<div dir="ltr">Gregory,<div><br></div><div>That is good to know and if therefore the true apples to apples comparison is 0.9.14 at comp 10 and 1.1 at comp 5 then things are fine. My ears are not good enough to hear the difference so for speed I would target comp 5 or lower.</div>
<div><br></div><div>I just did a quick test and 0.9.14 at comp 10 was 3.872</div><div>1.1 at comp 5 was 5.218</div><div><br></div><div>So if the output is comparable then we do in fact see a speed improvement.</div><div><br>
</div><div>Thanks for pointing this out. Is it documented? I admit I have only read some of the documentation.</div><div><br></div><div class="gmail_extra"><br clear="all"><div><div>Best Regards,</div><div><br></div><div>
Stuart Marsden</div><div><br></div></div>
<br><br><div class="gmail_quote">On 17 December 2013 23:50, Gregory Maxwell <span dir="ltr"><<a href="mailto:gmaxwell@gmail.com" target="_blank" onclick="window.open('https://mail.google.com/mail/?view=cm&tf=1&to=gmaxwell@gmail.com&cc=&bcc=&su=&body=','_blank');return false;">gmaxwell@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im">On Mon, Dec 16, 2013 at 5:03 AM, Stuart Marsden<br>
<<a href="mailto:stuartmarsden@finmars.co.uk" onclick="window.open('https://mail.google.com/mail/?view=cm&tf=1&to=stuartmarsden@finmars.co.uk&cc=&bcc=&su=&body=','_blank');return false;">stuartmarsden@finmars.co.uk</a>> wrote:<br>
> I have just started trying Opus with a view to using it in a project. I am<br>
> interested in embedded hardware and tried it on the Raspberry Pi using the<br>
> raspbian distro.<br>
><br>
> The version of libopus in the repos is 0.9.14. I installed this and tried<br>
> encoding 2 minutes of speech from a librevox recording. It managed this at a<br>
> respectable pace for complexity 10:<br>
<br>
</div>Complexity 10 is new analysis code that didn't exist in prior<br>
versions, setting complexity 5 gets you basically the same analysis<br>
that the 1.0 version had.<br>
<br>
On x86 and modern arm cores with fast FPUs the other speedups are<br>
enough that complexity 10 is about the same speed in the old software<br>
or the new software (but with much higher and consistent VBR quality).<br>
But on chips with slow FPUs the new analysis code is much slower, in<br>
particular because it has not been entirely converted to fixed point<br>
(e.g. in the fixed point builds) which is what I believe you're seeing<br>
here.<br>
</blockquote></div><br></div></div>