1. 15 Jul, 2021 1 commit
  2. 09 Jul, 2021 5 commits
  3. 05 Jul, 2021 1 commit
  4. 30 Jun, 2021 7 commits
  5. 28 Jun, 2021 2 commits
  6. 25 Jun, 2021 2 commits
    • Mathieu Duponchelle's avatar
      tttocea608: don't pad up to first input timestamp · 23219c3c
      Mathieu Duponchelle authored
      When receiving data from a live upstream element, the first buffer
      may have a non-0 timestamp (running time), and the element incorrectly
      pushed padding buffers up to that timestamp, resulting in much confusion
      
      Part-of: <!527>
      23219c3c
    • Mathieu Duponchelle's avatar
      awstranscriber: further decouple output from input · 9415c502
      Mathieu Duponchelle authored
      As awstranscriber might in theory push out gap events without
      any flow of input data, it needs to send its mandatory events
      (stream-start, caps, segment) independently.
      
      In addition, track a start time and use it to offset the 0-based
      timestamps returned by AWS in order to output buffers timestamped
      in the running-time domain, and perform item timing adjustment
      only when dequeuing, instead of when queuing.
      
      Part-of: <!525>
      9415c502
  7. 23 Jun, 2021 1 commit
  8. 21 Jun, 2021 1 commit
  9. 20 Jun, 2021 2 commits
  10. 19 Jun, 2021 1 commit
    • Mathieu Duponchelle's avatar
      awstranscriber: make use of new result stability AWS API option · 640ce43f
      Mathieu Duponchelle authored
      <https://aws.amazon.com/blogs/machine-learning/amazon-transcribe-now-supports-partial-results-stabilization-for-streaming-audio/>
      
      Amazon seem to have realized the previous iteration of their API
      made it difficult to identify items from one result to the next,
      which made the element much more complicated than it should have
      been. With that new "stability" option, we can enqueue items as
      soon as they stabilize, and simply rely on the current index in
      the transcript to output them exactly once.
      
      This also means the "use_partial_results" is now useless, as there
      will be no difference in accuracy between a non-partial result and
      and of its stable items that might have been pushed from previous
      partial versions of the result.
      
      The property is removed, instead a new option is exposed to let
      users control how fast results should stabilize.
      
      This greatly simplifies the code, and also improves the output as
      punctuation doesn't need to be randomly discarded anymore.
      640ce43f
  11. 18 Jun, 2021 1 commit
  12. 16 Jun, 2021 1 commit
  13. 08 Jun, 2021 1 commit
  14. 05 Jun, 2021 9 commits
  15. 03 Jun, 2021 2 commits
  16. 31 May, 2021 1 commit
  17. 26 May, 2021 1 commit
    • Guillaume Desmottes's avatar
      meson: fix csound detection · 1ec1352c
      Guillaume Desmottes authored
      csound-sys can detect the system lib using either pkg-config or using
      the CSOUND_LIB_DIR env variable.
      The former case just work but the second is trickier as we need to
      ensure that CSOUND_LIB_DIR is defined when building.
      
      So we no longer try to detect the lib using find_library() if user
      didn't define the env variable as the build will fail later.
      
      Also explicitly pass the env variable to cargo so user can now call
      'CSOUND_LIB_DIR=/usr/lib64 meson build && ninja -C build'
      and have it work without repassing the env variable to ninja.
      1ec1352c
  18. 18 May, 2021 1 commit