<div>Thanx for your response,</div>
<div> </div>
<div>The standard allows large motion vectors indeed (I recall a quarter window as max). However, motion estimation techniques generally limit their search range to 16 (or something close to that). Anyway, I encoded the movies with x264 using the option "--hex --merange 24" and "--esa --merange 16", where 'merange' sets the motion estimation search range. Thus, in principle I should not find any motion vector larger than 24 and 16 respectively. But I think I did, and that worries me.
</div>
<div> </div>
<div>with kind regards,</div>
<div>Cor<br><br> </div>
<div><span class="gmail_quote">2007/7/31, Panagiotis Issaris <<a href="mailto:takis.issaris@uhasselt.be">takis.issaris@uhasselt.be</a>>:</span>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">Hi,<br><br>CAdevel schreef:<br>>[...]<br>> THE QUESTION:<br>><br>> Why does x264 create this large motion vectors? Is something wrong, or
<br>> do I misunderstand something? Do I interpret the motion vectors<br>> correctly (even the reference encoder seems to produce vector too large)?<br>><br>> I woudl really appreciate if someone could help me out on this.
<br><br>If I recall correctly, the H.264 spec allows motion vectors to be up to<br>2048 pixels in horizontal direction and 512 pixels in vertical direction.<br><br>With friendly regards,<br>Takis<br>--<br>vCard: <a href="http://issaris.org/pi.vcf">
http://issaris.org/pi.vcf</a><br>PGP key: <a href="http://issaris.org/pi.key">http://issaris.org/pi.key</a><br>_______________________________________________<br>x264-devel mailing list<br><a href="mailto:x264-devel@videolan.org">
x264-devel@videolan.org</a><br><a href="http://mailman.videolan.org/listinfo/x264-devel">http://mailman.videolan.org/listinfo/x264-devel</a><br></blockquote></div><br>