DRM+ ready by 2007 - but the BBC are trying to delay it

Not content with providing low audio quality on DAB, the BBC are now trying to delay the introduction of DRM+. DRM currently only operates at frequencies up to 30 MHz, and work is underway to extend the system to work at frequencies up to 120 MHz. The updated system will then be called DRM+.

In a moment of honesty at a BBC R&D open day recently, one of the engineers told someone I know that the BBC is trying to delay the introduction of DRM+ "to allow DAB to have a clear run", i.e. so that DRM+ doesn't compete as a digital radio system with DAB.

The reason they want to do this is obvious: DRM+ will wipe the floor with DAB. DRM+ will be able to carry a CD-quality radio station (using 128 kbps AAC) in a 50 kHz channel bandwidth. To provide the same level of audio quality DAB would need to use 256 kbps because it uses the ancient MP2 audio codec, and you can only fit 4 x 256 kbps in a 1.75 MHz DAB channel, so DAB requires 438 kHz to carry a CD-quality radio station -- almost 9 times as much spectrum as DRM+ requires!!

The fact that the BBC is trying to delay the introduction of DRM+ is backed up by this figure in a BBC presentation on DRM+ (2.3 MB) last week at the IBC (International Broadcasting Convention) in Amsterdam:

The figure shows that DRM+ standardardisation should be completed by the end of 2006, and yet it then shows 2 years of DRM+ pilots and chip / receiver development before DRM+ receivers will be in the shops. This 2-year delay is a gross overestimation, for the following reasons:

1. Almost all current DRM receivers use software-defined radio receivers, which means that apart from some RF (radio frequency) hardware at the front of the receiver, the rest of the receiver is implemented in software. For example, Radioscape's new DRM/DAB module is a software-defined radio receiver, so modifying this module's software to incorporate the changes for DRM+ would be simple, and the changes could be implemented as soon as the draft system description is ready. Moreover, all of Radioscapes DAB, DMB, DRM etc modules are 'pin-compatible', which means that existing receivers don't even need to be re-designed to include the new modules. Basically, DRM+ receivers could be available from day 1 after the ETSI standardisation has been completed.

2. Because software-defined radio receivers can be continuously modified, DRM+ trials and pilots can start far sooner than 2008.

The presentation includes some interesting points about DRM+ though. For instance, it picks up on something that I identified a couple of months ago, and which I've been an advocate of on here with other systems for a long time (e.g. DMB), which is that "Once conjoined, DRM’s audio coding [will be] available to [the] DAB system", i.e. once we have combined DAB/DRM+ receivers there will be an AAC+ decoder in each receiver, which would allow broadcasters to use this state-of-the-art audio codec instead of the ancient MP2 audio codec. Because virtually all DRM/DAB receivers are software-defined radio receivers then re-directing an AAC+ stream transmitted over the DAB system to the AAC+ decoder is a trivial task.

The presentation also shows a possible scenario about how DRM+ could be introduced into the existing FM band without requiring any additional spectrum:

It is commonly said that the FM band is full, but that is referring to the difficulty of introducing new FM services into the band, because FM services require far higher transmitter powers and significantly wider bandwidths than DRM+ requires. For example, the FM band goes from about 87.5 MHz to 108 MHz. That's 20.5 MHz of spectrum, but how many FM stations can you receive? If you assume a 200 kHz bandwidth per FM station then the theoretical maximum number of stations you could receive is 102 stations. But in reality the number is far less than this to avoid distant radio stations interfering with one another, and this leaves a lot of empty space for DRM+ channels. The UK is quite lucky in this respect, because we seem to have very stringent interference regulations, which limits the number of FM stations that are allowed to broadcast. In comparison, in French cities you can receive about 50 FM stations and in Italy you can receive about 80 FM stations.

With DRM+ only requiring narrow bandwidths of the order of 50 kHz per radio station, this would allow lots of DRM+ channels to be fitted into this "empty space". This is especially important at the moment because according to a recent Ofcom announcement, 31% of existing UK analogue radio stations will never be able to get onto DAB even after the UK acquires 5 extra Band III channels from next year's Regional Radio Conference. The obvious solution is for the UK to use DRM+ for these stations, and the BBC should do all they can to allow the introduction of DRM+ as soon as possible rather than trying to delay its introduction, because an early introduction of DRM+ is in the best interests of the general public, and we own the BBC, or so they keep telling us.

And once FM is finally switched off, then DRM+ should have the full FM band to itself, which is a total of 20.5 MHz. If they use 50 kHz channels then that would mean that there's 410 channel 'slots'. I think that should be enough to be going on with.

The presentation also shows a possible way for multiple services to share a common transmitter:

Let's hope they do provide Radios 1-4 at CD-quality on DRM+, because they're closer to MW-quality than CD-quality on DAB...

This concept of transmitting multiplex stations on one transmitter could be take a lot further than this example, because, for instance, UK DVB-T multiplexes use 8 MHz channel bandwidths, and assuming 50 kHz DRM+ channel bandwidths then a single higher-power transmitter could transmit up to 160 DRM+ radio stations! Obviously you wouldn't be able to transmit 160 DRM+ stations while FM is still transmitting, but it shows what will be possible. The advantage of transmitting a lot of stations using the same transmitter is that the resources would be shared between a large number of radio stations so the transmission costs per radio station should be very low.

Even if DRM+ stations are transmitted individually the cost of transmission is very low because the transmission frequencies are low and the bandwidth is narrow. For example, with all else being equal, comparing a DRM+ station using a 100 MHz transmission frequency and a 50 kHz bandwidth with a 220 MHz DAB multiplex with a 1.75 MHz bandwidth, the reduction in transmitter power for DRM+ is about 22.3 dB, which is a reduction in power by a factor of 170. The BBC DAB multiplex uses a lot of transmitters that use a transmitter power of 10 kW, so a reduction in power by a factor of 170 would mean that a DRM+ station would only need to use a power of 59 W to provide similar coverage. Obviously that's an oversimplified example, but you get the picture that the transmitter powers required for DRM+ stations are very low, and that means that DRM+ stations will be able to use cheap, low power transmitters, so the cost of transmission for a DRM+ station will be very low.

All in all, DRM+ looks like it will become an excellent digital radio system and solving all the problems that plague DAB, such as DAB's gross spectral-inefficiency, lack of spectrum, and incredibly high transmission costs.

Let's see if the BBC and Ofcom manage to screw DRM+ up as well...

Bookmark with:
Slashdot	Digg	del.icio.us	Reddit