Faced with unfinished album sessions on analogue tape and a studio full of PCs and digital gear, Brian Heywood gets a chance to find out how flexible modern equipment really is.
Early in January, my band and I decided to make a concerted push to finish the album that had been on hold since our old drummer went off to Australia (maybe it was something we said?). Nightwatch is my folk/rock/jazz fusion band; we've been together about 10 years now, but we've never quite got around to recording a full‑length album. The timing seemed ideal, the new line‑up had settled down, and I didn't have any major projects tying up the studio. As the production relied heavily on the PC‑based systems in my studio, I thought it might be of interest to the readers of this column.
The studio is really quite modest by commercial standards, with three recorders: a single ADAT (eight tracks), Soundscape SSHDR1 and SADiE hard disk recorders, both the latter being hosted on a 486DX33 PC. Sequencing is provided by a 486DX50 PC running Voyetra's Sequencer Plus Gold — a DOS‑based package — and interfaced to the MIDI instruments via a Voyetra V24s interface. The synthesizers used for this project were a Yamaha TG77 and TG500, a Roland Sound Canvas (CM300) and an Akai S950 sampler. We had up to 16 audio tracks at our disposal, which is actually one more than when we did the original sessions for the album on 16‑track tape, since we didn't need to allocate a track for timecode.
Reeling In The Years
The original sessions had been recorded using a Fostex 16‑track, so we had two reels of half‑inch tape with some basic rhythm and melody tracks that we didn't want to lose, plus timecode on track 1. The first job was to transfer the material we already had to the ADAT. So I arranged with the truly wonderful Vic Lennard to hook his Fostex 16‑track to the ADAT via a JL Cooper DataSync2 and a Fostex synchroniser. As we had to redo all the drums, we didn't need to transfer them and — as there were never more than eight tracks of other instruments — the ADAT had enough tracks to take the material. Once we had worked out how to get the two sync boxes to 'talk' to each other, the process went very smoothly, with the only slight glitch being caused by a piece of discontinuous timecode.
Back in the studio, I transferred most of the tracks onto the hard disk systems, since we wanted to record the new drum parts onto the ADAT. The reason for this was two‑fold: firstly, the new drum parts were likely to need the most 'space', since they would need to run through almost all of the tracks; secondly, we wanted to keep the tracks together on the same recorder, and the hard disk systems only have four individual outputs apiece. The first of these points shows an advantage of traditional tape‑based recording: you have considerably more storage space than is common and affordable in a hard disk system. Consider a 42‑minute ADAT tape: you get the equivalent of 8 x 42 x 5.025Mb (digital audio requires about 5Mb of disk space per minute) — 1.7Gb of storage per reel. Just for the drums we would have needed more than 2Gb of hard disk storage (we went into a second ADAT tape). Hard disk systems store data much more efficiently than a tape‑based system, but in this case, this feature would have not made any difference.
Let's Work Together
For the rest of the project, the ADAT was the 'sync master', which is to say that the computers all derived their synchronisation from the ADAT via the DataSync2. This was because we had no way of getting the ADAT to 'chase' external SMPTE timecode generated from the computers. This system turned out to be very robust (at least when the ADAT was working) and only had the slight downside that the ADAT had to be continuously cycled when performing overdubs — which slowed things down a bit and increased head and tape wear. The JL Cooper device is a little gem, as it directly converts the ADAT's internal sync signal into MTC and SMPTE — and even MIDI clock and SPP (Song Position Pointer) if required. This meant that we could drive the Soundscape system directly off the MTC output while SADiE and the sequencer were being driven by the SMPTE timecode output.
Creating the final version of each track was a three‑stage process: track‑laying, mixing, and post‑production. The track‑laying consisted of overdubbing new performances onto the various recorders, which sometimes got slightly confusing, especially if we were swapping between the two hard disk recorders: their user interfaces have substantial differences, especially with respect to their keyboard shortcuts. On the whole, I think I preferred using the Soundscape over the SADiE for track‑laying, although it was slightly annoying to have to define the start and end points of each take before commencing recording (thankfully, this requirement has been removed from the latest version of the Soundscape front‑end software).
