STFbreadcrumb separatorSTFsbreadcrumb separatorSTF Homepagesbreadcrumb separatorT021

STF_WebPage_Template

TestingTask Force T021:

Characterization of the New ETSI speech codec on subjective Test data bases - CNET step 2

Who we are:

 

STF Leader:    Joachim Pomy, Focus Infocom Gesellschaft für Informatik und Telekommunikation mbH, consultant@joachimpomy.de

STF Experts:  Jan Holub, Mesaqin.com s.r.o (Ltd.), jan.holub@mesaqin.com

Markus Schnell, Fraunhofer Institut für Integrierte Schaltungen IIS, markus.schnell@iis.fraunhofer.de

Maximilian Schlegel, Fraunhofer Institut für Integrierte Schaltungen IIS, maximilian.schlegel@iis.fraunhofer.de

What we do

There are two more topics which had not been included in TTF T005, because they could only be initiated after it was confirmed by TTF T005 that the New ETSI Speech Codec (LC3plus) fulfils the requirements set out in TS 103 624. This has now been confirmed by the subjective tests conducted by the TTF T005 team.

 

Therefore, the proposed activity under this new TTF consists of the following two topics, which are required in context of the New ETSI Speech Codec:

 

1)     According to TS 103 624 clause 5.5, the subjective tests shall be followed by a verification with the objective method ITU-T P.863.             
NOTE: At the time when TS 103 624 was drafted, there was the expectation that these tests would be provided by a specific STQ member at no extra cost. However, that company is no longer a member of ETSI and will not provide these tests.

2)     For the purpose of inclusion of the New ETSI Speech Codec in transmission planning, ITU-T SG12 is requesting ETSI to provide information on the so-called Equipment Impairment factor(s) ( abbreviation: Ie) of the New ETSI Speech Codec for future inclusion in ITU-T G.113.
Ie -Factors are a characterization of the impact of speech codecs on the perceived quality on a psychological scale; their values are available for all codecs which are of relevance in modern networks and terminals.                     
In order to fully establish the New ETSI Speech Codec internationally, the knowledge about the
Ie -Factors is required and it would be highly promotional for ETSI to provide the international community with the Equipment Impairment Factors for transmission planning for the New ETSI Speech Codec for all three instances of the Transmission Planning with the E-model, namely narrowband (NB), wideband (WB) and fullband (FB).   
In order to provide the values for FB, a new methodology will have to be established; this can be done based on existing work in the field, like (1) , (2), (3) or (4)
[1] where the Ie factor has been derived for the OPUS codec [2].



[1] (1) Sebastian Möller, Nicolas Côté, Valérie Gautier-Turbin, Nobuhiko Kitawaki and Akira Takahashi, "Instrumental Estimation of E-Model Parameters for Wideband Speech Codecs", EURASIP Journal on Audio Speech and Music Processing, vol. 2010, no. 1, pp. 1-16, 2010.

(2) S. Moller, A. Raake, N. Kitawaki, A. Takahashi and M. Waltermann, "Impairment factor framework for wide-band speech codecs", IEEE Transactions on Audio Speech and Language Processing, vol. 14, no. 6, pp. 1969-1976, Nov 2006.

(3) M. Al-Ahmadi, P. Pocta and H. Melvin, "Instrumental estimation of e-model equipment impairment factor parameters for super-wideband opus codec", Proc. Ir. Signals Syst. Conf., pp. 1-5, Jun. 2019.

(4) M. Al-Ahmadi, P. Pocta and H. Melvin, "Derivation of E-model Equipment Impairment Factors for Narrowband and Wideband Opus Codec Using the Instrumental Method," 2020 31st Irish Signals and Systems Conference (ISSC), 2020, pp. 1-6, doi: 10.1109/ISSC49989.2020.9180160.

[2] IETF RFC 6716: "Definition of the Opus Audio Codec"

Objectives of the work to be executed

1)     Objective assessment of the subjective test result already produced by TTF T005.
This cross check shall verify that the test results on the New ETSI Speech Codec (LC3plus) show no statistically significant differences between subjective or objective assessment.

2)     Determination of the Equipment Impairment Factors for the different bandwidth modes of the codec.

2a)  Test plan definition:      
As the exact procedure for FB Ie is currently not standardized, it is necessary to design and agree upon the complete test procedure. Also test plans for the determination of NB and WB Ie are to be compiled based on ITU-T P.833, P.833.1 P.834 and P.834.1.

2b)  Preparation of test material, recordings:
To derive the
Ie  parameter, a combination of subjective and objective tests is needed. These tests will require a set of test recordings to be used as input materials for both test types.

2c)   Subjective test:
As ITU-T C-208 shows a discrepancy between subjectively (P.833.1) objectively (P.834.1) derived
Ie  values, Ie  will be derived based on both objective and subjective tests.

2d)  Objective assessment using POLQA (ITU-T P.863)

2e)  Statistical analysis, conclusion and derivation of one single Ie  value per bandwidth mode, if possible.

 

For more details, see our Terms of Reference

Why we do it

The required knowledge about the Ie -Factors for transmission planning using the New ETSI Speech Codec (LC3plus) will be available, and the codec will be established internationally outside ETSI. This has to be seen in the context that for the OPUS codec Ie - values are already known and can be used for network planning purposes.

The LC3plus codec technology is also contributed to the SIG Bluetooth providing a significant improvement to previous technologies. In order to provide the technical benefits also to groups outside ETSI this TTF activity is also performed.

How we do it

The TTF experts are working remotely from their locations (Germany, Czech Republic). The subjective tests will be performed in MESAQIN.com listening laboratory premises in Prague using native English speakers.  

Deliverables

The following detailed information will be added to TS 103 624:

·       Objective assessment of the speech data bases used for the subjective tests using POLQA (ITU-T P.863): new Annex to TS 103 624

·       Equipment Impairment Factors for the different bandwidth modes of the codec: new Annex to TS 103 624

After approval by ETSI STQ, the revised version(s) of TS 103 624 will be available to the general public. .

Time plan

STQ approval is scheduled for end of February 2024.

How to contact us

If additional information is needed, please send an email to consultant@joachimpomy.de

This information is based upon STF/TTF working assumptions.

The views expressed do not necessarily represent the position of ETSI in the context.