.mp3

The second thing is: i am currently editing some sort of late night show and mastered the regular voice from the moderator to -1db. The problem is when i try to add music in the background in premiere cs3 - the level meter thingy reports its on overdrive which is sort of resonable for me as two audio volumes are being added up but how can i make a 'resonable' mix of those two? I have constant applause being added and find it absolutely difficult to level the voice down every time applause is being played in. How are they doing this when they film the tonight show with jay leno for example? It sounds perfect and i doubt theres a guy mixing it live ;)

Thanks for all your help!
-Daniel

If you're putting your levels up at -1dBFS you're mastering far too hot. EBU standards call for average levels / alignment tone to be set so that 0VU = -18dBFS. That puts peaks at around -10dBFS with never exceed -8dBFS.

thanks for the info, this was new to me. I am producing it for web-playback which i figured needed not that much dynamic range as well as a high average level. got complaints from users that told me that they had to regulate the volume all the time as the sound was too low when someone was talking, too loud when someone was screaming :(

Excessive dynamic range can be controlled by the judicious application of compression and limiting in the final mastering stage. But simply setting the levels high like you've done won't do it, especially when the source material swings wildly - a range of XX dB is a range of XX dB, regardless of what the average level is but if your average level is set so high that the peaks exceed 0dBFS, you've clipped. That's where a compressor comes in, to reduce the range while raising the average level.

you're talking abou the audio file i've uploaded right? that was after applying the compressors to get this -1db fixed level

is there a tutuorial online somwhere where a decent production process is being described? basically i've rendered the timeline to audition, applied the compressor and that's it.

Sorry, I confess I haven't had a chance to give your file a good listen and have just been basing my comments on the numbers you gave. It could be that you're compressing too early. It should be the very last thing you do after the final mix

Assuming you mix your vocal to a more conservative level, as Steve suggested, here are some ways to deal with the audience masking your voice:

1. automate the levels of your audience and voice to keep a more consistent level. the level changes are much more transparent that way rather than reaching for a compressor right off.

2. eq the audience by dropping a few db in the 2-5k range range in order for the voice to pop out a bit more.

3. consider using a mild sidechain compressor to open and close whenever the voice is talking. be careful with this one because it could easily get out of hand.

As far as recording...if the dynamic range is really changing a ton then try adding a little more compression.

is that sound clip the raw or processed audio ? sounds like a gate opening and closing, or some compression problems.

thats after processing ;) i've used a very hard limiting thingy to get rid of the constant noise between the talking. "kill all below -45db", afterwards i've applied a limiter with raise +9db and limit at -1db which put everthing to a constant level

@Jonathan:

thanks for the tips! can you recommend a good camera mountable / portale recording system for that?

thanks!

Mountable? No I can't. These are all Post Production tips for when you get done recording. I guess you could do these things on the fly but that's out of the range of my expertise. As far as on the way in(I thinks it has been mentioned) I would use a a mixer to separate sources and a good limiter to catch the peaks.

ah i see. i think simple limiting would solve LOTS! of problems for me as my Canon AH 1 seems to have bit of a problem there.

thanks for your advice