[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Ccrtp-devel] timing problem
From: |
Federico Montesino Pouzols |
Subject: |
Re: [Ccrtp-devel] timing problem |
Date: |
Tue, 22 Jul 2003 14:15:15 +0200 |
User-agent: |
Mutt/1.3.28i |
Hi. Yes, it loses precision. However, at first sight it seems
to me that the loss is not cumulative within ccRTP, so it shouldn't
result in a stationary error. I will test it later.
On Thu, Jul 03, 2003 at 11:09:47AM +0200, Luuk van Dijk wrote:
> Hi,
>
> I have 2 timing problems with a streaming audio app. over ccrtp.
>
> I'm using 1.0.1
>
> one is that although I tell my soundcard to sample at 44100 Hz,
> it is slow compared to the system clock: I start the timestamp with
> getCurrentTimestamp, and increment the timestamp by the number of
> samples, but then the ratio timestamp/system clock starts at 1 but
> withing some tens of seconds settles around .9825
>
> when the difference between my timestamp and getCUrrentTimsstamp reaches
> the ExpireTimeout all transmission stops. (of course)
>
> The second timing problem is that about half of my packets are dropped.
> I saw a related discussion on this list's archive, and it was presumably
> fixed, (somewhere in the calculation of the reminder of the timeout),
> but I still have the problem, although, admittedly, it gets worse if I
> replace
> send.tv_usec = (1000ul*rem) / (rate/1000ul); // 10^6 * rem/rate
>
> by .... = (1000*1000*rem)/rate
>
> still I think the first one loses precision at rate = 44100
>
> any ideas?
>
> thanks,
> Luuk
> _______________________________________________
> Ccrtp-devel mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/ccrtp-devel