On Tue, Feb 28, 2006 at 02:53:58PM +0200, Ivan Kalvachev wrote:
2006/2/28, Guillaume Poirier <docmaintainerwannabe@laposte.net>:
Hi,
Ivan Kalvachev wrote:
2006/2/28, Guillaume POIRIER <poirierg@gmail.com>:
Hi,
On 2/26/06, Diego Biurrun <diego@biurrun.de> wrote:
On Sun, Feb 26, 2006 at 11:39:40PM +0100, Guillaume POIRIER wrote:
--- DOCS/man/en/mplayer.1 24 Feb 2006 20:28:16 -0000 1.1234 +++ DOCS/man/en/mplayer.1 26 Feb 2006 22:38:09 -0000 @@ -8550,6 +8549,16 @@ +The following option is only available in CVS version of XviD.
Either "in CVS versions" or "in the CVS version".
+Create n threads to run the motion estimation (default: 0). +The max number of threads that can be used is the picture height
maximum
Okay, thanks!
I'll commit the updated version sometime in the future, and drink the cola in the unlikely case of if it breaks smth.
I can't check at the moment, but i see that the "create->num_threads = xvidenc_num_threads;" is not shielded by any #if / #ifdef. In case it is not structure member since the beginning of api4 (xvid 1.0) it would surely break all previous xvid 1.x releases.
Nope, it's all right, it's been present at least present since Thu 13 Jun 2002: http://www.xvid.org/cvs/chora/co.php/xvidcore/src/xvid.h?login=2&r=1.10
Then commit.
BTW would thread=1 start encoding in single separate thread? This could boost speed on SMP machines without quality loss.
Huh? How would this work? The caller would just sit idle waiting for that thread to finish... Rich