Paul Wiffen has a sleep problem, sees the mLAN light, and rounds up the best of the Frankfurt launches for Apple musicians.
My work schedule was somewhat interrupted this month by not one but two of those Nightmare On Elm Street situations. I don't have space to go into both, but the first might contain useful information for others, so here goes. It concerned not one of my computers but a G3/233 desktop machine I look after for a colleague. In a spare afternoon when he wasn't using the computer, I was asked to install a USB PCI card for video conferencing.
I shut down the computer, popped the lid to install the tiny card, and rebooted. However, before the extensions appeared on the screen, a crash dialogue reading 'bus Error' appeared. Not panicking too much at this stage, I rebooted, and this time it said 'Illegal Instruction'. Subsequent reboots produced one or other of these messages with monotonous regularity. Holding down Shift to stop Extensions loading made no difference (presumably because the computer wasn't getting that far) and neither did moving the card to another PCI slot, or taking it out altogether. I asked Steve, the computer's usual operator, what had happened the last time he had re‑booted (attempting to ascertain whether the problem was tied to the appearance of the USB card). He said he couldn't remember, which turned out to mean that he couldn't remember the last time he re‑booted! He had just been putting the computer to sleep at the end of every working day and waking it up the following morning. The last time it had been rebooted was when someone else had upgraded it to OS9, about a month before.
The $64,000 question is: was constantly putting the computer to sleep instead of shutting it down causing it to refuse to boot now, or was it something to do with the arrival of the USB card? I know that keeping the OS running for a long time without re‑booting can fragment the memory (leading to something of the same inefficiency you get with a fragmented hard drive), but could it really cause this very early boot hang? The problem was that if Steve had introduced some conflict‑causing element weeks before, he wouldn't have noticed, as he had not rebooted. Maybe I was discovering some such conflict simply because I was the one who had re‑booted the computer.
I fished out an OS9 CD‑ROM and held down 'C' while re‑booting, to make the computer boot from the CD, but the G3 absolutely refused to see it, and repeatedly hung at exactly the same point. Disconnecting the boot drive didn't help either. Now the Mac just hung forever on the smiley‑face Mac screen. Nothing I did would persuade it to boot. Even zapping the P‑RAM had no effect.
In the end, after two days of no joy, to get Steve up and running I lent him one of my computers, in which I installed both his IDE boot drive and large audio drive so that his work could be rescued. But the G3/233 remained dead until I talked to Achim Kruese, designer of the C‑Lab TimeMachine and the only person who was able to shed any light on the problem. He suspected that the PCI buss was blocked by the USB card even after the card had been removed (which would explain the buss error message). He suggested switching off the PCI buss, by holding down Apple, Shift and 'D', and then zapping the P‑RAM again. I did, and it worked!
After the sheer hell of OS problems, I went to the opposite extreme in Uznach, Switzerland, finally having a first‑hand mLAN experience, after months of writing about mLAN, talking about mLAN, even dreaming about mLAN. It all started when I received a phone call a week before the Frankfurt Musikmesse from Daniel Feusi of Swissonic, asking if I could come over to Switzerland before the show and bring the prototype of their AD8 digital recording interface/converter, which I had had for demonstration since the Apple presentation I reported on in April's Apple Notes. They had received an mLAN board from Yamaha and wanted to try to get it working before the Musikmesse, so off I went.
On my arrival in Uznach, Daniel took the AD8 and disappeared into a back room, from which soon came the smell of solder. Half an hour later he emerged, with a board hanging from the back of the AD8, and connected one of the FireWire sockets it contained to a G3 Mac. We dragged the mLAN ASIO driver supplied by Yamaha into the computer's Logic Audio ASIO folder and booted the program. Ten seconds after opening the mLAN patchbay and doing an update, we had eight inputs available for recording in Logic, with a latency readout of 5mS. I couldn't believe my eyes. Surely it had to be more difficult than this. We hadn't even restarted the G3, for god's sake!
After a little tinkering with the configuration, we had a full clickfree setup with 24bit/48kHz sample fidelity. The next thing I knew we were actually making eight‑channel recordings into Logic. There was a slight hiatus while we tried to work out how to listen back to them (the AD8 has eight inputs, with a stereo monitor allowing monitoring during recording). We ended up enabling the driver for the Sonorus card that was also in the G3 and playing back through its stereo audio out. The quality was amazing, but even more exciting was the fact that we were bringing eight channels into the Mac on a single cable, with no PCI card in sight.
We very quickly decided that the AD8's mLAN option card would also have a stereo analogue out and an SPDIF output (even with three FireWire connectors, there is still lots of room on the option board for other sockets — the AD8's AES/EBU option requires four XLRs and a BNC connector!), as in Cubase you can't use another driver at the same time as the mLAN driver. The only thing we couldn't try was recording at 24/96, because this first generation of Yamaha chips maxes out at 48kHz (the new chipset jointly developed by Yamaha and Otari will be available in the summer, as detailed in the last part of my mLAN series, SOS November 2000). The AD8 mLAN option board will also ship this summer.
Following my initial brush with OSX, detailed in the March 2001 instalment of Apple Notes, I am still a bit reluctant to get back into that particular lion's cage, having been psychologically scarred, if not actually suffering any permanent injury (a melodramatic way of saying there was no actual data loss). But, with some prompting from Debbie at SOS, who tries to keep us columnists to our word (she often remembers what we promised to do in previous columns when we ourselves have forgotten), I have decided that it is time to take another look at that particular issue. This time I will start with a hard drive on which there is nothing at all — I think I need to get to the bottom of this business about why OSX wanted to reformat the drive... By the time you read this, the full release of OSX should already be shipping, so I will try and work with the full release rather than the beta, to provide a more accurate picture of what it will give you (other than compatibility problems!).
