Press releases and lies.
I hate the art of drafting a press release. It involved too much the art of lying. The number of times I’ve seen “world first” announced in press releases that say nothing is countless.
The one that really got me agitated recently is Imagination Technologies and their announcement around WebRTC. Imagination acquired some years back a competitor of a company I worked for that is called HelloSoft, a developer of VoIP client SDKs. Haven’t been involved with either for quite some time.
I want to go over this specific press release paragraph by paragraph and explain what my understanding of this press release is. The quotes are from the press release. The bulleted text is my own interpretation.
Lets start from the title:
Imagination introduces world’s highest quality WebRTC media engine
- This is about a WebRTC media engine, but it isn’t comparable to the one at webrtc.org (at least not from reading the content of the press release)
- There’s no real indication on why it is the world’s highest quality one – how was that measured, if at all – and against what exactly
Imagination Technologies (IMG.L) introduces an innovative WebRTC media engine that promises to deliver a new level of quality to companies building native or browser-based voice and video conferencing services based on WebRTC. Imagination’s HelloSoft WebRTC engine with its unique, proven algorithms provides enhanced voice and video quality for a superior user experience across mobile and tablet platforms.
- This reads like “there’s a new media engine that implements WebRTC”, which is promising – it reduces the dependency on Google somewhat
- It is suitable for native apps, so those who need to work in embedded environments and use WebRTC can certainly go for this one
- HelloSoft is a provider of tools for developers. If they have a WebRTC media engine ready for customers who need iOS or Android under an SLA, then this is a clear win
- That part of suitable for browsers? Total BS. Browsers who embed a WebRTC media engine are the only ones qualifying. There are 4 of them:
- Chrome has its own implementation
- Firefox has one as well
- Apple is probably using parts of Chome’s implementation
- Microsoft… I don’t see them purchasing this one any time soon
- Playing the enhanced voice and video quality tune means they are old-school. WebRTC is good enough for most. It is going to be a tough sell compared to free
Companies bringing WebRTC applications to mobile devices must consider challenges such as power consumption, packet losses in wireless environments, bandwidth availability, echo cancellation and resource/memory availability, among others.
- All true, but Google’s own publicized roadmap for WebRTC places dealing with these issues within 2014.
Krishna Yarlagadda, president, Imagination Inc., says: “Imagination has deep expertise in developing media engines and related technologies, and our HelloSoft V.VoIP and VoLTE clients have already shipped in millions of devices worldwide. With this experience and leadership in building a wide range of leading IP for mobile platforms and other resource constrained devices, we understand the issues faced by our customers in bringing WebRTC to mobile. With our advanced algorithms encapsulated in an off-the-shelf product, we are enabling our customers to dramatically improve the WebRTC user experience.”
- Millions of devices doesn’t seem like much when compared to the 1.6 billion devices referenced just in the next paragraph. Not much of an experience here… although both Imagination and HelloSoft know what it takes to build such tools
- “encapsulated in an off-the-shelf product” – so this isn’t a media engine or a WebRTC replacement after all…
Imagination’s HelloSoft WebRTC media engine incorporates a number of features for better performance, functionality and overall user experience:
- Improved voice quality for speaker phone functionality across platforms with pre-processing modules for acoustic echo cancellation (AEC), noise cancellation (NC) and more
- Superior quality voice and video under lossy wireless environments through Imagination’s DVQM (Dynamic Video Quality Management) and EVQM (Enhanced Voice Quality Management) algorithms that mitigate up to 30% of packet losses
- Optimized for major processor architectures including the MIPS CPU architecture, with support for Android, iOS, Windows and Linux-based operating systems
- Support for a comprehensive list of voice codecs including OPUS, G.722, G.722.1, AMR-WB, AMR-NB, G.729ab, G.723.1, G.711, Speex, EVRC, iLBC; plus video codecs including VP8 and H.264
- Leverage of the hardware video accelerators available on a device, such as PowerVR video codecs found in millions of devices
Imagination’s WebRTC media engine is also available as a complete turn-key client in combination with call management software and SIP stack.
- So we have a media engine with all the bells and whistles we used to have
- A lot of voice and video codes, as you might have expected
- Integrating with Imagination Technologies’ PowerVR by HelloSoft
- It can come with or without a full SIP stack and client on top
Me left unimpressed. Or at least a lot less impressed than the amount of publicity this PR made.
What do we have here?
This is a simple media engine that includes Opus, VP8 and the SRTP changes necessary to support WebRTC.
I’d say it is on par with Spirit-DSP’s offering or M5T.
It probably doesn’t support the WebRTC APIs – just handles the network in the same way.
I’d also assume it has no SDP parser/generator that is suitable for WebRTC built-in to the media engine itself.
This is on par with what GIPS was before getting acquired by Google, especially from the business model and concept angle.
Why is it important?
Such press releases put more confusion than understanding in the industry. It brings nothing new to the table rather than adds a few features to an existing SDK. The end result has the same business model, targeting the same old customers.
It shows that Imagination Technologies weren’t that imaginative (excuse the pun) when they took the route they had with WebRTC. This is rather sad, considering the important role they play in current day smartphone acceleration of codecs.
I believe they could have done more than stick the word WebRTC on their HelloSoft Media Engine, but that’s just me venting.
