In Part 5, I said I’d run through my audio-editing process. This is purely for the benefit of anyone who’s thinking of producing their own audiobooks, but who doesn’t have the first clue about editing.
I am not claiming this to be the only or best way to edit audio using Audacity. On the contrary, it is not even an advisable method because it is massively time-consuming.
As I write this, I’m picturing experienced audiobook producers rolling their eyes. What a ludicrously time-intensive way of doing things, I imagine them thinking. I completely agree with them. There must be quicker, more efficient ways of achieving the same outcome.
What this method has going for it is that it works—i.e. it results in audiobooks that meet Audible’s production standards—and works for narrators, like me, who don’t have a professional recording space and who aren’t professional narrators. I was extremely doubtful that my efforts would pass Audible’s quality control checks—why would they with the limitations on my recording studio and narration capabilities (see Part 5)? To have had both short story collections accepted first time without the need to make any changes was a huge boost. It also makes me reluctant to depart from the method that I know works, no matter how painstaking it is.
Painstaking is right. I have speeded up a little, but my editing time probably exceeds half an hour per completed minute of recording time. When you consider that the novel I’m currently producing in audio—The Beacon —is coming in at over thirty minutes per chapter, and there are twenty-three chapters altogether, that’s a significant time commitment.
Seriously, I’m not recommending you use my editing process. If you look around, you should be able to find a far more efficient method—if you do, please let me know. I’m setting out what I do for those who can’t find another way of doing it to a standard that meets Audible’s requirements.
Editing – Part 2
Always back up first—you really don’t want to have to make a new recording if something goes wrong during editing and you lose the original. I usually export the raw clip from Audacity as a WAV file and then save that file to the cloud.
Listen to the entire recording, deleting the mistakes. By ‘mistakes’ I mean the sections that I mucked up during recording or an external noise intruded or whatever and I noticed and so was able to re-record the mucked-up section immediately. There’s something quite satisfying about deleting the duff bit, leaving only a popping or clicking noise where that bit was. (That click will be eradicated later—don’t worry about it, or any other unwanted sounds, now.)
Raw recordings of each chapter of The Beacon might be as long as fifty minutes—I told you I make a lot of mistakes during narration. By the time I’ve completed Step 1, the recording will typically be reduced to around forty minutes.
Step 1 is easy since it is simply a case of deleting, without being concerned about removing clicks, etc. There’s no finesse required here and it might take me around an hour or two, depending on the length of the raw recording.
I now need to create my ‘good silence’. I usually record around thirty seconds of silence after speaking the final word. This gives me plenty of ambient room noise to play with.
Although I’m only looking for around two seconds of good silence for the editing process, Audible requires around four seconds of room noise at the end of each chapter so I make sure I have at least four seconds at this stage.
What do I mean by ‘good’silence’? It’s ambient room sound (a distant background hum) without any external noises like traffic or breathing or rustling. It will show on Audacity as a flat line, unbroken by the spikes that represents sounds.
Although I sit as still as a statue to record the thirty seconds of silence, you can guarantee my stomach will rumble or a noisy vehicle will go past or the house will creak for no apparent reason. (You notice sounds like that when you’re trying to be especially quiet.) So I need to remove those extraneous noises, leaving only the ambient sound, using the effect ‘Crossfade Clips’ (see below).
This is the time consumer. This is where the attention to detail comes in.
First things first—I add a second track (from the dropdown menu ‘Tracks’) and then copy around two seconds of the good silence I created in Step 2 to the clipboard*. You can paste that clip as many times as you want during each session, but it won’t remain in the clipboard after you close Audacity down. I therefore begin each editing session by going to the end and copying the two-second clip of good silence before resuming where I left off.
It’s then a case of working my way through every second of the recording to:
– shorten pauses between sentences and paragraphs to make them roughly the same length,
– insert a two-second pause between scenes, and
– remove unwanted noises: breathing, creaks, rustling, clicks, slappy mouth sounds (no matter how careful I am, I will inevitably make a few per recording that the microphone gleefully picks up), banging doors, passing vehicles, whatever.