There were several major items of particular interest to Mac users at the Frankfurt Musikmesse. Of course, Melodyne, the revolutionary new pitch–shifting software from Celemony, was wowing them in the aisles, but this was covered in the NAMM show news in the March issue of SOS. The most exciting thing for me at Frankfurt was an additional feature I was told about in Live, the new Acid‑style program from ex‑Native programmers Ableton — the ability to record and play back audio loops on the fly. I have been looking for this feature, which I used to have on the Plasmec ADAS running on the Atari, for 10 years. (In fact, I used it on the demo I produced for a Julian Colbeck home recording instructional video — take note, all those who still ring me to ask how the demo was made.) It will be great to finally use the same technique quickly and easily on the Mac, with multiple loops and the ability to record through plug‑ins. Eat your heart out, Robert Fripp!
A PC‑only bastion looks set to fall as IK Multimedia's new SampleTank VST Instrument invades Nemesys' GigaSampler territory with the ability to play really large multisamples, but from within VST. I spent a very pleasant hour going through the 2Gb of sample material that comes with the full version of SampleTank. The 128Mb piano was a joy, and drew admiration from passers‑by even with my poor playing, but it was when I started to go through the orchestral stuff that it dawned on me that I was encountering the sort of quality, playability and completeness you could normally only expect from CD‑ROM collections costing more than SampleTank itself. Even the 650Mb of sounds which come with SampleTank LE, at £179, are far better than the average thin set of sounds normally supplied with VST sample playback modules.
There was good news, too, on the USB front, as Steinberg announced that they would be licensing the technology which facilitates the rock‑solid timing in their Midex 8 MIDI interface to Midiman, for use in their smaller, cheaper USB interfaces (some of which are badged by Steinberg). This will mean that even if you can't afford anything more than an entry‑level USB interface you will still have the right to good MIDI timing, as long as you follow Wiffen's Law and only use the interface provided by the people who wrote your sequencer (or their strategic partners, in the case of Midiman).
As I prepared to leave Zurich for Frankfurt, I had a rather surprising phone call asking if I happened to have any DX7 sounds with me (I haven't been asked such a question in years). This was because Native Instruments (having produced the definitive Prophet 5 and Hammond Organ plug‑ins) have now turned their virtual attentions to Yamaha's best‑selling FM synth. As I would be helping them out with demos at Frankfurt, I was being prepared psychologically to play FM sounds in my demo. However, when I managed to look at the FM7 FM synth plug‑in, I found to my delight that it wasn't restricted to Yamaha's six‑operator sine‑wave implementation but also allowed other, more complex, waveforms on the operators, as did the early Synclavier and Oberheim Xpander. Also (thanks to an idea given to him by Peter Georges of Wizoo!), FM7 programmer Michael Kurz has managed to find a way to do Pulse Width Modulation (my favourite way of changing harmonic content in analogue synthesis). FM7 also has a distortion operator and built‑in chorus, flanging and delay, so maybe FM synthesis is for the likes of me after all...
I've just received news of the release by CPS Software of a new type of Mac/PC editor for Yamaha's CS1x synth. Essentially, it addresses the problems found in the synth's original design (ie. that you can only edit the first four channels from the synth's front panel when it's in Performance mode). I had a brief look at the editor, and although I don't have a CS1x to evaluate it properly, it looks very comprehensive. The editor is shareware and is available for just $15 from ShareIt (www.shareit.com). CS1x owners who would like to know more can check out the website of designer Chris Share (appropriate name!): www.labyrinth.net.au/~cpsmusic/C...
There has been an awful lot of misinformation recently about using Yamaha mixers with mLAN, much of which has been relayed to me by email or phone. When two friends of mine (one with an 03D, the other with an 01V) both contacted me within 24 hours to say that they were being told they couldn't use mLAN at the moment (the 01V's owner had been told never!), I decided that it was time to speak to James Baker at Yamaha Pro Audio and Nick Howes at Yamaha R&D to get the real story. It only took a couple of minutes to sort out, but I can see how the confusion has arisen.
The CD8‑mLAN card will only fit in the 02R and 03D mixers, being a full‑height YGDAI (Yamaha General Digital Audio Interface) card, so 01V owners will have to wait till Yamaha's mini‑YGDAI card is available in the Autumn (a lot better than never, as my friend Phil agreed). The firmware in the 02R is pretty much ready for CD8‑mLAN and will be final by the time Yamaha get me one for review next month. Although the firmware for the 03D is not quite so close to being final, Nick Howes tells me that the only problem he can find in the latest version is that the 03D cannot identify the card as a CD8‑mLAN, but that this doesn't prevent it from working reliably. With any luck, the 03D's final mLAN firmware will only be a few weeks behind that for the 02R.
The final confusion that seems to have arisen is the idea that you won't be able to use mLAN with PowerBooks. In view of the super–computer status of the new Titanium G4/500 PowerBook, it would be a shame if this were the case. Nick Howes thinks that this falsehood has gained currency because you do have to stop the Powerbook putting itself to Sleep, as the FireWire port won't then wake up properly (I have found the same problem with USB). So there's another reason to go straight to the Energy Saving Control Panel on your Mac right now and set at least the CPU Sleep parameter to 'Never'.