Author Topic: Tools Programmer here  (Read 10356 times)

0 Members and 1 Guest are viewing this topic.

Offline Surixurient

  • n00b Hero
  • *
  • Posts: 8
    • View Profile
Tools Programmer here
« on: April 11, 2006, 12:17:25 am »
any need for a live-evil programmer to write custom timing/translating/encoding software.

ive been experimenting with microphone controled timing.  voice recognition to auto set subtitles and times (times are automatically "stickied" to areas of peaked audio levels in the raw)  this is still experimental though.  another tool ive written is a clap timer (yup like clap on lighting :) )  so you can time a file by loading in a text file with all the translated subs, then claping to start and stop.  the next avilable subtitle is loaded up when you clap , and its end time is set when you clap again.

timing while lying in bed watching anime!

yeah there pretty useless :)

but the offers there

Offline Tofusensei

  • Admin
  • Rakuen Hero
  • *****
  • Posts: 406
  • Gender: Male
  • Believe in love.
    • View Profile
Re: Tools Programmer here
« Reply #1 on: April 11, 2006, 01:52:50 am »
O_O;

Sounds like fun :D

Reminds me of how timing was done with ZeroG... You timed in real time and hit the space bar to signal changes :X

-Tofu
The alpha and the omega of Live-eviL.

Offline Sindobook

  • Maetel Hero
  • *****
  • Posts: 662
    • View Profile
Re: Tools Programmer here
« Reply #2 on: April 12, 2006, 11:07:07 pm »
Have you ever thought of just using a bandpass filter to filter out  everything but the fundamental frequency of human voice (typically 85-155Hz for males, 165-255 for females) and then a simple volume-based autotimer that could attempt to time based on the filtered audio?  It wouldn't work all the time, but it could at least give a human timer a head start for the easier lines. 

Offline Surixurient

  • n00b Hero
  • *
  • Posts: 8
    • View Profile
Re: Tools Programmer here
« Reply #3 on: April 13, 2006, 04:56:49 pm »
going through and doing that to a whole episode would make a mess.  but what i do now is simular, in that when they set a time, it snaps to the beginning of audio data (actualy before it because its nice to have subs apear a split second before the talking starts)

filtering the data would be left up to the user, because they load a .wav file and an .avi file. so they can load a filtered .wav if they want. using soundforge or something