BARCELONA, Spain--What wireless carriers have been marketing as "HD Voice" may not actually be truly high-definition voice quality, according to an executive heavily involved in the development of the industry's audio codecs. However, carriers that have been using the term HD Voice to describe enhanced call quality stand by their use of it, arguing it has been endorsed by the GSMA.
The launch of what carriers call HD Voice around the world has started to take off. In the United States, Sprint (NYSE:S) and T-Mobile US (NYSE:TMUS) are its two largest proponents. However, HP Baumeister, who is director of mobile and communications markets, audio and multimedia for Fraunhofer, said that the term "HD Voice" currently in vogue for carriers is more a matter of marketing than technical voice quality.
For Baumeister, it all comes down to the codecs carriers are using. HD Voice uses Adaptive Multi Rate Wideband, or AMR-Wideband, technology to improve call quality, but the voice codec must be supported in handset hardware along with two microphones and noise-canceling software. In order for HD Voice calls to go through, both the handset making the call and the network infrastructure supporting that call must support HD Voice, of which there are several variants.
More importantly, the AMR-Wideband codec, while a large improvement over legacy voice telephony, isn't of such high quality compared to another codec that is likely going to be commercialized next year, or to a codec used for over-the-top voice and video applications, such as Apple's (NASDAQ:AAPL) FaceTime.
"HD Voice, it's a stretch of marketing," Baumeister said. He said most carriers know this, but that the general public does not.
"They [carriers] set themselves up for trouble immediately after they made the decision" to adopt AMR-Wideband. That's because there is another codec called AAC-ELD that is of higher quality for over-the-top applications. And, there is another coded for telephony voice being developed in the 3GPP called Enhanced Voice Service, or EVS.
Fraunhofer is working to standardize the EVS codec along with a technical working group of industry heavyweights that includes Ericsson (NASDAQ:ERIC), Qualcomm (NASDAQ:QCOM), Samsung Electronics, ZTE, Orange and NTT DoCoMo. The EVS codec should be completed in September and in commercial devices next year.
According to Baumeister, the AMR-Wideband codec runs on 7 kHz of bandwidth, which is an improvement from the legacy systems that used 3.5 kHz. However, EVS will run at least 14 kHz, and in a demonstration produced here by Fraunhofer at the Mobile World Congress trade show, a call using the EVS codec produced a much higher audio quality than an "HD Voice" call using AMR-Wideband.
In order for EVS to be commercialized, it must be implemented on chipsets, devices and in the network. It will likely first be deployed in mature LTE markets, such as the U.S., Japan, South Korea and parts of Europe, Baumiester said. The EVS codec also supports Voice over LTE. The most logical candidates to use it are carriers that have not yet embraced AMR-Wideband, he said.
Baumeister said he gives operators the benefit of the doubt because they had few options other than AMR-Wideband to improve voice quality. However, he said that because codecs on the market today, such as AAC-ELD (which also works for OTT apps on Android), already produce higher-quality audio than AMR-Wideband's "HD Voice," carriers who use it are setting themselves up for criticism.
"If you want to continue to monetize voice, you have to have the best voice, not the worst," he said.
The carriers defended their use of both the technology and the term "HD Voice."
"We stand behind our use of the term HD Voice, as it is an approved standard set by" the GSMA, Sprint spokeswoman Kelly Schlageter told FierceWireless. She said Sprint had no comment on any plans to use EVS.
On Sprint's fourth-quarter earnings conference call earlier this month, Steve Elfman, president of network operations at Sprint, said, according to a Seeking Alpha transcript: "Another important feature for our customers is the deployment of HD voice, which we expect to be complete across our nationwide network by midyear. We currently have over 12 million customers in our base today with HD voice capable handsets that will experience the enhanced call quality this year."
T-Mobile also declined to comment on its future plans for EVS, and it echoed Sprint in saying that HD Voice is an industry standard defined by the GSMA. "With HD Voice, T-Mobile customers with capable smartphones automatically hear fuller, more natural-sounding voice quality with significantly reduced background noise when calling another capable smartphone on T-Mobile's nationwide 4G LTE network," the carrier said in a statement to FierceWireless.
"T-Mobile launched HD Voice in 2013 marking a major step forward in dramatically enhancing in-call audio performance on many of the latest smartphones across its 4G LTE network," the carrier said. "Our HD Voice is a set of features on the network and the handset including enhanced audio processing, multiple microphones, speakers and improved echo cancellation. While codecs do play a part in supporting HD Voice, T-Mobile also works with OEMs to provide specific hardware components that are a key ingredient in the HD Voice experience."
For more:
- see this AMR-WB page from 3GPP
- see this EVS page from 3GPP
Related Articles:
AT&T admits to VoLTE delay, won't offer new launch date
GSA: HD Voice starting to gain traction worldwide
HD Voice: AT&T, Sprint promise it this year, but Verizon targets late 2013, early 2014
T-Mobile turns to YouTube to promote HD Voice
Sprint builds on HTC Evo success with LTE version featuring HD voice