[wdmaudiodev] Re: RTAudio packet streaming is inefficient by design?
- From: Eugene Muzychenko <reg.wad@xxxxxxxxxxxxxx>
- To: Robert Bielik <wdmaudiodev@xxxxxxxxxxxxx>
- Date: Thu, 21 Jun 2018 12:46:31 +0200
Hello Robert,
Actually, I've never understood how it is possible to have anything
else BUT 2 buffer notifications
With relatively large (dozens of ms) buffers, there is no reason to
have packet size (and number of notifications) more than 2, because
a time equal to a half of buffer duration is enough to react and
process data.
But with small (units of ms) buffers, the time is not enough. For
example, with 10 ms buffer, you have only 5 ms to get the CPU, while
internal kernel delays may be long (see LatencyMon measurements).
Using at least four parts/packets, you might have 7.5 ms, which is more
reliable.
Regards,
Eugene
******************
WDMAUDIODEV addresses:
Post message:
mailto:wdmaudiodev@xxxxxxxxxxxxx
Subscribe:
mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=subscribe
Unsubscribe:
mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=unsubscribe
Moderator:
mailto:wdmaudiodev-moderators@xxxxxxxxxxxxx
URL to WDMAUDIODEV page:
http://www.wdmaudiodev.com/
Other related posts:
