Voice Over Recording Tip: Don't Just Normalize

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
if you're recording your own voiceover auditions or jobs you might be normalizing the audio file are you crazy that's the last thing you want to do hello I'm William Williams from Alisa Craig voiceover classes actually I'm not saying don't normalize what I really mean literally is that normalizing is the last step you want to do there are two audio effects you want to be familiar with limiting and normalizing if you want your recordings to really pop you need to understand these effects and when to use them let's say you record a voiceover audition of course you use my famous Goldilocks theory of recording not too hot so it doesn't clip and distort and not too cold so it's not down there on the noise floor okay good but then you need to pump up the volume a bit to get it to its optimum level the effect to do that is called normalizing before we get into that let's take a quick look at digital audio you see everything in life is measured from the ground up how tall are you from the ground up how tall is that building from the ground up but digital audio has limited numbers it can use to represent height 16-bit audio can either go to about 32,000 numbers high or 32,000 numbers law so digital audio has an absolute ceiling of 32,768 steps up or down so to make sure you don't bust through that ceiling digital audio is measured from the ceiling down the ceiling is called 0 dbfs the FS stands for full scale full scale means that's all there is and any No More kind of like that guy in the Tamiflu mercial so what exactly does normalizing do all it does is turn up all your audio evenly so that the loudest part of it just touches the ceiling if your audio has very even levels that's fine but it usually doesn't there may be some random Peaks that are much louder than the average and that's the problem with normalizing let's look at a typical waveform see how normal it looks yeah except for that one peak so the normalizing algorithm turns everything up until the tip of that peak just touches the ceiling I've normalized this waveform to minus 3 dB that's 3 decibels lower than the 0 DB ceiling and whoops the volume of the waveform is actually lower than the original so if you only normalize your finished audition it may have little effect or actually turn down the volume of your recording so what's the problem and how do you fix it well to find out let's take a stroll in the forest oh I happen to have one right here let's pretend all these trees want to grow as tall as possible but they all have to grow evenly and they have to stop at the top edge of the sky like they say the sky's the limit in this case literally there'd be a lot of room to grow if all the trees were about the same height but there's that one pesky tree so up they go and the growth of all the other trees is restricted by the tall tree hitting the ceiling okay now say you're in the lumber business one tall tree means less lumber from all the other trees or if you're in the audio business which we are one tall spike means less loudness from your other audio so how do you fix this well like this ah now all the trees can grow to the top of the sky like this and there's much more lumber right and in audio that means much more sound so where can you get an audio chainsaw that effect is called a limiter the effect looks like this the line that gradually slopes up is called the linear part of the effect any sound along this part isn't changed at all but the line that's horizontal is called the nonlinear part any sound on that part of the line is limited and that means it's restricted from getting any louder and where the line bends is called the threshold okay enough about trees let's get back to that waveform with the pesky peak and run the audio through a limiter we want to set the threshold so it just touches the tops of the average sounds of the waveform and voila notice the entire waveform is much more even the tall peak is lowered to the average level and the entire waveform is a little quieter but now we can apply the normalizing and will normalize up to minus 3 dB and look how much stronger the average volume is here's the original waveform and the limited and normalized waveform so when do you want to use these effects I do it on all my auditions this is an unaffected audition I'm speaking now and this is a boost at audition I'm speaking now all things being equal which do you think is going to win the job and I do it for jobs when I feel a producer might not be smart enough about all things audio of course if they tell me no processing I won't do it and if you're into specifications ACX the audiobook production site likes to have their audio processed to minus 3 DB FS for peaks and minus 22 to minus 18 dbfs rms which is the average loudness so those are the Goldilocks numbers for spoken word notice those specifications are what I'm achieving with the limiter don't overdo the limiting or you'll get pumping breathing effects a louder noise floor and a generally smashed unpleasant sound like this 6 DB of reduction is usually just fine so the moral of the story is this limit the audio first to tame those random peaks and then normalize to minus 3db your auditions will sound stronger and your listeners will thank you for it for further information on all things voice-over go to Aliso Creek net and click here to subscribe to my youtube channel I'm William Williams thanks for watching and remember keep talking you
Info
Channel: Aliso Creek Voice Over Classes
Views: 204,164
Rating: undefined out of 5
Keywords: voiceover, limiter, normalize, audio effects, audio processing, audition, commercial, narration, audiobook, animation, aliso creek, recording, audio engineering
Id: mZ8p_Mtug4w
Channel Id: undefined
Length: 8min 17sec (497 seconds)
Published: Thu Feb 25 2016
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.