Making good, consistent Voiceover recordings

Special Interest Groups Recording Techniques
1

Hey there, all! I hope this is the right forum for this, but here goes:

I’ve been recording home VO for videogame mods for a little while now, and have been basically successful, but I have several questions about how to make my life easier. I know VERY little about recording, and have basically just been doing ‘what works’ for some time, so I know nothing about “hard-pass limiting” or “dynamic ranges” or what have you. I can successfully Normalize a track, do Noise Removal, and Fade In and Out, add Silences, and even change the Speed slightly, but that’s about all I know how to do. Are there other tricks I can use to work with my voice files to get them to sound good?

  1. Is there a good way to get a consistent volume level across multiple tracks?

I ask this because some lines of dialogue have more energy in them (“Attack!!” vs. “Hello, stranger”), so Normalizing them at the same time will make the more energetic line (or the line that has a few more loud moments in it) much softer. Is there a way to preserve the integrity of each track but make them of a similar volume level in general, but still let the loud parts BE a little louder?

  1. Can the Draw tool help me clean up otherwise wonderful tracks?

I’ve never used the Draw Tool, so I don’t know it’s best uses, or even really what it does. Can it help with click and pop smoothing? Can it help remove or clean up high volume levels inside of tracks (like when I got too close to the mic, or my voice cracked, or what have you)? Do YOU use the Draw tool to fix recordings?

  1. Is there any effective software way to De-S a track?

Occasionally I get strong S’s in my recordings, and a friend at a professional studio I know has a piece of equipment called a De-S-er that helps him smooth out or tone down big S sounds. Is there anything like that in Audacity, and if not, where can I find it?

2

Hey,

The peope here are great with tech answers. I’ll be watching to see what is recommended. For Voice techniques check out voice-overs.com and go to the forum. You’ll find some good stuff there.

Lynn

3

Yes you could use the draw tool - but you would be better off upgrading to the Beta version 1.3.5 (which is actuallu pretty stable now). In 1.3.x the developrs introduced an excellent new Repair effect, which enables yo to select up to 128 samples of damaged sound and then Audacity will interpolate a repaired waveform based on the surrounding waveforms. Note that to use either the Draw or Repair tools effectively you will need to zoom in until you can see the little individual soound samples.

IIRC I think that Stevethefiddle (a frequent poster on the forum) is developing a de-ess plug-in. BUT you will be much better off avoiding such sibilants in your recording in the first place. Wjat tou need is what is known aas a “pop shield” between your mouth and the mic. You can make one at home with a wire coathanger and a pair of ladies tights/pantyhose. And avoid getting you mouth too close to the mic.

WC

4

When you have words like “Attack!!”, there will be some high amplitude peaks in the sound.
When you have words like"Hello, stranger", the level will be much more even.

When you use “Normalize”, it will bring the “peak” levels to the same level, making the overall volume of the “softer” words louder than the harsh words, which if anything is the opposite of what you want.

To get round this problem, you need to use a “dynamic compressor” or a “dynamic limiter” that will “squash” the peaks in the word “Attack!!” to allow you to amplify it further without distorting. You can find a lot more information about these effects if you look in Wikipedia.

An effect plug-in that I suggest you try is the “Fast Lookahead Limiter” available from here: http://plugin.org.uk/ladspa-swh/docs/ladspa-swh.html

To use it, set the “Input gain” to a positive value - the higher you set it, the “louder” it will sound, but too much will sound distorted. Set the “Limit” to just below 0dB (say -0.2) and the release time below 0.5
A bit of experimentation will be necessary.

The draw tool works on single samples and is very time consuming to use. For repairing single clicks I’d agree with WC and the “Repair” tool in 1.3.x

Yes indeed, and it’s virtually finished apart from a bit of tweaking and documentation.
I’m looking for some good samples of sss’s to try it on for the tweaking, and also for some people to try it out and give feedback on it. If you’re interested in trying it DoomMunky, PM me your e-mail address and I’ll let you know when it is available (which will be before Christmas).

Also. if you could provide some short audio clips which you think need de-essing, I’d find them very useful for fine tuning this effect and setting useful default settings.

I would agree with that, but don’t get rid of them altogether just yet, I still need some ss’s for my plug-in :smiley: