[x264-devel] [Tom Jacobs <T.R.Jacobs at lboro.ac.uk>] Re: Re: Rif: Re: JM 9.6 vs. x264 rev 228: results

[System administration]

 The deleted attachments are at:
    <http://www.videolan.org/~admin/20050523-videolan/coastguard-352-288-br.eps>
    <http://www.videolan.org/~admin/20050523-videolan/coastguard-352-288-rd.eps>


----- Forwarded message from Tom Jacobs <T.R.Jacobs at lboro.ac.uk> -----

From: Tom Jacobs <T.R.Jacobs at lboro.ac.uk>
Date: Mon, 23 May 2005 12:29:49 +0100
To: x264-devel at videolan.org
Subject: Re: [x264-devel] Re: Rif: Re: JM 9.6 vs. x264 rev 228: results
User-Agent: Internet Messaging Program (IMP) 3.2.5
X-Spam-Status: No, score=-4.4 required=5.0 tests=IN_REP_TO autolearn=failed 
	version=3.0.3


thanks for all your help. i have managed to make a few graphs :)

i have attached 2 of the graphs (not sure if attachements work in the
mailing list or not) showing how the multisliced implementation works.
these 2 graphs are taken with 25frame click with 1 ref frame. the major
difference between the sliced version seam to be at lower psnr the relative
bitrate of the multislices are bigger. this is understandable since the rate
control for multislice isnt that good.

having said this what do people think? is it aceptable (assuming im getting
good benifits from using my multi-prossor system)?

thanks

Tom


Quoting Tuukka Toivonen <tuukkat at ee.oulu.fi>:

> On Fri, 20 May 2005, Tom Jacobs wrote:
>
> > ive had a look at your scripts for doing the graphs and im alittle
> stuck.
>
> No wonder. The script is absolutely horrible; I wouldn't be surprised
> if writing a new one would be easier than understanding plot.sh ...
>
> > my problem is im not sure how to edit plot.sh to take into account
> these
> > changes. also when have changed plot.sh im not sure what the inputs i
> need
> > to pass to it.
>
> I use plot.sh like this:
>
> ./plot.sh "" "" "o0r0-250-0-2-all-5 o6r1-250-0-2-all-5"
>
> when I have file names like this
>
> x264-coastguard-352-288-o5r0-250-0-1-all-5-20.log
>
> The first parameter is the set of test sequences. I always leave it
> empty (""), and the script will then determine automatically what log
> files
> are available and use all of them.
>
> The second parameter is set of JM encoding methods, since we
> test x264, we can leave it empty too.
>
> The last parameter is the set of x264 encoding methods, starting
> after the image size "...352-288-" and ending before QP
> parameter value, which is the last number "...-20.log".
>
> So, after we remove the prefix and postfix from the example log filename,
> we get "o5r0-250-0-1-all-5" which is the string describing the "coding
> method" part of the log file filename.
>
> You should move qp to last, something like
>    filename-width-height--frame-reframe-analyse-slice-qp.log
> and then run plot.sh like
>
> ./plot.sh "" "" "-frame-reframe-analyse-slice
> -frame-reframe-analyse-slice"
>
> (replace frame, reframe, analyse, slice with their actual values)
>
> Next I should describe how plot.sh works; but I must leave now so I'll
> left
> it to next time.
>
> Anyway, shortly:
> the loop "for met in $JMMETS; do" loops over files encoded with JM;
> you could probably remove that loop completely.
> $qpva = QP value
> $met = coding method
> $seq = video sequence...
>
> (must go now. bye)
>
> --
> This is the x264-devel mailing-list
> To unsubscribe, go to: http://developers.videolan.org/lists.html
>
>

----- End forwarded message -----

-- 
System administration <admin at via.ecp.fr>
VIA, Ecole Centrale Paris, France

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