The past few weeks the blogosphere has been alight with the conjecture that “VoIP is dead.” A great many have weighed in with opinion on the matter, including; Alec Saunders, Andy Abramson, Dan York*, Ken Camp, Jon Arnold, Irwin Lazar, Om Malik, Tom Keating, amongst others. And a little unexpectedly, Jeff Pulver joined in on the conversation.
All of this seems to have culminated in a Calliflower conference call the evening of Monday January 5th. Those in attendance (39 people!) were essentially the most respected folks in the VoIP blogosphere. The debate was interesting but nothing I care to comment on here. The call is available as a podcast. It’s a good listen.
Where I will offer comment is about the nature of Monday’s conference call. It seems to me that when telecom or VoIP luminaries hold such calls the actual quality of the call (and resulting recording) is often sadly disappointing.
I must ask Why? Especially when it’s completely unnecessary.
I’m not picking on Iotum’s Calliflower service, which I have used numerous times and find generally useful. My complaint is about the nature of the PSTN and the manner in which we deal with conference calls and podcasts. Many conference calls and podcasts simply suffer lousy sound quality.
I’ve heard so many podcasts by and for telecom experts where there were basic problems with the call. From QoS issues to hearing music when they put the conference on hold, it seems like the proverbial cobblers children’s shoes.
Even if the technology is working perfectly there are many other potential sources that can cause trouble. Some people may have called in via cell phone, or they’re in a noisy location, or have a poor headset, or a laptop without a headset, they could be using a speakerphone and wander off-mic, or don’t bother to self-mute on the conference bridge. ZipDX has great advice on conference call best practices, something everyone should be aware of.
It may also be that people have been trained to not expect better. One of the as yet undelivered promises of VoIP is better than PSTN call quality. With so much of the VoIP industry focused on cheap replacement of POTS lines there’s been little focus on quality until relatively recently.
Happily, there is progress on this front. In the past year wideband telephony has been gaining greater traction. Companies like Polycom, Cisco, Mitel, Avaya and snom have introduced phones capable of improved audio bandwidth using newer codecs like G.719, G.722.1, G.722.2 and even the more vintage G.722. Cellular carriers are starting to use AMR-WB making possible improved call quality even from mobile phones.
Of course, the PSTN is limited to G.711, which means 3.4 khz of usable audio bandwidth, not unlike AM radio. In contrast, wideband telephony supports at least 7 khz of audio bandwidth, more like FM radio. The difference is startling. You really need to hear it to understand. Like so many things, once you’ve experienced truly better quality it’s hard to go back.
Since the PSTN does not support anything beyond G.711 any call dialed by a traditional phone number is going to be limited to traditional call quality. Alternatively, calls placed by SIP URI may traverse an IP network end-to-end. This provides an opportunity for dramatically improved call quality using wideband capable systems.
For most people Skype is their initial introduction to wideband telephony. Since its introduction Skype has been capable of better than PSTN call quality when network conditions allow. Initially using GIPS iSac codec, Skype can sound remarkably good without using too much data bandwidth. Skype calls to/from regular phone numbers are passed to commercial PSTN gateways and so are negotiated down to one of the normal PSTN codecs. The wideband advantage is lost.
Ever since the VUC call on wideband conferencing back in November (recording here) I’ve really been bugged by lousy quality conferences. Fortunately, my employer has just moved our west coast offices which gave us the chance to deploy more wideband capable phones, including a Polycom IP6000 conference phone. Now much of our inter-office calling is wideband, and people have noticed the improvement.
Even in the residential setting the DECT CATiQ initiative has laid out a plan to deliver consumer oriented DECT cordless phones that are wideband capable. Some of these are also SIP capable. The Siemens S685IP is a great example. Several of my UK based associates have purchased these for their homes. When signed into our OnSIP account we get wideband interoperability with wideband SIP phones in our US division.
Early in the evolution of VoIP the industry struggled to attain “toll quality” calling over the IP network. It’s is an interesting term as it implies “worth paying for.” I think that in 2009 “toll quality” as measured on the traditional PSTN has little value. Toll quality is now merely adequate. We should expect better.
Now lets get back to where we started in reference to the idea that VoIP is dead. In a recent blog post Jeff Pulver cites new interest in wideband telephony as one of the things driving a possible renaissance in VoIP. Shouldn’t those of us who are promoting enhanced communications solutions practice a little more of what we preach? Let’s push beyond PSTN quality in 2009.
*Incidentally, Dan York and Jonanthan Zar’s excellent Blue Box VoIP Security podcast is one show that is always an excellent quality recording. My understanding is that they often use Skype in its production. My hope is that along with VoIP in general, it also isn’t dead yet.