I just discovered the awesomeness of this program yesterday and so am beginning to migrate everything in from my various other writing applications. I’m wondering if I can create tags in scrivener to switch between voices with Speech Control?
As a dyslexic screenwriter, Text to Speech (TTS) is a personal savior for proof-reading and hearing dialogue spoken out loud. The best tool in Final Draft is the ability to “Assign Voices" to my characters. In Scrivener I found the Speech Control tool will read text aloud and record it to iTunes as a spoken track (awesome and not in Final Draft).
But it seems to be limited to a single narrator’s voice. Is there a way for me to do ‘assign voices” within Scrivener? I assume it might require some automation or additional/specialized software but it seems tantalizingly possible. I’ve been playing around with the trial versions of some other TTS applications like Wrise and GhostReader Plus, that better integrate Scrivener (for listening to and editing individual scenes vs. than the entire script in final draft).
In these specialized TTS programs you tag the text with the voice to speak it (you can also manipulate voices) but if you open a scene in their text editors you’ll need to individually replace each character name with a tag for the character’s voice, which is a cumbersome non-writing task. If you want to hear rewrites as you make them, it means you end up editing your script outside of Scrivener’s wonderful interface and no longer benefit from it’s interconnectedness.
Is it something I could potentially create a meta-data tag for and then automate a find and replace function to carry out? Any idea of a path to a viable workaround?
I’m hoping so given how vibrant the Scrivener community is! If anyone can point me toward some resources on how to get started, I’d be grateful. Glad to have this product and you lovely folks.
Highlight the text you’d like to have spoken and then ctrl+click to open a menu, the bottom option should be called “services” in here in the I get a sub menu “send to evernote”, etc. and one option is “send as spoken track to iTunes” which is a very hand feature indeed.
That is a brilliant idea. Speaking dialogue (using a dictation application or Siri) makes for better dialogue but it can be cumbersome and inappropriate in some situation (can’t be dictating a sexy scene in a crowded commuter train). Playing dialogue back later is a great idea.
With you on the dyslexia too. But maybe if everyone, dyslexic or not, listened to their dialogue the quality of fiction would go up.
Adding a track to iTunes is good, especially if you want to listen to what you’ve written when you’re out and about.
But if you just want your Mac to read text to you (in any app) you need to visit sys prefs > speech > text to speech and turn on “speak selected text…”. Then just highlight the text you want (or put your cursor at the start of a block of text to get the Mac to read the whole text), press CMD T and your Mac will read the text to you, without having to create an iTunes file. CMD T stops it again.
This is very useful for proofing: it can only read out what you have written … so it highlights typos that your eyes might miss. It also helps you to hear the rhythm of what you have written fairly accurately.
It works for single words, phrases, sentences, paragraphs, or even whole novels.
I created an audiobook (mainly for visually-impaired people) using the add-to-iTunes facility. Been downloaded by people around the world thanks to Twitter and recommendations by numerous charities supporting visually-impaired people.
There’s no built-in way to do this, no. The Mac OS X text-to-speech function isn’t controllable by commands embedded in the text. You would need another program, designed to scan the text for embedded voice change commands. You would then probably have to compile to a text file and open it with that program, which would feed the text+commands to the OS.
I can’t say I’ve ever run across such a program though. Sorry.
Ah, thanks - I was just looking through the menus. I found Edit | Speech | Start Speaking. But I didn’t right-click, so I didn’t see the send to iTunes option.
I was also looking for a “record what you hear” option in the sound preferences, but I gather you need to install soundflower or something to do it that way.
Edit: I actually recently had a need to anonymize audio - take some spoken audio, but change the voice so that the speaker couldn’t be identified. If there was any way to get the TTS system to use tags to automatically change voices - that would be too cool…
Actually Windows has been making it harder to record what you hear as well in recent versions - probably the recording industry leaning on them behind the scenes…
My old iMac and MBP both have headphone and mic sockets. I plug a cable between the two (headphone and mic sockets on whichever machine I am using) and then record streams using QuickTime. Cable looks like the link below: don’t know what it is called. img.alibaba.com/photo/11730506/A … _Cable.jpg. Cost about £1—£2 IIRC.
Newer Macs only have one headphone socket, so I assume they must need a headphone/USB combo of some kind.
Or, by the sound of things, Soundflower might do what you want. Never used SF, as the cable solution works so well. Expect another Mac user will be able to help if they read this thread.
Yeah, I’ve got a new mini with both ports, so a loopback cable would also do the job for that.
On Windows though, (I forget if I turned it on or if it came with my sound card driver) there was a “What U Hear” option in the recording mixer. So I was just seeing if there was an equivalent of that.
Another thing I like from QuickTime is the one-click ability to output just the audio from a video recording. Allows me to study how narrative and dialogue work without images. Grabbing the audio and putting it on a Shuffle is a good way of working while exercising.
the third party programs Wrise or Ghostreader can carry out the voice switching but it requires replacing each “CHARACTER” element to be replace with a voice tag. Any thoughts on how to automate a find and replace function so I could replace these tags with custom metadata?
I would also agitate for at least some Scrivener-to-GhostReader (or other TTS engine) functionality.
Even at the level of a compile target – Generate an entire audiobook, where each scene (or chapter, since compile is so configurable!) is its own individual .m4a file, that could be imported into iTunes as a single playlist (or Album)*.
Being able to tag-which-voice, tag-reading-pace, tag-silence etc would be nice, but I really would be happy first with being able to compile to .m4a without having to export as individual files, load as indiv files into GhostReader, generate .m4a individually, upload to iTunes (individually), change the meta-data for each individually loaded file… etc.
Just a compile would be nice.
{tag voice=“whinychild”}Is that too much to ask{/tag}
iTunes and export to iOS has some sort of weird non-functionality, like Apple left something out accidentally. If you mark a file as an audiobook, it will download on sync – but it won’t be shown anywhere. You have to create a playlist (not an album, also not available from the Artist listing), and add the audiobook files to the playlist, and sync -that-. Then, hay presto, you can listen to your artful prose while driving to/from work…
