Raising Volume on Voice Also Raises Room Noise

I think BG is BackGround...
Click to expand...
Yes BG is Background, AMB is Ambience.

ADR has different meanings "Automatic dialogue replacement", "Advanved Dialogue Replacement", "Alternative Dialogue Replacement"... chose one :-p
Click to expand...

I don't think there is a definitive one. ADR is also called "looping".


Idealy the actors see the action on a big screen with written text under it. The text is written in "pronouncing" style.
For exameple: "Hey dude, where 's my cow ?" can be written (and pronounced) as "heeeeeeeeeeeeeeeyy duuuuuuuuuuuude wheereeeeeeeeeeee is myyyyyyyyy cooow"
Click to expand...

That would actually be the European style of doing a foreign language dub. The sub-title part is not done when doing ADR. At least in the US.
 
LuckyProducer said:
Question: in wider shots, how do you get the mics close enough without having them show up.
Click to expand...

On a wide shot you use the audio from the closeup shot. If your actors are decent and you warn them that this is what you are doing
then they can nail a repeat performance easily. I tend to keep dialog on wides to a minimum usually at the end or beginning of a scene.
 
Spoken voice only goes down to say 110 hz usually, unless you are a heavy smoker. So EQ is a powerfull tool here. On female voice you can almost double your rolloff point as well--depends how much or that room sound is rumble. Som a/c units have a lot of mid and low treble energy and are a real pain to manage.
I would also use splits and fades, which to me sounds better than volume curves.
I split the audio where I want to make a volume change, and make a nice crossfade into the spoken material. That way, you can trim the noise floor and still get a good clean into in to the spoken word. I don't particularly care for Vegas as a sound editor, but it is more important that you be comfortable with whatever NLE you are using. If you are using DX or VST plugins, then it doesn't matter much what the underlying NLE is--the big difference is of course whether you can do four point crossfades, and that costs extra money for that feature.
Lastly, you can try context differential NR, where you use a different noiseprint taken directly before the person speaks. This allows you to compensate for variations in the noise.
 
Noiz2 said:
Yes BG is Background, AMB is Ambience.



I don't think there is a definitive one. ADR is also called "looping".




That would actually be the European style of doing a foreign language dub. The sub-title part is not done when doing ADR. At least in the US.
Click to expand...

You're right it's for the dubbed stuff. But it could be usefull to for ADR
 
DrDave said:
Lastly, you can try context differential NR, where you use a different noiseprint taken directly before the person speaks. This allows you to compensate for variations in the noise.
Click to expand...

What do you mean by that?

And I'm not sure what you meant by EQs. What should I do? Bring the trebles higher? Lower? What specifically? If you don't want to get deep into it, it's ok.
 
LuckyProducer said:
What do you mean by that?

And I'm not sure what you meant by EQs. What should I do? Bring the trebles higher? Lower? What specifically? If you don't want to get deep into it, it's ok.
Click to expand...

It depends on the nature of the room noise you wish to even out. Let's say as a general rule, some to most, but not all, of that noise is concentrated between 30-120 Hz. You can just use a high pass filter--start at 120Hz--to remove or lower that, but you can also use EQ to *add* some to even it out in quieter spots (only if you really need to). Move the roll-off point up and down till you hear it affecting the voice, then back off to find your ideal roll-off point. Then use the other bands to see if there are notches in the sound. Specifically look at harmonics, or use waves x-hum forthat job. Again, it depends on the room noise. If you are using noiseprint software, take more than one print to fine tune the NR, unless the noise is absolutely constant. I use spectral noise reduction, but the noiseprint can do OK for dialogue.
If the noise is concentrated in the mids or highs, you can still use EQ, but it is trickier, and spectral analysis is the way to go. That software, like Cedar or Renovator, can be pricey, but you can hire someone to do it for you.
 
You must log in or register to reply here.
Back
Top