Hoping that the open-source Android would be the one to adopt standards-based video calling? We weren’t holding our breath either…
Google has quietly made the statement that they do not appear willing to support the current industry standard for phone-to-phone video calling inside of Android.
UMTS (and LTE) cell phones are encouraged to comply with the optional 3GPP standard for video phone calling. The standard is part of the 3GPP specification and utilizes H.263 for phone to phone calling.
Support for the standard has been sparse, but always present since the launch of UMTS cellular phones. It is also still present in LTE as part of LTE’s backwards compatibility. LTE did not, however, propose a succeeding standard.
The first phones in the United States that supported UMTS video calling were the original UMTS phones that landed in the USA; the Nokia 6651 and Motorola A845 on AT&T Wireless. While AT&T’s network was effectively a regional test launch – it would not launch nationwide until years later on Cingular’s consolidated network – it did work consistently for almost four years on all capable devices.
Some phone makers have gone to great lengths to support the UMTS video calling standard. Samsung wrote custom code for Windows Phone devices to support the standard and it also was integrated into select Android phones in Europe through custom code too.
However, Apple has avoided the standard in iOS, opting instead for FaceTime. While Steve Jobs himself promised to open up the FaceTime standard, Apple never has, and has refused to comment on the commitment by Apple’s founder on every occasion since. Microsoft has since purchased Skype and integrated it into all Windows products.
And that leaves Google. Google has struggled with its own video calling standard, first using Jabber atop Google Talk – and then closing the platform by forking Jabber to create the closed-source Google Hangouts platform.
Where this leaves the industry is with three video calling standards that are completely private, and completely incompatible. So it was no surprise when Google closed the feature request to add 3GPP video calling as “obsolete.” It’s also sad that the industry can’t cooperate on standards anymore.
This adds insult to injury because H.263 is mandatory for Android-compatibility. Google mandates devices support H.263 but appears uninterested in closing the loop and allowing UMTS/LTE Android phones to communicate with non-Android UMTS/LTE phones that support H.263 too.
When will this change? Not any time soon, likely. If Google’s VP8 codec (now WebM) survives court patent challenges, that may emerge as an open calling standard for video codecs. WebRTC remains mired in patent land, now proposing to please nobody by requiring both H.264 and WebM on browser-based clients (IoT devices could then use either protocol – but this doesn’t explain how one IoT client with WebM and one IoT client with H.264 could communicate together).
It likely will not be until H.264’s patents expire in 2027 that the industry has a truly patent free choice to replace 3GPP video calling standards. WebRTC’s future remains questionable, and there already is a functional standard in place dating back to devices sold and in-use for over a decade. Supporting 3GPP and the H.263 codec remains the best option for standards-based calling on UMTS/LTE smartphones.
I can sympathize with Google, though I agree they should not have closed the bug/feature request as obsolete.
Google is trying to replace 3GPP with WebRTC in order to welcome in VoIP, tablets, etc that do not have phone numbers… 3GPP’s video calling requires an MDN (phone number).
It is (more than) a bit pot-calling-kettle-black for Google to not support 3GPP, but encourage everyone to support WebRTC… when they themselves won’t go the open route on Hangouts. Has Google even promised that Hangouts would become WebRTC should the standard get approved?
I would hope that Google would at least offer to allow others to write 3GPP video calling support for Android, and agree to accept it into AOSP should someone take up the cause. I can’t see anyone undertaking that (considerable) effort unless Google was willing to agree to incorporate it beforehand, however.
I find this article pretty baffling. H.263 was pointless in 2004, and 10 years still haven’t done anything for its relevance.
In fact 3GPP video “calling” is one of those areas which probably shouldn’t have been standardised in the first place. As usual with the mobile industry bodies, nobody bothered to think about design and end-user behavioural considerations first, or the use-cases. They simply assumed that a video call would be like a phone call, would use the same model of interruptive call-flows, the same number, the same dialer experience, any-to-any connectivity etc. And that’s without thinking about the pricing and business-model aspects. Most of those have no basis for assumption.
