Mixing and Recording
As of 1995, Numb had become the priority focus for me in terms of music, but I had also accumulated this handful of T4M tracks archived on disk in various stages of completion.
I’m not sure what the precise circumstances were, but sometime around the Spring/Summer of 1996, Don generously suggested that I use his gear to put together the newer 4th Man material while he was away for a week or so. I determined that I had about 45 minutes of new material, so it made sense to take advantage of the situation, hole myself up with the gear, feed Don’s cats, and see how much I could get done. This was, at the very least, a great chance to dig into the Numb equipment. It was also my first chance to work on T4M material in a reasonably proper mixing situation.
Besides having access to an unprecedented 32 mixer channels and multiple outboard effects devices, Don was using "Logic" as his sequencer, a far more advanced version of Creator, my sequencing program. Thanks to backwards compatibility, Logic would load my Creator disks – almost flawlessly.
I finally had a chance to record the vocals in a slightly more professional manner. On the previous T4M CD, we usually recorded the vocals "wet" through effects onto the typical 2 track allotment on our 4 track in order to free up our one effects unit for mix-down. In this case, I was using an 8-track digital recorder, keeping the vocals clean until mix-down. Factoring out the rough band track and SMPTE, that left me with a whopping 5 tracks to record alternate takes and double the vocals, etc. In short, more options.
Having mentioned SMPTE, this seems like a fantastic time to take a stumble down memory lane and discuss that wonderful artifact from the past – the process known as "striping the tape". For those of you who don’t know what I'm talking about, get ready to be bored and misinformed. For those who do, bear with me while I lie to these other people.
First, think of high speed internet, and then compare that with two empty soup cans connected by a string. That’s what striping tape is like. More specifically, in the days of analog tape, prior to the common use of having audio recorded digitally, there was the mechanical issue of analog tape stretching and small variations in the speed of the motor controlling the tape reels. This on its’ own was not an issue, but as I discovered back in 1988, vocals or instrument tracks recorded onto analog tape will quickly go out of time with instruments being played by a MIDI sequencer unless there is a way to synchronize the timing between the tape tracks and the MIDI tracks. Enter the world of the "sync tone".
The principal of a sync tone is similar to dial up modems and fax machines. A digital signal is converted to an audible signal that can then be transferred over the then limited bandwidth of a phone line. In the case of a sync tone, a series of specialized audio tones are recorded sequentially onto one of the tape tracks. On play back, these tones can then be routed to a device, which will then convert the tones to MIDI time code so that the MIDI sequencer (controlling the digital instruments) can match the small fluctuations in speed on the tape. Sounds great, right? I hated having to piss around with time code.
As much as the recording equipment was an improvement over the entry level gear we used for the first CD, I was once again contending with the gear being located in a residential apartment with all the contingent issues connected to that - thin walls, neighbours to consider, and (as usual) my own crippling self-consciousness. As a result, the vocals for Artman, 'onaniershow', Boredom, High Plains Castrato, Filler, Tasteless Killing Spree, The Happy Song, and Belgian Sock Fite were finished in one afternoon, doing about 3 takes per song and then quickly moving on. The entire process took about 4-5 hours.
This process had good and bad aspects to it. On one hand, there was little time for hemming and hawing, the result being that shit got done. I had to commit to the (*cough*) performances as they were. The down side was that, excluding a few of the live tracks, I“d never actually vocalized the lyrics for these songs before, so I was unsure as to how they were going to work. The result was that the performances were less than nuanced. Instead of taking the time to perhaps pull together a more dynamic performance, I fell back on my old party tricks of shouting and growling, hoping that I could "fix it in post", which essentially meant running the resulting vocals through distortion. This changed nothing in terms of the usual method T4M had always used as far as vocals went – print 'em and don't look back (maybe later, with regret).
There was another issue with the time factor. In order to take advantage of a mixing board with significantly more possibilities, a large chunk of time was required to configure my sampler for multiple outputs so I could mix the virtual tracks. If you don’t know what I'm talking about when I say "virtual tracks" and love to be bored, then you just got lucky, because I intend to explain the now moot concept.
Everything is Virtual Back in the "day" (before the cheap ubiquity of digital multitrack recording and VST instruments – oh, say, the 2000s), most MIDI devices could be synched up to the tracks recorded on tape using my favorite dubious ally, SMPTE time code, and be played back live by the sequencer and run through the mixing board with the rest of the recorded tracks. These MIDI generated outputs were often referred to as virtual tracks. There were multiple advantages to virtual over tape. It freed up the then-limited number of tape tracks for other purposes, and it also meant that the signal was cleaner. As much as people crow about the charmingly warm sound of analog tape, it tends to bring up the noise floor, and in the wrong hands (mine), things can really turn into a hiss-fest (to this day, even when dealing almost entirely in the digital domain, if there is a point where noise from analog circuits, tape, or microphone leakage can feasibly be accidentally introduced to a recording, it will happen on my watch). Another advantage to virtual tracks is the flexibility in terms of being able to change sounds and notes, something that isn’t really feasible once you’ve recorded it onto analogue tape.