The live parts were laid down in sync with the sequencer, which we used to record a live performance rather than pre‑sequencing the parts. The live instruments tended to be placed on the same tracks for each song, so that the basic monitor/mix layout remained constant. While this removed some of the flexibility of the system, it prevented us from getting confused when recording new tracks. This also paid off at mixdown, since the EQ for these tracks stayed pretty much the same. It wasn't too much of a problem to transfer an ADAT track to SADiE, correct any minor glitches, and then transfer it back, though we only really did this with the tracks recorded two years previously. When recording live tracks, we tended to repeat the overdub if there was some problem. The whole point of this approach was to keep the sessions as natural as possible, so the technology wouldn't get in the way of the creativity and feel of the recordings.
Mixing It
After taking a break for a couple of weeks, the producer and I settled down to mix the tracks. As we don't have an automated desk, we used a technique of setting up a static — or at least a very simple — mix, and then recording it onto DAT. Each segment of the mix was recorded with a couple of seconds of material on each side of the edit points, so that we could crossfade between them. The segments might be several minutes long or only a couple of seconds, depending on what we were trying to do. As we created the mix in stages, we were able to generate an extremely complicated final product.
The post‑production stage consisted of digitally copying the audio segments from the DAT back onto SADiE and then stitching them together in the editor's Trim window, with crossfades to ensure that the joins were seamless. Then it was simply a matter of putting the finished tracks in the right order, balancing their overall levels, setting the inter‑track gaps to the correct length, and copying the material back to DAT via the S/PDIF digital output. The DAT could then be used as a production master. We backed up the EDL (Edit Decision List) of the master from SADiE to DAT, so that the mix can be tweaked at a later stage if required. One nice feature of the SADiE is that it allows you to burn a Red Book standard 'gold disc' CD using a CDR drive (in our case, a Yamaha CDE100) and then play the mix in any domestic CD audio player — which is a great way of sending the finished product to interested parties.
The Last Word
As this was the first album‑length project I had recorded with this particular studio configuration, I learned a few things about how to best use the equipment. The first is: don't forget to back up your data. This does not apply only to the hard disk systems: the ADAT went down during the course of the project, causing damage to a 30‑second section of tape. Unfortunately, the ADAT data had never been backed up, so that section is lost forever. If you don't have two ADATs, consider hiring a machine for the day on a regular basis — say for once a week, or at key points in the project.
In an ideal world, I think I'd prefer a single hard disk system working with the ADAT, so I'm rather looking forward to seeing the planned new Soundscape hardware, which it seems will expand the number of separate inputs and outputs — the Soundscape would be my system of choice for track‑laying and mixing at present if it had enough outputs. We could also mix directly to SADiE rather than to DAT, simplifying the mix and post‑production phase, and reducing the wear and tear on the DAT recorder. One last point about using a hard disk system in this kind of project is that you can never have too much hard disk space. We ended up with only 40 seconds of free space on the Soundscape system. Still, we couldn't have done it without the two hard disk systems.
Let Your Fingers Do The Walking
While I'm talking about hard disk recorders, Sypha have just released the 5th edition of their excellent Tapeless Audio Directory. This A4‑format book describes over 300 non‑linear audio systems, ranging from Windows soundcards and software packages up to expensive professional dedicated systems. The directory is an indispensable aid to anyone buying a system for professional use — or just for keeping in touch with what's going on in the tapeless recording world. The coverage of some of the low‑end systems is fairly brief, but as contact details are included, more information is always available for the cost of a phone call. Highly recommended. It's available from the SOS Bookshelf at £14.95 plus £2.50 UK p&p (£4.50 overseas). The order code is B304.