Convert 3gp audio to Text (googl/speech-api)

3gp audio file (aac 44100hz) to flac then segmented then to Google speech-to-text


convert it first flac output at slower sample rate (step down 44100 to 16000 )
 ffmpeg -y -loglevel verbose -i voicet.3gp -ar 16000 -ab 16000 voice.flac

Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'Downloads/test.3gp':
  Metadata:
    major_brand     : 3gp4
    minor_version   : 0
    compatible_brands: isom3gp4
    creation_time   : 2012-10-09 16:57:24
  Duration: 00:00:04.57, start: 0.000000, bitrate: 69 kb/s
    Stream #0:0(eng): Audio: aac (mp4a / 0x6134706D), 44100 Hz, mono, s16, 64 kb/s
    Metadata:
      creation_time   : 2012-10-09 16:57:24
      handler_name    : SoundHandle
The bitrate parameter is set too low. It takes bits/s as argument, not kbits/s
Output #0, flac, to 'voice.flac':
  Metadata:
    major_brand     : 3gp4
    minor_version   : 0
--- END STep
---split file to 10 Sec segments

 ffmpeg -i voice.flac -y -c copy -flags global_header -map 0 -f segment -segment_time 10 -segment_list fileList.txt ../data/beacon/segment_%02d.flac

[flac @ 0x2bfe260] max_analyze_duration 5000000 reached at 5040000
Input #0, flac, from 'voicet.flac':
  Metadata:
    MAJOR_BRAND     : 3gp4
    MINOR_VERSION   : 0
    COMPATIBLE_BRANDS: isom3gp4
    ENCODER         : Lavf54.29.105
  Duration: 00:00:54.03, bitrate: 163 kb/s
    Stream #0:0: Audio: flac, 16000 Hz, mono, s16
Output #0, segment, to '../data/beacon/segment_%02d.flac':
  Metadata:
    MAJOR_BRAND     : 3gp4
    MINOR_VERSION   : 0
    COMPATIBLE_BRANDS: isom3gp4
    encoder         : Lavf54.29.105
    Stream #0:0: Audio: flac, 16000 Hz, mono
Stream mapping:
  Stream #0:0 -> #0:0 (copy)
Press [q] to stop, [?] for help
size=       0kB time=00:00:54.00 bitrate=   0.0kbits/s
video:0kB audio:1071kB subtitle:0 global headers:0kB muxing overhead -100.000000%

--> list OUT files each 10 second
rob@ beacon$ ls -l *flac
-rw-rw-r-- 1 rob rob 220508 Oct 19 15:37 segment_00.flac
-rw-rw-r-- 1 rob rob 213396 Oct 19 15:37 segment_01.flac
-rw-rw-r-- 1 rob rob 203085 Oct 19 15:37 segment_02.flac
-rw-rw-r-- 1 rob rob 204558 Oct 19 15:37 segment_03.flac
-rw-rw-r-- 1 rob rob 213091 Oct 19 15:37 segment_04.flac
-rw-rw-r-- 1 rob rob  92195 Oct 19 15:37 segment_05.flac

--> process the above list submitted to google/speech-api for text OUTPUT

$ for f in *flac; \
 do curl -X POST -H "Content-Type: audio/x-flac; rate=16000" \
 -T $f "https://www.google.com/speech-api/v1/recognize?xjerr=1&client=speech2text&maxresults=1&lang=en-US"; \
  done

--Responses to 6 requests: text in 'utterance' TAG

{"status":0,"id":"c90030465ed0deb7c0723687a0657c02-1","hypotheses":[{"utterance":"streaming service did you did on Netflix in 2007 offering movies and television programs in a","confidence":0.7910386}]}
{"status":0,"id":"8b8ac5cd6b8a1effc350ab55fb3c7123-1","hypotheses":[{"utterance":"sweet corn on with subscribers for the price of a subscription to Netflix which is about 10 dollars a month you could walk","confidence":0.8551374}]}
{"status":0,"id":"91c1ab4c90c557332436ffa7b06d7e55-1","hypotheses":[{"utterance":"is many movies as you want and if you didn't like 1 you could start another it brought channel surfing to movies cancel","confidence":0.76847774}]}
{"status":0,"id":"0550f4606da7bb58458fc7ae379f3ba3-1","hypotheses":[{"utterance":"says for many people Netflix was the first glimpse of a kind of content Holy Grail limitless choice","confidence":0.80531996}]}
{"status":0,"id":"9125c7c98f5a83170421398918bcd3b6-1","hypotheses":[{"utterance":"all on demand available on any internet connected device by 2010 streaming have become a major part of","confidence":0.8495959}]}
{"status":0,"id":"81dc3362d64ad2c6af3ec2319fce4f9b-1","hypotheses":[{"utterance":"business its stock price sword","confidence":0.66059893}]}