CUTTING EDGE

Tiny portable hardware sequencers have been available for several years, but the clever Crusoe chip is now making possible incredibly portable software studios, running on pocket-sized PCs.

I'm writing this on a computer that's running a word processor (obviously!) and a music application. While I'm typing, I've got software emulations of the following devices hammering away at the processor: a digital mixer with fader automation, four samplers with 16-note polyphony, two polyphonic 'analogue' synths, three sample-loop players, a drum machine, a digital patchbay, a compressor, two delay lines, a phaser, an envelope-controlled filter, and a reverb unit. Not to mention a MIDI sequencer. The software is, of course, Propellerhead's Reason.

You might be forgiven for thinking that this computer is not a PC or a Mac. But it is indeed a PC, and, with no apparent effort, it's running Windows 2000 Professional. The processor is one I mentioned over a year ago in Cutting Edge, when it was first announced — a Transmeta Crusoe. When I first came across descriptions of the Crusoe, it caught my attention because if it was ever built it would represent an important development in processor design, that could have significant implications for the way studios could work in the future. Here's why.

As we've already established, the Transmeta device is not a device from the Intel family of chips that has powered PCs since IBM (unintentionally) established the PC standard a couple of decades ago. It's not an Intel clone, either. It has an architecture that is completely novel and unlike anything that's ever been soldered to a PC motherboard before. It might as well have come from Mars. So how is it, then, that it can quite happily run a Microsoft operating system?

Well, the remarkable fact is that you could probably run any operating system that works with Intel chips on the Crusoe. Linux, for example. I suspect you could even have a go at running BeOS (sadly ironic in the light of the company's recent history). That's because Transmeta have written a special 'hardware translation' layer of software that runs on their chip. It sits between the chip and the operating system and makes the Crusoe look like a conventional Intel device. So Windows 2000 doesn't know it's not running on a Pentium. Indeed, the Windows 2000 system information tool reports the presence of an x86 processor (which is the general shorthand for an Intel). I've never had it confirmed, but it probably follows from this that Transmeta could write alternative hardware translation software to make the Crusoe look like a G4. It would be pretty cool, wouldn't it, to have a computer that you could dual-boot as a Mac or a PC? But I won't go there, because this is speculation based on supposition.

To sum up, then: this device sports a processor that uses a software emulation layer to appear to be a different kind of processor than it actually is. It quite happily runs a software package that can itself behave, in real time, as if it is a complete recording studio. (I should point out that, as far as I can tell, the Transmeta chip runs software more slowly than the Pentium device it is emulating. There were a few glitches in the more complex Reason projects.)

This 'software emulation of hardware' technique is significant because it's the first time I can recall ever seeing studio-quality audio processes running on hardware that the software writers may have had no conception of. It means that in the future you could create, let's say, a reverb setting and run it on anything that can support the application, and it will, without qualification, sound exactly the same.

The computer concerned is a Sony Vaio PCG-C1VFK, and it's by far the nicest gadget I've ever seen!

Digital Radio Mondiale is an exciting idea that is the first suggestion I've come across for rejuvenating the part of the radio spectrum that we call the AM bands: Long Wave, Medium Wave and Short Wave. These were the frequencies on which we used to listen to radio, before, over the last 20 years or so, FM became our default waveband. I love radio and think it's still the best way to maximise the amount of music in our busy lives.

At the risk of sounding ever so slightly nerdish, I need to introduce some explanations at this point.

AM stands for Amplitude Modulation, a simple way of getting around the fact that radio waves at audio frequencies do not travel very far. It's a good job that they don't, because if they did, every amplifier and loudspeaker voice-coil would be broadcasting to the world, which would make reception a nightmare. Radio waves only travel any distance at frequencies well above the ones we can hear (although you don't have to go too much above the audio spectrum before they can travel truly huge distances: Radio 4 goes out at a frequency that is only 10 times that of the highest sound we can hear, and only twice the frequency of the standard for DVD Audio sample rates).

A tiny Sony Vaio PC, with the Transmeta Crusoe chip at its heart, running the self-contained Reason software studio.

FM sounds pretty good. To most people it sounds nearly as good as a CD, but in fact its frequency response is limited to 15kHz and it is prone to hiss (especially in stereo) when there is less-than-ideal reception. But the biggest problem with FM is that it doesn't go very far. In fact, reception is almost limited to 'line-of-sight'. It's only by putting the transmitters on top of very tall masts, and having pretty powerful transmitters, that any sort of useful range is achievable. Absolutely the worst possible circumstances for FM reception is when the receiver is moving — and guess what? Most of us listen to FM when we are in our cars. Brilliant.

There is an answer, of sorts, to this: Digital Audio Broadcasting, or DAB. Designed to take over from FM for high-quality audio broadcasting, it's available now. But I think it's safe to say that I'm as likely to win the lottery as I am to meet a reader of this column who has a DAB receiver. Which is ironic, because the only way to afford one is to win the lottery.

