- From: Jim Barnett <1jhbarnett@gmail.com>
- Date: Tue, 14 Oct 2014 08:21:51 -0400
- To: public-media-capture@w3.org
For us and anyone else who is interested in using MediaRecorder to get audio samples to feed to a speech recognition system, the typical timeslice will be 200ms. - Jim On 10/14/2014 7:44 AM, Harald Alvestrand wrote: > In looking at the implementation of MediaStreamRecorder, we're finding > that we have to make an implementation choice (at least in version 1). > > One will make the most sense for use cases where "timeslice" is in the > order of tenths of seconds. > > The other one will make the most sense for use cases where "timeslice" > is on the order of tens of seconds to minutes - or "infinite". > > For the people who have definite ideas on what they intend to use > MediaStreamRecorder for - what order of magnitude of "timeslice" do > you think will be more important to you? > > Harald > > >
Received on Tuesday, 14 October 2014 12:22:18 UTC