Is there more?
Already, I am wondering if there is more. I enjoy doing my LibriVox projects, but still. Volunteering is a good thing: for me, I get some feedback on my performances; for the world, they get the benefit of listening to my performances. Win-win, right? But if you could get paid for volunteering – win-win-win!!!
There are a lot of things that can be done to an audio file after it has been recorded. Some of them actually need to be done sometimes – just because you can do something doesn’t mean you should. I lean more toward the Rob Mayzes school of thought that is summarized bit “Less is More.”
Throughout these posts, one important thing to remember is that your goal should be to prevent having to do any post-production of your audio files, or to minimize that as much as possible, by addressing the issues while you record. Of course there are great plugins for Audacity or whatever DAW you use, but assume that any processing will reduce the quality of the recording, and if you can avoid it, you keep your recording sounding cleaner.
This is the point of a previous post about Get Good Glass. Put your money (or effort) into your lens (or mic and recording process) and you won’t have to compensate in the darkroom or with Photoshop (or in post-production with plug-ins), If your recording is clean, you won’t need to run Noise Reduction on it. With good technique, you can avoid most mouth noise and not have to do Click Removal. Position the mic correctly and you won’t have to try to get rid of the breathy pops and booms from plosives. Work hard for the best original recording you can get, and you will be glad you did on the back end.
Realistically, though, you can’t do it to industry standards without some processing. And the first one (and maybe the easiest) to learn is normalization. Put succinctly, to normalize an audio file is to set the overall volume so the loudest part of the recording is at an acceptable level. It is your overall volume control built into the file itself.
LibriVox has a technical specification that files have volume peaks no louder than -3 dB (negative three decibels). What does that mean to the average voice actor? Probably not much, which is fine because as the talent you don’t need to know all the engineering! What you do need to know is how to make your files conform to a spec of -3 dB peaks.
In Audacity, it is very straightforward. After you edit your file (hmm, I just realized maybe that topic should have been the one before this post…note to self: write that down for future reference) ensure that you don’t have a small section of your file selected, or select the entire waveform, and go the the plugins menu list and choose Normalize. Set the value to -3 and hit OK, and that’s that. Easy peasie.
You should make it a regular practice to critically listen to your audio before and after you do any processing to it. Not only that, you should look at your audio – use the waveform display in your DAW to find out what your audio recording looks like. Hear with your eyes – kind of like breathing through your eyelids. Most likely, the volume of your recording and the amplitude of the waveform will be noticeably greater after you normalize. This is a good thing.
What you don’t want (okay, one of many things) is a recording that people will have to turn the volume up to hear correctly, and then turn back down before the next thing they listen to comes in too loudly. It’s like listening to classical music as you drive, constantly having to adjust the volume (although that is done on purpose, and normalization doesn’t really apply to that phenomenon). Your recordings should have about the same volume as most other recordings people will listen to. It should sound not too loud, and not too soft. It should sound … normal. Huh. Is that where the term came from?