Suffice to say that DAB hasn't reached critical mass yet, which is a pity, in some ways, because DAB certainly addresses the biggest problem of mobile reception: multi-path interference. When you listen to an FM station in your car, you are likely to drive through the 'standing wave' patterns created as multiple reflections of the original signal add and subtract. Neither the transmitters nor the buildings reflecting their signal move, but if you do, you will hear regular swooshing noises as you pass through the peaks and troughs. Not ideal as you listen to the Adagietto from Mahler's Fifth Symphony.

DAB reduces the fragility of the information in the radio signal by using multiple carrier waves (up to an incredible 1536 of them), at 1kHz intervals. Spreading the information across all these carriers makes it much easier to reconstruct the original signal. In fact, it has even been claimed that multi-path reflection actually reinforces the signal — like reflected light in a room. The BBC's snappy title for this technique is Coded Orthogonal Frequency Division Multiplexing, or COFDM, to make it only slightly shorter.

For general radio reception, DAB sounds like a good bet. (I do have reservations about quality, which I will talk about in more detail in a future article.) Still, for all its cleverness, DAB is anything but a worldwide standard. A bit like NICAM, the technique we use for stereo television, it only works in the UK and a few other places — and definitely not in America.

Digital Radio Mondiale, on the other hand, is designed from the outset to work anywhere on the planet. The original document setting up the technical specification for DRM states that "a DRM digital receiver, operating below 30MHz [that encompasses the long, medium and short wavebands], bought anywhere in the world, should operate anywhere (where the frequency bands are available)."

I was at one of the first ever public presentations of DRM, at the International Broadcasting Convention in Amsterdam. The technology is at the stage where you need a wardrobe full of electronics to make up a working radio, but don't worry about this. All consumer technologies go through this stage and rapidly become smaller in size and cost.

DRM uses some incredibly clever and complex compression techniques — including one where the highest frequencies are not fully encoded but are 'hinted' at, and are regenerated synthetically at the receiving end. This won't please audio purists, but don't forget that the whole idea is to make AM reception sound better than before.

In practice, the results are pretty good. The system allows the broadcaster to choose between audio quality and the robustness of the transmission (its ability to withstand interference and adverse signal conditions). I don't know where the sample I listened to was on this scale. It sounded to me rather like a commercial FM chart-hits station, with masses of (dynamic range) compression, and with the kind of unmusical grittiness you expect from MP3 files at too low a bit rate. And it was in mono, although the system does allow for stereo transmission where appropriate.

Not perfect, then, but a huge improvement on anything I've ever heard on the AM bands — and especially impressive when you consider that the system can use existing transmitters and aerials. I look forward to listening to Radio Moscow in stereo on my hi-fi. The DRM roadmap tells me that this may be possible by 2003.

Bluetooth: Will It Happen After All?     The little Sony PC discussed earlier in this column has one more trick up its sleeve: Bluetooth. We've been talking about Bluetooth for so long now that "it will never happen" seems almost to be part of its definition. This wireless 'cable replacement' protocol is so cheap and so economical with power that it's set to become a standard feature of computers, hand-held organisers, mobile phones, and anything else that needs to move data between devices. I'm rather hoping that we'll see it on some electronic music and studio devices before long, as it easily outpaces MIDI for capacity and would work nicely to connect remote controllers to rackmount units. But it's here and now in the Sony PCG-C1VFK. Bluetooth is supposed to be completely self-configuring. Devices in a Bluetooth transaction would ideally discuss with each other what kind of facilities they offer and devise a way to collaborate. You can imagine the conversation. Say I'm a mobile phone and I want to print out a text message. Hearing this, an inkjet printer would say to the phone, "I can print your message. Just send me the text and I'll do it right away." Or, alternatively, another mobile phone might say, "Sorry, I'm the wrong kind of device and I can't help you, so don't waste time talking to me." It would be very nice if it worked like that, but because it doesn't — yet — Sony have included an application called Bluespace that lets you intervene in the Bluetooth discovery dialogue and tell the computer what kind of devices are out there searching for partners. I managed to borrow an Ericsson phone with Bluetooth and had no trouble getting each of the two devices to acknowledge that the other was there. I never did manage to exchange any information, but it might have worked if I had had manuals for either device and more than 10 minutes to mess around with them. You can normally tell when a device has Bluetooth. Look for a bright-blue LED that probably consumes more power than Bluetooth itself. (By the way, you can see why Bluetooth was named after the viking of that name and not after his son, who was called Forkbeard.)

Home | Search | News | Current Issue | Digital Editions | Articles | Forum | Subscribe | Shop | Readers Ads

Advertise | Information | Links | Privacy Policy | Support