I’ve talked recently about the use of a timecode sync in home multitrack tape recording, and for this post I’m going to expand on that with an example recording, then a breakdown of exactly how the track was put together.
It should be recognised that whilst a timecode sync did keep some instruments live until the mastering stage, the finished song was mastered to an audio cassette. Everything you hear in the example, then, has been on at least one audio cassette – some of it has been on two.
Sadly, this is the Internet, and you’re not going to hear the quality of sound the tape has preserved. I obviously had to digitise the recording, then compress it to an MP3, and Soundcloud’s upload/processing has degraded the compression further, so the result is noticeably inferior to the tape master.
You do, nevertheless get an idea of the recording’s basic aura. I could have gated out the tape hiss, but on a blog about tapes and tape recording that would have been stupid. You’ll notice that the general brightness of the instruments masks the hiss for most of the recording, but in the middle, where the instrumentation drops right back, you can hear some blatant combined Portastudio and master tape noise.
Here’s the recording…
HOW IT WAS DONE…
The recording was actually made back in the ‘90s, but when I was using an analogue setup I’d typically log the details for future reference, and sure enough, diarised for Monday 19th April 1999, is a full rundown of the recording session.
The multitrack recording devices were a Tascam 424 Portastudio, running a timecode track to sync with a Pentium 1 MMX PC, which recorded the MIDI instruments using the Cubasis MIDI sequencer software… Ah, Cubasis MIDI – how simple sequencers once were! Cubasis MIDI did have editing facilities, but still, a lot of the more complicated passages were played live, in real time, and then simply copied and pasted on a long loop as and when necessary.
The final result was mastered onto a Yamaha KX-260 stereo tape deck (shown at the head of the post). The tapes used were a Sony SF 60 for the Portastudio multitracking, and a That’s VX 60 for the final stereo master. Both cassettes are high bias Type IIs.
Above: what an exciting time we had in the 1990s, with our technology magazines and gadgets. The black box, in case you don’t recognise it, is an audio tape to MIDI sequencer timecode sync.
BREAKDOWN…
The piano, which is really the main instrumental component, was an Alesis NanoPiano module, played via a Fatar Studio 900 full length, weighted action keyboard. The preset was Bank A, Number 1 – ‘TrueStereo’. I recorded the NanoPiano parts in real time, to the Cubasis sequencer, and then let the timecode box sync them with the tape so they wouldn’t need to use any Portastudio tracks.
The drum sounds came from a Boss Dr Rhythm DR-660 beatbox. Again, I recorded the parts to Cubasis, and kept them live throughout the recording using the timecode tape sync. There are no sample loops – it’s 100% Dr Rhythm. Here’s a bit more about the drum content, exactly as I diarised it in 1999…
I used a custom kit rather than a preset, and since the nicest sounding snares tend to be a little innocuous volume wise, I reduced the level of the rest of the kit to bring them roughly into line. The main beats were quantised at 70%. Some of the fills went down as live takes, with any remedial work undergoing individual note by note attention in the [Cubasis] editor.
I’d forgotten that I used to pull down ALL the other drum sounds to emphasise the snares. I suppose there are some advantages to having a trainspotter mentality and logging random information.
Two separate organ parts came from two separate hardware devices. The rhythm organ sound was produced by a Hammond XM1 module, whilst the high pitched ‘lead’ organ notes were added with a Korg X5 synth. Both of these parts were recorded to Cubasis’s MIDI tracks and kept live throughout the process. Surely, I’ll have to start using the actual tape soon?…
Yep. It’s tape time! The brass sounds also came from the Korg X5, but because the synth’s live output would be occupied playing the sync’d organ part, I had to put the brass sounds onto the Portastudio tape. Let me explain that a little further…
The X5 is multitimbral (meaning it can play multiple sounds at once), but its onboard effects are not multitimbral, and since the onboard effects settings would be very different for the organ and the brass sounds, the two components could not run live simultaneously, and one of the synth’s two roles had to be tape recorded.
If you’ve ever tried using an old S&S synth for brass, you’ll probably have found that it’s a lot more convincing when you set the output to monophonic, like actual brass instruments – and then layer up single note lines. With a three-part brass, like the one I wanted, this meant playing each of the three monophonic lines separately. I could have done this on the digital sequencer and then recorded the result to tape with the timecode sync, but I found it easier just to do the whole thing on tape, with one bounce…
So, I’ve got the timecode on Track 4 of my Portastudio tape. Three tracks available. I need one track for the bass guitar, but I haven’t taped that yet. So to get my three-part monophonic brass I record the first part to tape Track 1, overdub the second part to Track 2, then bounce Tracks 1 and 2 over to Track 3, whilst also playing the third brass part live. The result is that I now have timecode on Track 4, my full brass ‘ensemble’ on Track 3, and Tracks 1 and 2 become available again.
The bass guitar now went onto Track 1 of the Portastudio tape, leaving one more analogue track spare.
I could have used the final tape track for the vocals, but I chose instead to sing live, as part of the final mastering process. The advantage of doing this was that the vocals could be recorded with stereo reverb (a single Portastudio track would only render mono), and a stage of degradation would be omitted, meaning better quality for the vocal sound.
LOOSE ENDS
To tie up a loose end… The mixing of the live electronic hardware instrument sounds was done using the actual volume controls on the instruments – not with the Cubasis sequencer. Cubasis did have a mixer, but I couldn’t be bothered trying to link up the digital control parameters with those of the outboard hardware devices. There were only a few live instruments to adjust anyway – I wasn’t mixing a 48-piece arrangement.
And basically, that’s how a tape-plus-timecode Portastudio recording would work. As I mentioned in the Ten Things You Didn’t Know About Portastudio Recording post, it was all about picking the right instruments to tape, and more importantly, the right ones not to.
Tape Tardis 