It appears to be that it isn't that a fully digital system that has failed, but that since the project was first started there are now many off the shelf systems available that do the same thing but are far cheaper.
From what I hear, the problem was more like entering to run a marathon before you could walk. I don't think many people disagree that in principle the "tapeless" concept is the way of the future - but as with many things, it can often be extremely foolish to try to force "the future" too soon. That's especially true the more complex the project - and DMI was not just a production system, it was (supposed to be!) a complete end to end system from production, through post, playout and to and from the archive. It's worth reading what the Guardian has to say about it - http://www.guardian.co.uk/media/2013/may/24/bbc-digital-media-initiative-failure . In particular, ""This is because much of the software and hardware which has been developed could only be used by the BBC if the project were completed, a course of action which, due to technological difficulties and changes to business needs, would be, I fear, equivalent to throwing good money after bad."
Interesting article but I was intrigued by the 'reader comments' at the bottom, the second one in particular.
The second comment is really hitting the nail on the head in some respects, but is a bit out in other ways. The whole "tapeless" terminology was a bit of a red herring - "file based" may have been better wording - so basically the idea is that instead of dealing with video signals, you're dealing with video files. Accept that, and LTO tape for backing up makes complete sense. It's just not intended for the viewing purposes the author of the comment thinks. Secondly, it's important to remember that a big organisation has advantages over a very small one (economies of scale etc) - but also has complimentary disadvantages (far more difficult to structure any change). And if you change one significant aspect - you have to change everything else to go with it. By the sound of things, that's what went wrong. The project tried to be all encompassing, that meant unique systems to the BBC, so if details of it go wrong the whole workflow falls apart. Bearing in mind the fate of other very large IT based projects (NHS, anybody?) it shouldn't really be any surprise. Sometimes extremely clever IT people can show a remarkable lack of appreciation of what happens in the real world.
See - http://www.broadcastnow.co.uk/has-the-tapeless-future-arrived/158279.article . The whole idea of "tapeless" and Starwinder was well under way in May 2006 - which begs the question of how much money the BBC spent before 2008?
The BBC, which estimates that it spends £8.5m a year transferring footage to and from tape, is at the forefront of the move towards tapeless and has three major projects involving it underway: Starwinder, Gateway and the Creative Desktop.
As with all projects of this nature the BBC involves itself in, controversy is never too far away. Starwinder, which according to BBC head of technology production Paul Cheesbrough, has a rolling brief to look at the spend on broadcast equipment and the transition to digital, is currently focusing on camera technology and has settled on the P2 and Grass Valley's new Infinity camera as its preferred acquisition formats.
Secondly, don't attempt to build vertically integrated systems if you're a broadcaster. Just make sure that your equipment and systems use standard interfaces and protocols and buy the "best of breed" at each layer
I think that given the BBC's reputation for R&D in the past they may have thought they could do something similar. I'm not sure about now but the BBC were always something/someone the rest of the industry looked at with some envy given some of their technological developments.
Yes, but as well as some very good work they had some pretty spectacular lemons as well. The difference being that none of them were on the scale, or as far reaching as Starwinder/DMI, and/or were at a time when there was less direct accountability. (Though past direct accountability may have stopped some successes by too much risk aversion!)
As examples, VERA? Technologically brilliant, but at what cost - and rendered completely obsolete by Ampex before it was operational. DAB, even? Again, brilliant technology, but ahead of it's time. And being deployed with a non-upgradeable codec that was obsolete before the first set made market probably has put digital radio back decades. It really needed somebody to have the guts to kill it (yes, rendering existing DAB sets useless) and relaunch digital radio with a far more efficient compression system to make any idea of analogue radio switch off viable.
DAB, even? Again, brilliant technology, but ahead of it's time.
Often, the brightest technical minds can be very blinkered. Be far too much into technology for it's own sake - not as serving the end user.
.... and here's an example
I remember well looking round at some of the first DAB receiver offerings only to be a bit disappointed when I read that there was likely to be a different system because the one launched was limited. Further reading led to the limitations of the codec and the fact there was no upgrade route. Needless to say I'm still with FM!
And you are far from alone! The basic engineering behind DAB was brilliant, and I was lucky enough to see a demonstration at a very early stage. But my memory is that one of the main perceived advantages at the time was a lot down to in-car use. Because two transmitters can be set up adjacent to one another on the same frequency, and only serve to add constructively, it meant that any network across the entire country would only need one frequency, and a listener in a car driving the breadth of the British Isles would be seamlessly passed from one transmitter to another with no retuning necessary. And all the time with perfect reception - no fading etc - due to the characteristics of the DAB signal. The other point given in it's favour was quality, digital techniques to give quality surpassing FM.
Well, that was the theory. Again, with hindsight it was put into service too soon. And the decision was taken to market it primarily on the basis of extra channels. With limited bandwidth available in the first place, that meant the channels had to be far more compressed than initially expected, and guess what? Quality suffered. So dramatically, that complaints led to the BBC being embarrassingly forced by Advertising Standards to withdraw a major advertising campaign on the grounds of misrepresentation. (Remember "1's and 0's" adverts about 15 years ago? Ever wondered why they stopped suddenly.....?) Worse, for a long time the transmitters broadcasting DAB were very limited, which made it's mobile and in-car features pointless. Worse still, the limited deployment gave it a bad name with early adopters who found themselves either getting a perfect reception - or total silence. Which was generally perceived as more disconcerting than the gradual fading and deterioration of the in-car FM signal.
As for the extra channels, I don't know how successful they are now, but for years the number of listeners were such that they were below the measuring scale, and even now they heavily rely on people listening by other means than DAB - online, Freeview, etc. Finally, it also didn't take early adopters long to find out that DAB sets ate batteries like there was no tomorrow, and effectively meant that "the radio" got tethered back to a mains point - back to pre-transistor days!
I'm not knocking digital radio per se - it was the "too much too soon" facets of the rollout that were wrong. For a very long time the number of listeners was very small - so little overall benefit there - but it saddled the UK with an outdated legacy problem by the time when technology had matured enough to make it worthwhile. That's in particular down to the codec used for compression, extremely inefficient by todays standards. But move to a modern system and all existing receivers get obsoleted overnight. Worth a read is http://www.telegraph.co.uk/news/9657238/Radio-switchover-delayed-as-listeners-shun-digital.html but I think there may be a lot of spin going on even there. It says
Digital sets were expected to account for half of all radio listening by 2013 but the figure stands at just 31.3 per cent.
I suspect that the 31.3% figure may not actually represent 31.3% of all radio listening, but rather the number of households with at least one DAB set. Which is not the same thing. We have a DAB set in the kitchen - but due to reception problems on DAB, it now spends it's entire life tuned to FM.
In the Telegraph article, the head of the industry body says (rather patronisingly) “A lot of people still don’t get it – they don’t understand what digital radio offers and why it’s important,” I think a lot of people get the importance only too well. Turn off analogue and they get money from selling off the spectrum. Or is that just too cynical?
Lots of other information (and teccy stuff about why DAB+ is so superior to DAB) at http://en.wikipedia.org/wiki/Digital_Audio_Broadcasting#DAB.2B