[Alsa-user] # of periods / input latency
Brought to you by:
perex
From: <ale...@gm...> - 2005-09-27 12:26:34
|
Hi to all, capturing the line in and measuring how long it takes I figured that the whole latency is twice the blocking delay of the soundcard : In case of using 48 kHz, 256 Samples/Frame I'm dealing with a blocking delay of 5,6 ms. Processing audio internally I get exactly this value. Capturing from line it gives me values of 11,2 ms. In fact I used 2 periods because 1 results in tons of under/overruns. My questions : Is there a way to reach capture values equal to the blocking delay or is it alway twice of that ? Does this maybe depend on the soundcard I'm using ? Is there a common way to process input signals in any soundcard or do exist different approaches (i.e. doublebuffering) ? Thanks in advance -- A l e x -- Dipl.-Ing. Alexander Carôt Email : Alexander@Carot.de Tel.: +49 (0)177 5719797 phd-candidate at www.isnm.de 5 GB Mailbox, 50 FreeSMS http://www.gmx.net/de/go/promail +++ GMX - die erste Adresse für Mail, Message, More +++ |