Re: Limiting intersample peaks
Steve Harris <[email protected]>
| Newsgroups | gmane.comp.audio.jamin.devel |
|---|---|
| Message-ID | <[email protected]> |
On 20 Dec 2007, at 17:07, John Rigg wrote: > > Is this suggestion impractical? It would obviously increase > CPU load, but OTOH the interpolating algorithm wouldn't need > to be as high quality as one used for SRC, as long as it could > detect the peak level well enough. It's not impractical, but it would have to be optional. It is necessary to use something on the order of SRC's sinc interpolation otherwise you won't be detecting the true peaks, which is the problem. It may be that we only have to interpolate up by a few times, but I couldn't say for sure without trying it. It would also complicate the latency issue, which is already quite complex. A good idea though. - Steve ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/