Audio editing order of operations (i.e. compressor, levels, normalize, etc)

Discussion in 'Digital Audio' started by budha, May 25, 2010.

  1. budha macrumors member

    Mar 5, 2008
    Hello everyone.

    Is there a specific step by step process you should do when post processing narration/voice over audio as good as possible? I am recording podcasts for work now and I want them to sound great.

    I understand you need a good mic, etc, but I am wondering what you do once it's recorded (I already have a good USB mic). I'm using STP3.

    Also, is there a good way to make my voice sound richer/deeper, with an effect?

    So for example (this probably isn't right, just throwing a list out there):

    1. remove bg noise
    2. adjust audio levels
    3. run a limiter effect

    Thoughts? Thanks!
  2. ChrisA macrumors G4

    Jan 5, 2006
    Redondo Beach, California
    Some might argue that "good USB Mic" is an oxymoron. But I guess you need to use what you have.

    You can't remove background noise. You can try and EQ it out ot bring the volume down during pausies or mask it with music. But once you got noise, you got noise.

    One of the big problems with a USb mic is that the USB cable is short. With a real mic able you can get 50 feet away from your noisy computer and its fan and hard disk. You need a noise free space tomake the recording.

    Other kinds of noise you make your self for example the voise caries past the mic, hits the wall and bounces backinto the mic. Get rid of that.

    In short do absolutly as much as you can to the sound BEFORE it reaches the microphone.

    Ideally any limiter would be placed right after the microphone before the signal is converted to digital. But if you record in 24 bits you can run the gain down a little and the 24 bits will get y so much headroom you don't need to compress or limit. You can do that later if need be in software.
    But if yo are recording in 16 bits with a fairly dynamic source you need to do the limiting in analog. Yes you can turn the gain down but with 16-bits you have fewer to throw away like that.

    Once yuo have the digital file, the order that the plug-ins are applies matters less than you's thing as all the work interally is done in 32 bit floating point.

    How to make the voice sound better. Mostly this is a functin of the mic. People will say to use a large diaphram condensor mic. But just using EQ can boost the lows and have a sorta of the same effect.

    To keep the noise out use a pop filter or good wind screen on the mic and keep the mic very close, inches away, lean back to breathe. move far from walls or other reflective surfaces, a long cable helps.

    For good resilt the most important things are what are closest to the sourec. The number one thing is the person speaking, nothing will fix a poor voice actor. Next is the space between the mic and speaker. Is is the space silaent and non-refective and is the mic close enough and is there a filter? next on the list is the mic, is is any good? Ok this is 90% of the battle. Next comes the mic preamp and the D/A converter. Software is last. Work and money spent at the top of this list will pay off more then work spent on bottom of this list. Wouldn't it be nice it this were not true then we could use cheap gear andno-talent actors and fix it all with software. Maybe some day? But today recording has many stages sound goes through and each one is "garbage-in garbage-out" so you have to be very careful with those first stages.
  3. TheLOGICalone macrumors regular

    Jun 7, 2010
    Your best bet

    Depending on budget, you could buy a shure sm7 for 350$, and a stand alone preamp with a bunch of tools (ART makes a nice one for like 200$) If you want to work with what you have, first normalize your vocal track. then look at plugins, first thing on vocal track is a noise gate, it will kill the level when you are not talking, set the threshold as low as possible while still getting un interupted vox, then try a denoiser (if you have) to filter the bad noise, then eq, i do low cut from about 60hz and down, dip 100hz, dip around 300-500hz, dip just under 1k, and probably kill everything over 10k-15k (if you have noise), then compress, for speech i would say to look for about 10-15db of gain reduction(the light or graphic that moves left to right or up and down with db meter) use btw 5:1 and 10:1 ratio, attack all the way to left and slowly move it right until you get too much definition to your speech, release in the middle, and set gain to 0. Then you can try a chorus effect, make all of the settings very low and slow to just thicken it up a bit, flange also works. You can also try gentle reverb for this low mix small size (bus them if you can), then throw on a limiter with max 3db gain reduction.
    Oh, and another trick is sidechain compression: send your vocals to a bus, say bus 6, then turn down the fader for that bus to 0, then put a compressor on your non vocal track, and make it rather strong (low thrshhold, high ratio, 0 gain, med attack, fast release) and set the side chain input to bus 6. This way, when the compressor detects you speaking it will lower the non vocal track (this is why music dips for radio djs talking)

Share This Page