[mythtv] [PATCH] time stretch with libsoundtouch

Ed Wildgoose lists at wildgooses.com
Thu Nov 4 08:17:16 UTC 2004


Mark Spieth wrote:

>>I'm also shooting for pretty high audio quality so if the quality is
>>better I'd also like libsamplerate to stay. Tempo-changing would be
>>*very* cool though, but maybe it's a bit heavy on the CPU and the
>>latency didn't seem that good. I guess a setting or a compile option
>>would be in place, perhaps with a fallback to libsamplerate for speed
>>change if one doesn't want to use soundtouch?
>>    
>>
>
>there is no problem for libsoundtouch and libsamplerate coexisting. this is
>how I have implemented it.
>libsoundtouch does not even get instantiated until the first non 1.0
>stretchfactor set. thus if you dont use it it has no impact. this should
>satisfy you I hope.
>  
>

Agreed.  There should be no disadvantage in having this available in 
general. As you say, lipsync is easily handled - actually I would also 
like to include a general param to allow lipsync offsets for people with 
video processors

I didn't look at the patch too hard yet, but in terms of control keys, 
we already have a speedup and slowdown option.  Did you piggyback onto 
that code?  (Which is probably the best way I think?). 

Ed W

P.S. If you wanted to fix something interesting, then I would personally 
love to be able to do like 25x speedup or more - ie for forwarding 
through the adverts.  Because the current method decodes all frames, I 
currently run out of processor power at around x6-x10 or so (I think?).



More information about the mythtv-dev mailing list