[x264-devel] Re: x264 error reporting

[Måns Rullgård]

"B. G. Kim" <bgkim at etri.re.kr> writes:

> Dear, L. Aimar.
>
> nice to meet you. I'm an researcher who is interested in video coding.......
>
> So, I usually have used x264 encoder for H.264/AVC contents in Linux
> (Red Hat 9.0, Intel 3.4 GHz CPU).
>
> At recent, I found a misoperation of the x264 during encoding the
> video as follows, and much slower than
>
> the previous verion (x264). I have updated it on 17, January 2005.
>
> ====== printed message ==============
> ~]# ./x264 -B 1024 -o ../x264Videos/starwars_vga_1Mbps_20050113.264 ../yuv/starwars_vga_250frame.yuv  640x480
>
> x264 [warning]: rc buffer size 0 too small
> x264 [info]: using cpu capabilities MMX MMXEXT SSE SSE2 
> x264 [warning]: buffer underflow -28264
> x264 [info]: slice I:1    Avg QP:31.00 Avg size:  8674 PSNR Mean Y:35.42 U:40.86 V:40.53 Avg:36.58 Global:36.58 MSE*Size:1.905
> x264 [info]: slice P:29   Avg QP:32.38 Avg size:  4192 PSNR Mean Y:35.41 U:41.25 V:40.91 Avg:36.60 Global:36.34 MSE*Size:0.973
> x264 [info]: slice I   Avg I4x4:37.4%  I16x16:62.6%
> x264 [info]: slice P   Avg I4x4:3.6%  I16x16:4.6%  P:34.4%  P8x8:2.5%  PSKIP:55.0%
> x264 [info]: PSNR Mean Y:35.41 U:41.24 V:40.89 Avg:36.60 Global:36.35 kb/s:868.3
>
> encoded 30 frames, 14.23 fps, 869.37 kb/s
>
> ~]#
> ===============================
>
> To the third row, there are some errenous operation in my opinion.
>
> As a result on another Linux PC, I've got the same message.
>
> I hope this report is helpful of the advanced development of x264. And 
>
> reply to me if you have any solution for this problem.

The first warning means you requested a constant bitrate without
specifying the buffer size, so a reasonable default value is used.
The second warning indicates (surprise) a buffer underflow.

Was your previous version from before the rate control was added some
time last summer?

-- 
Måns Rullgård
mru at inprovide.com

-- 
This is the x264-devel mailing-list
To unsubscribe, go to: http://developers.videolan.org/lists.html