Audio Quality in UC - All Ears Matter

Speakerphones have always been a luxury compared with handset phones, but they have a few drawbacks associated with them. Although speakerphones eliminate the need to hold a phone or wear a headset, they often have a negative impact on sound quality. The reasons for this vary - cheap speakers, small and leaky enclosures, mediocre audio design, etc. The assumption has always been that because it's only a phone call, it's only voice and a quality device is unnecessary.

However, even the old Bell phones had good speakers and mics. When I was at Nortel, the phone design team understood how to design a great speakerphone. They used a good quality speaker and ported the speaker, much like high-end stereo speakers that have ports to manage the sound and to emphasize the mid-bass. That's where most human speech falls, especially male speech. The result is that the phones from the end of the last century often sound better than a modern phone when used as a speakerphone. Many modern speakerphones sound "tinny" in comparison.

Today, there are additional challenges when it comes to high-quality audio. More and more of our conversations are conducted using IP, and more often than not, the IP audio path is High Definition (HD) Audio. Virtually all of the new IP-based communications apps like Skype®, Lync®, Spark®, WebEx®, etc. enable HD audio. Whereas the old G711 phone standard used a low sample rate, yielding a narrow frequency range (20-3,200Hz) and low resolution (8 bits), the new technologies are generally based on the CD standards. You remember CDs, don't you - those silvery discs we stored music on before we started streaming? They sample at a rate that allows 20-20,000Hz, the "maximum" range of human hearing, and have a resolution of 16 bits, yielding a dynamic range that is 255 times greater than the old G711 phone standard. Although these audio streams are typically encoded to reduce bandwidth (similar to mp3s for music), the resulting audio quality is transformational. On a call using this technology, you can really hear all of the conversation with the right mics and speakers.

The result is that the transmission technology is no longer what's limiting phone call quality; rather, it's the design of the endpoints themselves. For various reasons, many users have resorted to using earbuds or Bluetooth headsets for their in-office conversations. It's important to realize that a good headset can make a world of difference because it produces better sound and also manages the echo that can impact others on the call. However, when a speakerphone is required, the challenges are greater. The old "bat wing" conference room speakerphones are good for basic audio, but they generally lack either HD on VOIP for new services or a USB input to connect to a PC. Several small-form-factor USB speaker devices are available; these so-called "disc" speakerphones provide sound quality that's somewhat better than a typical laptop speaker but not much. The result is that you often see users joining a conference on Skype®, Jabber®, WebEx®, GoToMeeting® or WebRTC using their PC speakers and microphone. The result is poor sound quality for these users, as well as for the other participants, because the single mic in a typical laptop is poorly suited for a room, and echo is often an issue.

Recently, Revolabs sent me a FLX UC 500 USB speakerphone to try. Although the same device comes in a UC 1000 version with VoIP support and a UC 1500 version with extension mics, all of the versions have the same quad microphone array and advanced speakers so they are equivalent from an audio quality perspective. The first thing I noticed with the UC 500 was that it weighs two pounds and has both a tweeter and a woofer speaker that can be seen clearly through the mesh speaker cover on the top. Being a bit of an audiophile, I am very aware of the challenges in reproducing sound. A speaker designed to move enough air for good bass response generally cannot move quickly enough to reproduce high frequency signals.

Conversely, a speaker capable of high frequencies does not move enough air for reasonable bass. Most audio equipment addresses this conundrum by having two separate speakers - one optimized for high and one for low. An electrical crossover splits the signal to the correct speaker. To make this all work takes scale, as evidenced by the weight of the UC 500. Without even listening to the UC 500, I knew it had something different than devices with a single small speaker.

Over the next few weeks, I switched from my typical Bluetooth or DECT headset to the Revolabs UC 500 for all of my PC-based communications, including Lync/Skype for Business, Skype, WebEx, a number of WebRTC applications, and even a GoToWebinar event. I asked the other participants if they thought I was on a speakerphone and how I sounded. Most of the time, they could not tell I was on a speakerphone, hear any echo, or hear any difference in how I sounded. From my perspective, the sound was much richer than a disc speaker from another major UC device vendor. This isn't surprising because the Revolabs UC 500 is at least six times larger and more expensive. If you want to understand why it takes size to make a speaker sound good, just go visit your local stereo shop and listen to a few speakers. I also tried moving around to test the multiple mics and found that the quality remained high.

One final event convinced me that the Revolabs UC 500 was truly different. I recently ripped my CD collection to my PC using FLAC (Free Lossless Audio Codec) at full bandwidth. I typically play music through a high-end DAC (Digital to Audio Converter) and through a very high-end stereo system or through desktop speakers in my office. While I was testing the Revolabs UC 500, I was doing some editing in my media manager applications (Media Monkey) and happened to play the song "Diamonds on the Soles of Her Shoes" from the remastered Paul Simon album, "Graceland." This is actually a very high resolution 192kHz/24 (196kHz sample rate, 24-bit resolution) recording that is essentially the studio master file (purchased from HD Tracks). I was, quite frankly, astonished with the UC 500's music reproduction quality. Although it was obviously not stereo, the sound from the UC 500 was great - much better than my desktop Altec Lansing PC speakers that have a separate subwoofer and produce really good sound. In fact, I have since used the UC 500 on in my office for playing background music while I work. If the UC 500 can effectively produce music so well, it is clear why it's great as a speakerphone.

We all have heard the rule that only 7 percent of the value of a conversation is in the words spoken. This is often used to justify the value of video. In fact, in the 1971 book, "Silent Messages," Albert Mehrabian indicated that the non-verbal components were body language (55 percent) and tone of voice (38 percent). In other words, for many conversations, the audio reproduction of the tone of voice is almost as important as video. And it takes a high quality device to capture and reproduce tone of voice. In a PKE Consulting White Paper, I argued that for most task-based conversations, video is not required, although audio intonation is an important part of those conversations. Audio intonation can indicate which word is most important in a sentence, as the emphasis changes the potential meaning.

We need to pay attention to audio quality in Unified Communications, not degrade it further by using PC speakers or small speakers when we are in a group conference. The Revolabs UC 500 does just that. In the end, our ears matter - what we hear is as important as what is said. Try listening to a superior speakerphone and you'll agree.

This paper is sponsored by Revolabs.