<div class="gmail_quote">On Sun, Mar 15, 2009 at 9:43 PM, Adam Langley <span dir="ltr"><<a href="mailto:alangley@winscribe.com">alangley@winscribe.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Thanks for the insight Carsten.<br>
Basically, we're building a dictation system, with SPEEX at its core. Currently we have this working but the problem is, the SPEEX encoding process takes approximately 1 minute per MB of PCM audio, resulting in lengthy 'encode' jobs after a user has finished their dictation.<br>
I am attempting to find a way around this, and my hope was that I could perform some type of incremental encode during the record process, yet still enable the user to trim/rearrange their dictation without requiring a complete re-encode.<br>
And most users do edit the file to some degree...<br><div class="im"></div></blockquote><div><br>At least you've picked the right tool for the job. That's good. You might want to ask for advice on the mailing list that's dedicated to Speex. Speech coding is an entirely different animal than general-purpose audio encoding, and you have a better chance of reaching Speex experts, including Jean-Marc himself, on the Speex list.<br>
<br>A Speex expert will be much better qualified to gauge the feasibility of your approach, and might be able to provide insights for how to avoid or reduce the artifacts you might get from naive frame-splicing. Then again, you could just try the naive frame-splicing to hear what artifacts you'll get. Maybe it won't be too bad.<br>
<br>-Carsten<br><br></div></div>