Offtopic: Into Commodore 64 (6502) coding, pixeling or music?
Xentax is looking for new members for the C64 activities!
Just drop us a message at forum@xentax.com and join the Scene Team!

Forum rules: Click here

Looking for wav2text converter .....

Get help on any and all audio formats, or chip in and help others!
Post Reply
Xela
VIP member
VIP member
Posts: 225
Joined: Sun Jul 31, 2005 4:12 am
Been thanked: 1 time

Looking for wav2text converter .....

Post by Xela » Wed Oct 25, 2006 3:58 pm

We all know programs which convert text to speech. Are there any to convert the other way around :?: Freeware, if I may add and most importantly .... really working :)

Mirrodin
advanced
Posts: 71
Joined: Tue May 31, 2005 7:36 pm
Has thanked: 3 times
Been thanked: 1 time
Contact:

Re: Looking for wav2text converter .....

Post by Mirrodin » Wed Apr 16, 2008 5:02 am

The only one I can think of is in the retail side of the software industry: Dragon Naturally Speaking. I know Various games have implemented Speech to Text but it's usually implemented directly into the game's installation (pc's) or directly into the game engine (consoles).

grimdoomer
advanced
Posts: 70
Joined: Fri Mar 21, 2008 8:11 pm
Location: New York
Been thanked: 1 time
Contact:

Re: Looking for wav2text converter .....

Post by grimdoomer » Sun Apr 20, 2008 12:52 am

You can't take a audio file and have a program write the words to a text file or any other file. Each byte in the file represents a specific tone, which gets played through your speackers. So some ones voice could be any tone in the world. But im sure you could do it if there wasn't any background noise.

Mirrodin
advanced
Posts: 71
Joined: Tue May 31, 2005 7:36 pm
Has thanked: 3 times
Been thanked: 1 time
Contact:

Re: Looking for wav2text converter .....

Post by Mirrodin » Sun Apr 20, 2008 10:24 pm

grimdoomer wrote:You can't take a audio file and have a program write the words to a text file or any other file. Each byte in the file represents a specific tone, which gets played through your speackers. So some ones voice could be any tone in the world. But im sure you could do it if there wasn't any background noise.
Not true. To be technical, there are two important things to consider with an audio file (waveform) the bit depth, and the sampling frequency. Today's standard is based on the Nyquist theorum: sampling frequency = double the range of human hearing which is ballparked at 20hz to 20,000 hz (hertz = cycles per second aka frequency) bit depth determins how many bits it takes to define a singular sample. multiply this by the number of samples per second you're recording at (usually the standard today would be 44,100 samples per second) and then again by the length of the audio file to get the actual file size. It takes multiple samples to create a cycle for any complex waveform. Speech recognition software is pretty complex but it definitely exists. Like I said above, Dragon Naturally speaking is the only software I can think of that does it, and it's expensive.

http://www.nuance.com/naturallyspeaking/

The reason it gets that expensive is because they make the software interface with other applications like Outlook, Word, and more advanced features might include recognizing functions to do actions by voice in Access or Excel. If this is what you're looking at doing, I'd recommend it. However, it only works via live interaction.
If you're looking at doing some advanced archiving like taking audio files from a game, and saving a text file next to each one with the dialog, or even just saving a text file with the dialog from each wave file, you could cue up all the audio files into a media player, and change the input of the Naturally Speaking software to listen to all audio input from the OS. It's definitely possible, I've had to do this for several mods in the past.

Post Reply