[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lojban] uploaded them - may need redoing/recoding (was: missing audio files)



This is the bash line Dag used to generate the other audio:

cat gismu_places | while read v; do echo $v; espeak -vjbo --stdout "$v" | lame -S - "valsi/$v.mp3"; done

On Sun, Mar 7, 2010 at 9:01 AM, Timo Paulssen <timonator@perpetuum-immobile.de> wrote:
Jonathan Jones wrote:
> Alright, so if we can get someone to upload the audio for those 7 gismu
> sumti, we'll have the complete set. (And then I can download them, delete
> the cmavo, and combine them all into one archive.)

Another job easily done with one line of script :D

read lines you paste into the terminal and speak&encode and save to the
appropriately named file:

 while read WORDS; do echo $WORDS | espeak -v jbo -w /dev/stdout | lame\
--preset phon+ - "$WORDS".mp3; done
lo se rutni
lo te dzena
lo te lanme
lo te mensi
lo te solri
lo te stagi
lo te tamji

http://www.wakelift.de/lojban/missing_espeak_valsi.zip

unfortunately the format differs from the original files, but i couldn't
immediately figure out what preset would fit and i feared i might lose patience
fiddling with parameters, so whoever wants to tell me, go ahead, or use the line
up there to do it automatically.

lo se rutni.mp3: MPEG ADTS, layer III, v2,  64 kbps, 16 kHz, Monaural
../lo zvati.mp3: MPEG ADTS, layer III, v2,  32 kbps, 22.05 kHz, Monaural

mu'o mi'e timos




--
mu'o mi'e .aionys.

.i.a'o.e'e ko klama le bende pe denpa bu