Date: Sun, 3 Nov 2002 06:59:52 +0100 From: Steve O'Hara-Smith <steve@sohara.org> To: Mario Sergio Fujikawa Ferreira <lioux@FreeBSD.org> Cc: carl@stagecraft.cx, freebsd-multimedia@FreeBSD.ORG Subject: Re: ffmpeg compile optimisation helps no end! was Re: FreeBSD 4.7R: no sound when recording with ffmpeg? Message-ID: <20021103065952.41c47392.steve@sohara.org> In-Reply-To: <20021102220232.40685.qmail@exxodus.fedaykin.here> References: <20021028063325.224adec5.steve@sohara.org> <CB5FB6EA-EA6F-11D6-8981-0050E445BE6D@stagecraft.cx> <20021102173507.352ddc71.steve@sohara.org> <20021102220232.40685.qmail@exxodus.fedaykin.here>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, 2 Nov 2002 19:02:10 -0300
Mario Sergio Fujikawa Ferreira <lioux@FreeBSD.org> wrote:
MSFF> > I think this is worth committing, with it I can use default
MSFF> > motion estimation and even high quality settings at full
MSFF> > resolution without dropping frames.
MSFF>
MSFF> Humm, I am improved a bit on your idea. :)
Oh good.
MSFF> I tested these changes with several mpeg output formats:
MSFF> mpeg1video mpeg4 msmpeg4v1 msmpeg4v2 msmpeg4. Everything seems fine
MSFF> but I did not benchmark anything. Steve? Could you try some
MSFF> benchmarks since you've been timing stuff? Unfortunaly, I do not
MSFF> have a booktree board to test capture performance.
CPU utilisation is down by 3-5% (of the total available) on my
AMD compared to my optimisation attempt. Usage seems to be less peaky
too. With motion estimation off I could probably record two streams at
once in real time (CPU usage for one is around 40% and steady, unoptimised
it is around 70% and peaky).
It's a winner!
--
C:>WIN | Directable Mirrors
The computer obeys and wins. |A Better Way To Focus The Sun
You lose and Bill collects. | licenses available - see:
| http://www.sohara.org/
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-multimedia" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20021103065952.41c47392.steve>