Hi everyone<br><br>I am currently working on a research project trying to modify the x264 encoding part to improve the encoding-time.<br><br>One of the means we try to implement uses a direction detection of textures, and for that I would like to access to the data of the proccessed Macro Bloc during the mode prediction phase.<br><br>I modified <br><br>static void predict_16x16_mode_available( unsigned int i_neighbour, int *mode, int *pi_count )<br><br>into<br><br>static void predict_16x16_mode_available( unsigned int i_neighbour, int *mode, int *pi_count, uint8_t macrobloc )<br><br><br>in order to be able to access to the p_fenc[0] data, and process this information to detect the texture direction.<br><br>My problem is that I don't know how to handle this data structure. p_fenc is described in the common.h file, in the "pic" structure, as:<br> /* pointer over mb of the frame to be compressed
*/<br> uint8_t *p_fenc[3];<br>and I guess that p_fenc[0] is the luma part, and p_fenc[1] & p_fenc[2] are the chroma parts, but once I get the p_fenc[0], I don't know how to access to the different pixels of the MB.<br><br>I tryed to crowl in the code, thinking that this should have already been done somewhere else in the coder, but I always find some un-understandable code, relying on assembler code I guess...<br><br>As you probably noticed, I am not too familiar with the x264 code, neither with the C language actually, but I really need to perform this (school project).<br><br>Any help would be greatly appreciated. THanx in advance, and happy new year :)<br><br>Sam<br><p> 
<hr size="1">
Ne gardez plus qu'une seule adresse mail ! <a href="http://fr.rd.yahoo.com/mail/mail_taglines/trueswitch/*http://www.trueswitch.com/yahoo-fr/">Copiez vos mails</a> vers Yahoo! Mail