WebRTC allows for many different models of both voice and video interaction, as can be seen with the huge array of applications and services that have already appeared, including those baking the main protocols into mobile apps. (Hangouts is already supporting WebRTC, by the way).
There are far fewer questions over WebRTC’s future, given the broad ecosystem of developers involved. H.263 and even the proposed IMS ViLTE are highly unlikely to provide viable alternatives to other video apps/services that are well-tailored to particular use-cases. There is no need for a lowest-common denominator, unlike voice telephony.
I disagree that the article is “baffling” – Google should weigh in here, this standard is in use today quite broadly in Europe on millions of phones. Just because carriers over here have conspired to limit its use (which I suspect is so they can cut costs and offer their own strategic platforms) doesn’t contest the standards use today.
It would be great if WebRTC or IMS sort out their differences and standardize. But this case is about the here and now.
Frankly AOSP could use 3GPP as a starting point to implement WebRTC in the dialer app and other services, thus allowing Androids to auto negotiate the vest protocol based on the carrier and platform choices loaded into the device. 3GPP/H.263 would then be the fallback if WebRTC or IMS aren’t available.
Meant to say “best protocol” not “vest protocol”. Sorry, typing on an Android. At least I’m using the device I’d like to see 3GPP (and WebRTC) used on.
Used in Europe on millions of phones?! You jest.
I see nobody, ever, using it, neither at home in the UK, nor when I travel anywhere else. Maybe there’s a few hundred ancient cellular CCTV cameras or similar, but there’s simply no evidence of its widespread use – I’m not even sure if any networks still support it. I challenge you to name a single individual that is an active user.
WebRTC is almost nothing to do with IMS. I expect that at least 90% of WebRTC use-cases will never touch IMS, nor ever need to interoperate with any public phone or video services. For the handful of IMS/WebRTC uses such as VoLTE extension, there are already numerous gateways and even a 3GPP standard.
The whole point is that “video calling” is a (poor) solution in search of a non-existent problem. Video communications exists exclusively within other broader applications, or in well-designed standalone apps like FaceTime.
3GPP video-calling is like its modern-day equivalents RCS and ViLTE – pointless technologies dreamed up by engineers & standards bureacrats, not created by designers and developers who understand what users actually want.
WebRTC *is* here and now. I have three video-capable WebRTC apps on my phone (appear.in, WeCam & SnapChat) and two voice-only ones (Talko & Tuenti)
Based on my experience with UMTS video calling when I lived in the UK for a couple of years in the early 2000’s, what killed its chances in the mainstream was the fact that carriers had the now idiotic idea of charging for video calls on a per-minute basis, when the standard was developed so that flat-rate usage would have been the best way to promote and sell the capability.
I recall rates as high as 4 pounds 50/min for a video call on Vodafone and even higher rates if another caller was in another country on the same network. It never had a chance of taking off and by the time Hutchison’s 3 network launched a few years later, it was too late to pitch UMTS video calling as a compelling feature anywhere.
Standards matter. WebRTC doesn’t even have final how WebM and H.264 will interop in some scenarios. I agree WebRTC is great, but jest I do not.
AOSP could easily support both platforms. If Google doesn’t want to do the work, fine. But they should accept commits if offered up for joint 3GPP and dialer integration. Or at the least, commit to a path for WebRTC natively in the AOSP stack. Note I said AOSP, not Hangouts/PlayStore.
After talking with some sources in the industry, AT&T and T-Mobile basically are unwilling to implement the server-side support even if Android supports the standard.
Of course, that might change under pressure from users, if the standard got implemented… thus becoming a painful Catch-22.
While Verizon could probably opt-in via IMS on VoLTE phones, neither Sprint nor Verizon could do it across-the-board. That leaves the US without any carrier support, and without that, it would be hard to accomplish much traction-wise.
I do sympathize with the sentiment, however. It shouldn’t be this way, but AOSP is only one piece of the puzzle.