[Speex-dev] Speex Codec Delay Problem

[John Ridges]

I believe if you add the *decoder* lookahead time to the encoder 
lookahead time you are already referencing, you will get the numbers 
that you have calculated.

--John

On 4/11/2012 1:00 PM, speex-dev-request at xiph.org wrote:
> Message: 1
> Date: 10 Apr 2012 22:07:20 +0200
> From: Thilo K?hler<koehlerthilo at gmx.de>
> Subject: [Speex-dev] Speex Codec Delay Problem
> To: speex-dev at xiph.org
> Message-ID:<20120410200733.53EAF100D06 at fraxinus.osuosl.org>
>
> Hello All!
>
> SPEEX introduces an additional delay to the audio data,
> I found out by reverse enginiering (it is NOT the lookahead time):
>
> narrow band : delay = 200 - framesize + lookahead = 200 - 160 +  40 =  80
> samples
> wide band   : delay = 400 - framesize + lookahead = 400 - 320 + 143 = 223
> samples
> uwide band  : delay = 800 - framesize + lookahead = 800 - 640 + 349 = 509
> samples
>
> To get the timing right, you must skip those samples before you get the
> actual audio
> data you have feeded into the codec.
> Or, better, feed the encoder with (framesize - delay) dummy samples first,
> throw away
> the entire first frame, and continue encoding. This avoids that the first
> data you feed in must
> share its bits with the preceeding zeros, and you get a more compact
> encoding.
> (this might matter if you have a lot of short samples).
>
> This all sounds like a bad hack, and actually it is. What do the auhtor(s)
> of SPEEX think,
> can this be considered as a bug and be fixed?
>
> I ran into this problem and it took me a while to work around it, and I can
> observe
> on the internet that other people have problems with this too.
> If you encode 1000 samples, and then decode 1000 samples, you expect
> to get the same 1000 samples back.
>   (or, if the codec is lossy, an approximation)
>
> Regards,
>
> Thilo
>
>
> On 10.04.2012, you wrote:
>> Send Speex-dev mailing list submissions to
>>    speex-dev at xiph.org
>>
>> To subscribe or unsubscribe via the World Wide Web, visit
>>    http://lists.xiph.org/mailman/listinfo/speex-dev
>> or, via email, send a message with subject or body 'help' to
>>    speex-dev-request at xiph.org
>>
>> You can reach the person managing the list at
>>    speex-dev-owner at xiph.org
>>
>> When replying, please edit your Subject line so it is more specific
>> than "Re: Contents of Speex-dev digest..."
>>
>>
>> Today's Topics:
>>
>>     1. (no subject) (Bush Josh)
>>
>>
>> ----------------------------------------------------------------------
>>
>> Message: 1
>> Date: Tue, 10 Apr 2012 05:59:42 +0100
>> From: Bush Josh<darthbush at hotmail.com>
>> Subject: [Speex-dev] (no subject)
>> To:<a2xde4 at msn.com>,<pink.caro at hotmail.com>,<avahod at hotmail.co.uk>,
>>    <alexinanutshell at hotmail.co.uk>,<speex-dev at xiph.org>,
>>    <finchn.n1c0ll3 at hotmail.co.uk>,<haleigh-x-x-x at hotmail.co.uk>,
>>    <dujimache at hotmail.com>,<james_e397 at hotmail.com>
>> Message-ID:<DUB113-W36807C4B7517AB0B77B5A1A8340 at phx.gbl>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>>
>>
>>
> http://www.bbqol-tr.com/wp-content/themes/timecrunch/rytj.html?mg=ghy.jng&amf=fh.msg&gsyj=uylp
>>                       -------------- next part --------------
>> An HTML attachment was scrubbed...
>> URL:
>>
> http://lists.xiph.org/pipermail/speex-dev/attachments/20120410/c73280ab/attachment-0001.htm
>>
>> ------------------------------
>>
>> _______________________________________________
>> Speex-dev mailing list
>> Speex-dev at xiph.org
>> http://lists.xiph.org/mailman/listinfo/speex-dev
>>
>>
>> End of Speex-dev Digest, Vol 95, Issue 4
>> ****************************************
>>
> Regards,
>
> Thilo Koehler