In the past, since we’d usually only had a 4-track (and in some cases a "no-track") recorder, almost all of the tracks from the first CD were initially "virtual", however, due to the limited number of adequate mixing channels on our equipment, we'd never been able to properly take advantage of separating the song elements and mixing them in the traditionally effective method studios used. With access to the Numb studio, I had this opportunity, which takes me back to the original challenge.
The basic rerouting was theoretically a simple process, but I was once again working with the Yamaha TX16W as my core sampler, and as I mentioned in the liner notes for the first CD, the Yamaha TX16W sampler's operating system was a bit awkward to use. I have heard the term "pig", "beast", and/or "piece of shit" used to describe operating the 16W (depending on who you asked). As comfortable as I was with it, routing the instruments to separate outputs was time-consuming. In some cases, it couldn’t be done, and re-sampling to another device was required. This basic and unglamorous librarian work was 20-30% of the process.
Once the librarian work was done, pulling it all together and making the creative decisions of mixing was kind of an anti-climax. This was before the days of unlimited vst effects chains on every track and user friendly mixer automation. Cutting and pasting variations of the mix in retrospect was possible but not really on the radar yet for me, and I only had two shaky hands for the mix-down process. In retrospect, I possibly would have not made some of the choices I did, but it was really triage work, particularly since the clock was ticking. Add to it my dyslexic approach to all things goal-oriented, and you’ve got a real tuxedo and gum boots sort of a mix.
For all my supposed knowledge in terms of audio recording, I was a long way from being effective in terms of practical application in 1996 (that could also be said at the time of this writing in 2014). One of the fundamental major errors I made was in determining my reference mix. Be prepared to nod off again…
When mixing in a new environment (and in general), it is good to have an example of a mix you are familiar with which has the qualities you would like to emulate (or avoid). As an extreme example, if you want a recording to have similar sonic qualities to Abbey Road by the Beatles, you would not want to use a recording of Louis Armstrong from the 1920s as a reference. A common mistake (or at least the mistake that I made) is to use songs that appeal stylistically, but not in terms of overall eq and dynamics. Much of the music on popular radio might not have been to my taste, but the mixes can be and usually are fantastic compared to many of the bands I was inspired by.
So right from the start I was referencing to mixes by industrial bands from the 1987-1991 period that were often not quite as clean and punchy as what was being done in 1995. Later, when hearing my mixes next to the latest by Numb, Front Line Assembly, or pretty much anyone else I didn't use as a reference, I was a wee bit disappointed. That being said, were I to be slightly dishonest, I could have said I was showing a "punk" mentality to mixing.
Most of the songs got at least half a day's worth of work before committing to a mix. Since "Boo" was such a simple track with no vocals to contend with, a large number of elements to the track were worked out during the mixing process, with plenty of live knob-twiddling, something I’d only had limited capability to do prior to using the Numb gear. It was like being a kid in a small but reasonably well-stocked candy shop, and in a way, my glee might have surpassed my attention to detail in terms of the final product. I was also perhaps a bit overly enthused at the constant Blade Runner-esque view of Vancouver’s Robson Street, squirming with people filing in and out of the tourist traps.
I ran out of time towards the end, and was unable to pull Artman and Boredom together as full mixes. They were quickly run straight out as two channel mixes with some basic "industrial" effects processing slapped on the vocals – much more like the process done for the first T4M CD.
For better or worse, it ended up sounding like 4th Man material, with heavy reliance on the then-woefully unhip fm synth sounds from my Yamaha TX81Z, plenty of mismatched sound elements, and typically trashy sounding lo-fi vocals. It was perhaps cleaner sounding than the 1994 CD, but all said, I was skeptical to consider the songs "finished". With Gabriel and Martin no longer involved in the project to function as a glee club, I became biased by my own self-doubt. If left alone with my head, I tended to deem most material unfinished or too substandard to release.
It was entirely likely that Don would have allowed me more time to refine the songs after he returned, but I felt that I had at least put together some decent demo versions, and I didn't want to push my luck in terms of the hospitality, still being new to the band. Besides, moving forward, Numb was starting to work on new material, and Don had much better equipment for me to play with.
Around this time, there was some minor renewed interest in the original CD, as well as potential to put out the new material. Tasteless Killing Spree and a quickly remixed version of Dare ended up on a compilation. Because I didn't have easy access to a multi-track for the Dare remix, I ended up doing the vocals live while I was recording the band track onto the master. The remixed version, 7 years since the original had been recorded, took mercifully less time.
After this, a Gabriel-less T4M did one or two more live shows (which actually turned out really well for a change), and then I ceased to work on any more T4M material. By 1999, when I could have feasibly returned my focus to it, most of the original T4M gear was in various states of non-functionality, and new equipment still seemed either too expensive or too clunky for what had become basically a hobby. The idea of really pulling ideas together started to become feasible again by the mid to late 2000s, when multi-core systems, DAWS, and vst plugins significantly came down in price. This was also around the same time that a certain Mr. Abney suggested that maybe we could consider kicking the tires on the old T4M jalopy again.
Mr. Abney and myself will most likely be wearing adult diapers by the time we have assembled the equivalent of an albums worth of new material. On that subject, I am of two minds simultaneously – too little, too late, AND better late than never.
Feebly drooling and mumbling,
David Collings, 2014