I am sure HelloSoft will find interested customers. This product, as it seems, is VERY suitable for those looking to add to their existing VoIP clients the ability to communicate with browser-based WebRTC – or for those developing mobile apps use it instead of “porting” WebRTC on their own.
–
A few disclaimers on this one:
- Sorry for being blunt on this one. Had a bad morning
- I haven’t been in contact with Imagination Technologies or Hellosoft in the past 5 years
- The solution in this press release is good for some vendors, but the way it is presented in the press release isn’t targeted at the potential customers, but rather at getting more eyeballs
Feelink kranky like Chris ? 😉
Well… he did just annouced we’re upping our game, so I guess I need to get Krankier as part of that process.
Tsahi,
I completely agree! That press release just looks wrong. It described a thin layer of the new wrapped around a big old wad of the old. It points to a struggle to appear relevant in a brave new world.
Michael
Michael,
This is a problem I see with many of the vendors (and consultants/analysts) around WebRTC and VoIP. They don’t understand the potential of disruption here that goes in all levels – from the development processes, to testing and deployment, and to the business models and offerings that WebRTC enables.
This one just got on my nerves because it is so close to what I did a few years back and left behind BECAUSE of things like WebRTC.
I like to put this kind of news that I see out there for discussion. Most of the “claims” we see and hear even from our clients are usually met by me saying things like “do you remember when x did that?” to which I get “no, who are they?”
I’ve also learned not to dismiss things either. Instead, lets see what they have to say if they reply to this and get the facts and the SDK out to see just what it does or doesn’t do.
Andy,
I definitely don’t dismiss this one – just the press release. If this is half as good as they say it is, they will find a few customers willing to integrate it and happy with it.
What worries me most about this Press release is what it doesn’t talk about. No mention I could see to ICE or STUN – which are the weak points in webRTC at the moment. No mention of use of platform native EC (the echo canceller on the iPad is brilliant, anyone implementing their own is a fool). No mention of DTLS – or which crypto suites are supported (do they support Perfect Forward Secrecy?).
I’m also concerned by [DE]VQM – does it work if the other end isn’t running it? Does this stack play nice if the far end is chrome?
So yes, I agree, an annoying release, but I’d just ignored it along with all the other announcements of vague re-purposings of old tech that need a new home.
What was it that riled you specifically in this one ?
Good question.
Probably my upbringing. I used to develop and market similar technologies when I worked at RADVISION and Hellosoft was a competitor of sorts. I think I learned to read press releases, and this one, just up my alley just seemed impossibly ridiculous – but more than that – it just missed the point and their specific target market as far as I can see.
Tsahi
Given I agreed to be quoted in the release, it probably falls to me to explain why I think it’s more interesting than you suggest.
The main thing in my mind is thinking about who *Imagination’s* main customers are, rather than “legacy” HelloSoft:
Handset vendors.
Imagination sells silicon IPR, especially for multimedia/GPU chips in smartphones & other devices. It also owns MIPS, which makes processor cores. Apple is one of its largest customers.
In other words, services like Apple FaceTime implicitly already use various of Imagination’s products. It’s presumably powering the chip that does the H264 acceleration in iPhone & iPad.
It doesn’t take a huge leap to think that some device vendors might want a combination of graphics chip + media engine suitable for WebRTC or WebRTC-like applications. With, say, H265 or VP9 hardware acceleration.
You are assuming that HelloSoft’s main targets with this announcement may be the wider voice/video developer community. I suspect that’s partly true, but I suspect the real strategy might be more about signing up more handset OEMs and their *internal* development teams.
Consider if (for example, I have no inside information), HTC or Huawei wanted to create its own FaceTime-equivalent. This could be a pre-integrated solution “down to the silicon”. Or perhaps a consumer electronics or medical device vendor. They may also be in a position to create custom/optimised browsers as well.
Silicon IPR companies don’t really care about supporting random conferencing or VoIP developers. They care about selling more processor designs, for which having additional functionality for “native” WebRTC may well prove to be a benefit.
My analysis, anyway – yours may vary.
Dean,
This is where our opinions diverge.
I’ve been working for many years with chipset vendors and OEMs on voice and video solutions. I also been in the business of selling the same voice and video solution to application developers in various areas. Been working closely with the likes of TI, Qualcomm, HTC, Samsung, LG, Asus, etc. I know this industry. I know the way of thinking in it rather well – their needs, their constraints.
I also know what HelloSoft was selling at the time, and what Imagination as a company is all about.
Somehow, when you bake this all into the PR – the end result is not tasty enough for me.
If this was a PR coming from Qualcomm, Broadcom or other chipset vendor that can do RF and application chip, package it all and provide the base OS on top with all relevant drivers – then it is a different ball game.
Coming from the likes of HelloSoft and Imagination, I don’t see the huge value.
I have been wrong before, if that is a comfort 🙂
Hi,
Thank you for sharing your opinion on the matter but obviously we stand by our press release.
We’ll be at WebRTC Expo for anyone who would like to come and see demonstrations of our technologies – after all, we know that seeing is believing.
Best regards,
Alex.
Alexandru,
Thanks for the response. Sorry this one turned out this way. I myself probably won’t be at WebRTC Expo this time – got an interesting announcement to make later this week about my own plans.
I do hope to meet you in one of the future events out there on WebRTC, and I truly hope this initiative of yours with WebRTC turns out for the best. There are definitely some things I would have done differently in your place – both on the technical side and the business/marketing side.