An audiobook is made in four basic steps:
1. A voice actor records the narration for the book.
2. The editor takes the raw audio recordings and edits these files for pacing, flow, and aesthetic flavor. The editor also makes note of any errors in the narration which require a re-record.
3. After making all re-records and inserting the new audio, the audio is listened through one more time. They call this a "QC pass."
4. Finally, the QC'd audio is mastered and encoded for delivery.
I was the owner of step two for all Audible Studios productions.
Do they read all of the dialogue for one character in a big chunk, then move on to the dialogue of the next character? Or do they read through the book naturally and switch voices? I would think that the latter would be very difficult.
Narrators actually do switch all character voices in real-time. This does not commonly cause issues that require recording to stop, as the narrators are usually very experienced professionals that come from the Broadway stage or from other voice acting professions (cartoons and the like.) I've also recorded and edited my fair share of bad voice actors (for instance, former Labor Secretary Robert Reich narrated his book 'Beyond Outrage' even though he was told that it really wasn't a good idea.)
That's a very interesting aside. I could see how a non-professional narrator would struggle with the demands of the process. Would it not still have made sense for someone like Mr. Reich from a personal branding point of view? He is, after all, a somewhat well-known public figure and paid public speaker.
Audiobooks are a very intimate media, and the wrong voice can really put off the experience. With training I think he could do it, but to just give a public speaker the task of narrating a book he/she wrote without such training is usually a recipe for disaster.
If you don't mind my asking, when people want to do their own reading and are willing to get training, how are they trained?
I ask because (like every geek) I sometimes think about doing a podcast some day, plus I just want to be a better public speaker, and there seems to be a lack of options between free/cheap resources focused on basic skills (like Toastmasters) and intensive training for people who want to do voiceover for a living.
Then I listened to Leonard Mlodinow's "The Upright Thinkers"; Oh god. Euclid's Window by Mlodinow was absolutely fantastic, and it was narrated by the slightly pompous-sounding -- but fitting -- Robert Blumenfield. Leonard had a somewhat slow, drawling voice, and he often stumbled over words. This was pretty disappointing because he seems like such a smart dude. He just shouldn't narrate his own books :)
The pauses between lines/phrases were often tightened up, yes, by pasting room tone over the pauses. This way, the near-silence of the room tone is consistent throughout each pause, which makes the narration easier to comprehend and allows the listener greater immersion into the experience.
The exact duration of room tone impacts the pacing of the narration. This is, in a way, the audiobook editor's "art." They can't over-edit because it would ruin the narrator's dramatic delivery - they need to be able to edit within the stylistic flow of the narrator. At the same time, if they are not careful enough to make the resulting delivery consistent, then the extraneous audio may distract listeners and lead to a lower quality result. That balance is acquired with experience.
The reason this all happens, really, is to ensure a clean "noise floor" - the background sounds heard in addition to the narrator's voice. The sound of the room and the gear, if you will. The narrators are such pros (my experience) - there are many who I could get away with not editing at all, if it weren't for the incredible need for near-silence throughout the entire finished production.
I was thinking I really wish there was a good app where you can listen to audio articles. It is partly solved in iOS as you can speak selected text, but problem is you cannot select all text easily and you don't wanna mess with it while driving.
Was wondering what would be a great UI for the transcriber and whether you could create p2p network to share the transcriptions.
Anyway, on iOS if you enable voiceover you can read the entire screen by swiping down from the top with two fingers (I think - check the Apple website for accessibility information if not!)
Another possibility i just realised is the possibility to pass texts from reader mode into 3rd party text-to-speech application. Will have to research some of them.
Honestly, though, it was a great job. It never felt exhausting. Not only was the pay rather good for a 21 year old fresh out of university - the job was really fun. I got to read medium-to-high quality novels and create the audiobook experience.
Later, I got to witness Audible's rise into the mainstream and evangelized producing audiobooks for ACX [1], bringing it to a new generation of actors and engineers.
(I discovered a couple years ago that I really like audiobooks.)