Re: [codec] #25: Supported audio contents

"Dr. Christian Hoene" <> Sun, 02 May 2010 12:11 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id A99E83A68F0 for <>; Sun, 2 May 2010 05:11:45 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -3.966
X-Spam-Status: No, score=-3.966 tagged_above=-999 required=5 tests=[AWL=-0.317, BAYES_50=0.001, HELO_EQ_DE=0.35, RCVD_IN_DNSWL_MED=-4]
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id 9-fPxqlt030Q for <>; Sun, 2 May 2010 05:11:44 -0700 (PDT)
Received: from ( []) by (Postfix) with ESMTP id 453943A68C8 for <>; Sun, 2 May 2010 05:11:43 -0700 (PDT)
Received: from ( []) by (8.13.6/8.13.6) with ESMTP id o42CBSnK011916; Sun, 2 May 2010 14:11:28 +0200
Received: by (Postfix, from userid 30) id 7C4B57241D7; Sun, 2 May 2010 14:11:27 +0200 (CEST)
Received: from ([]) by (Horde Framework) with HTTP; Sun, 02 May 2010 14:11:27 +0200
Message-ID: <>
Date: Sun, 02 May 2010 14:11:27 +0200
From: "Dr. Christian Hoene" <>
To:, codec issue tracker <>
References: <>
In-Reply-To: <>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; DelSp="Yes"; format="flowed"
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
User-Agent: Internet Messaging Program (IMP) H3 (4.3.6)
X-AntiVirus-Spam-Check: failed (checked by Avira MailGate: version: 3.0.0-4; spam filter version: unknown; host: mx05)
X-AntiVirus: checked by Avira MailGate (version: 3.0.0-4; AVE:; VDF:; host: mx05); id=318-l0sojS
Subject: Re: [codec] #25: Supported audio contents
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Codec WG <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Sun, 02 May 2010 12:11:45 -0000


some time ago I wrote that the CODEC shall support these contents:

"  At all bitrates the ACS must deliver speech in any language at good
    quality. The ACS must be tested for different speakers and at least
    with two languages and should support tonal languages as well.

    Frequently, speech needs to be transmitted not only without
    background noise but also at conditions including car, office and
    street noise. Background signals shall be considered not as the noise
    but as a part of the signals that convey information. Background
    signal can include background music at a SNR of 25 dB, office noise
    at a SNR of 20 dB, car noise at a SNR of 15 dB, babble Noise at a SNR
    of 25 dB, interfering talker at a SNR of 15 dB and street noise at a
    SNR of 20 dB.

    At high bitrates the quality must be excellent for any audio signal,
    especially music. Stereo is considered as a must. Also, for high
    quality audio conferencing, reverberant input signals should be
    considered for testing the modes.

    The speech and audio signals might have varying loudness. The
    transmission shall support a wide range of dynamics. The nominal
    input level of -36 dB, -26 dB and -16dB with respect to the
    overlapping bandwidth limit (OVL) point (-20 dBm0). "

Actually, I copied those requirements from G.718. The testing plan for  
G.719 considers mixed content (ads, film trailer, news with jingles)  
and music (classical and modern).

Are they important for the Internet CODEC as well?

With best regards,


> #25: Supported audio contents
> ------------------------------------+---------------------------------------
>  Reporter:  hoene@…                 |       Owner:
>      Type:  defect                  |      Status:  new
>  Priority:  major                   |   Milestone:
> Component:  requirements            |     Version:
>  Severity:  Active WG Document      |    Keywords:
> ------------------------------------+---------------------------------------
>  Which kind of audio content should be used to test the CODEC?
>  Of course speech. Music at all rates? Mixed contents? What is about
>  background noise and reverberations?
>  Currently, the draft says: "Quality should be measured for multiple
>  languages, including tonal languages.  The case of multiple simultaneous
>  voices (as sometimes happens in conferencing) should be evaluated as
>  well."
> --
> Ticket URL: <>
> codec <>