What does HD voice mean? These days, it depends on whose marketing agenda you want to follow. Fraunhofer is raising the stakes to put its AAC codec front and center as the "new" standard for voice quality, but it should be careful how it does so.

 Within the (admittedly small) community discussing the issue back in 1999, the technical baseline for HD voice was set at the characteristics for G.722 -- roughly double the Hertz (Hz) of a 3.4 KHz narrowband phone call and providing enough audio information to make a significant difference over narrowband.

AMR-WB  was established as the next generation voice codec for the GSM world and was dubbed  "HD voice." The GSMA, with prompting and participation from AMR-WB stakeholders, established a branding and logo program for "HD voice" to be used by service providers and handset manufacturers. Service providers wanted to raise consumer awareness with HD voice and make sure there was a minimum standard for handsets and capabilities.

An HD voice logo for use in the GSM world was released in October 2011 along with an "Annex C" technical document describing the requirements to be met by mobile networks and handsets to use the logo. Meet the specifications and you can use the GSMA trademarked logo in your advertising.

To cover Sprint and Qualcomm's implementation of HD voice for CDMA2000 networks, GSMA issued an "Annex D" document. Sprint currently is the only documented cellular carrier using Qualcomm's 1X Advanced technology and EVRC-NW codec as covered under Annex C. 

Both Sprint and T-Mobile are correct in saying they adhere to GSMA's HD voice standards (plural). Can the two standards seamlessly talk to each other? No. An HD voice call between GSM and 1X advanced networks would have to be transcoded (translated) in order to get more-or-less the same quality.

To further muddle the picture of what HD voice "is" these days, GSMA and the DECT Forum reached agreement in 2013 to license the HD voice logo to be used with CAT-iq 2.0 handsets. The CAT-iq 2.0 definition uses the G.722 codec  as its version of HD voice.

Technically speaking, this means the GSMA-trademarked HD voice logo can be used in three different types of non-compatible HD voice systems -- cellular AMR-WB, cellular Qualcomm 1X Advanced, and broadband G.722 CAT-iq 2.0. 

Enter Fraunhofer and its multi-year efforts to change the debate on what "HD voice" means.  HP Baumeister, a director at Fraunhofer and longtime participant on HD voice and voice quality panels at IT EXPO, stirred up things at Mobile World Congress this past week by putting the AAC-ELD codec front and center along with 3GPP efforts to create an Enhanced Voice Services (EVS) codec standard for LTE networks.

AAC-ELD already delivers a superior voice experience through Apple's FaceTime services, but hasn't been put into a carrier LTE standard. EVS should deliver much better voice delivery than AMR-WB due to a combination of using better sampling and more bandwidth when available.

Baumerister is correct in asserting that AAC-ELD and EVS will give a better voice experience than AMR-WB.  Neither new codec has been codified into a standards document standard by GSMA for use with the HD voice logo. 

HP, if you're really serious about this, stop all the spin and tell me when AAC-ELD will be approved for use in a VoLTE network, followed by a GSMA annex document.  Maybe then we can talk about shaving around the definition of HD voice.