There are two main methods I use—which one will depend on the type of change I’m trying to make. You can only work this out through trial and error initially, but it gets much easier the more accustomed to it you become. Don’t be afraid to experiment—one of the big pluses of Audacity is that it allows you to undo any number of steps (within that session), so if you make a mistake, simply undo it and try again.
I use this mainly to decrease spaces between sentences, to shorten mid-sentence pauses and to get rid of clicks left over from Step 1. It can also be used to eliminate clicking noises mid-word, though this can be tricky to achieve without losing part of the word and thus making it noticeable to a listener. You might need to use fade instead—you’ll have to experiment.
Here’s an example:
In A1, the space between two sentences, at over two seconds, is too long. I want to reduce it to about a second. Simply deleting a chunk of gap will introduce popping noises. To avoid this, use the Effect ‘Crossfade Clips’.
As shown in A2, highlight the area to reduce and apply ‘Crossfade Clips’ from the dropdown menu.
A3 shows the result. The gap has been reduced to just over a second. If I want to reduce it further, I can repeat the process, highlighting a smaller area if I only want to reduce the gap slightly. The larger the area highlighted, the greater the reduction. Only trial and error will give you a feel for it, but it will come with practice.
Fade Out/Fade In:
In the above example, I’ve reduced the gap between sentences, but I haven’t addressed the sounds (the clicks, pops, creaks and sighs) picked up by the microphone during recording and represented by the thicker dots and dashes. If I use Crossfade Clips again, the gap will become shorter, whch I may not want. Here’s how to eliminate unwanted sounds in the gaps between words and sentences without shortening the gap.
As shown in A4, apply Fade Out from roughly the end of the preceding sentence to around the midpoint of the gap. This will eliminate any unwanted noises, especially towards the end of the highlighted area. To get rid of noise from the start of the gap without shortening it, you can start/end a fade in a different place (but you’ll have to be careful not to introduce extra clicks or pops—generally speaking, so long as you perform a corresponding Fade In/Out that slightly overlaps the first one, you shouldn’t introduce any new clicks). Again, it’s a matter of trial and error—practice and you’ll become proficient.
As shown in A5, you then highlight from the centre of the gap to the first word of the next sentence and apply ‘Fade In’. You must ensure that the highlighted area begins just before the point where the previously highlighted area ended—the overlap I mention above—as otherwise you’ll create a new clicking/popping sound. Again, if this doesn’t eliminate sounds towards the end of the gap, you might need to do a Fade Out, but practice will make perfect.
The sounds you can see in the highlighted section in A3 have disappeared; the line in that section is now perfectly flat. However, if you do nothing further, you’ll be able to hear where the fades start or end, so you need to cover them up. This is where the clip of ‘good silence’ comes in. Simply paste it over the gap onto the second track, making sure the start and end of the clip corresponds with speaking on the main track (adjusting the length of the clip as necessary using ‘Delete’ from the ‘Edit’ dropdown menu) as otherwise the start/end of the clip will be heard as a clicking noise.
Okay, so that’s how I do it. If you’ve been struggling to find a way to edit that satisfies Audible’s requirements, feel free to copy what I do. It’s worth repeating the warning, though: it’s hugely time-consuming and there must be a better way to do it.
Next time, in a much briefer post, I’ll talk a little about mixing and mastering. (That makes it sound as though I know a lot about them—I don’t, I really don’t, but I know what to do to get it approved by Audible.) Till then…
* Since drafting this post, a way to reduce the editing time occurred to me. I recorded a lengthy section of ambient room noise. Then I used Crossfade Clips to remove any extraneous sounds, leaving only good silence. I spliced the clip together a few times (and used Crossfade Clips to conceal the joins) to leave me with a track of thirty-six minutes consisting entirely of good silence. It’s saved as a WAV file and labelled ‘Ambience’. When I embark on Step 3, the first thing I do is import ‘Ambience’ as the second track. This avoids having to paste a clip of silence over each edit and is saving me a fair amount of time overall. Wish I’d thought of it sooner. Here’s a screenshot of the current Beacon chapter I’m working on showing the Ambience track beneath the main track.