Studio is a place to discuss my interest in the tools of the music makers. I have collected such tools over the decades, with the thought of doing some composition and recording in my free time. Most of the studio hardware is ancient, relative to expected electronic gear lifespan; it is heartening to see that it all still works and still reaches state of the art.
Our first studio consisted only of a Mac program called Deluxe Music Construction Set. Owen wrote and scored music with it for JrHi music projects, and played them on a piano and recorded them. That got me interested in keeping up with the state of the art and acquiring more powerful tools and computers and software, and acquiring our first keyboard synth, which remains the heart of the studio.
Studio investment ceased 20+ ago, with only the minimum upgrades continuing to keep software current and hardware maintained. The core of the studio has always been Yamaha and Apple, with Roland, Korg, Kurzweil, Kawai, and E-mu a strong supporting cast, and Presonus, Alesis, and Emagic making everything work together. It is good for my aging brain, keeping it all up to snuff, and keeping myself current with it.
The purpose of a musical studio is to create music, everything from song to symphony. Creating music used to mean composing music that available instruments could play. But now a new dimension is included, actually designing musical instruments to play the composed music. Using a process called music synthesis, an instrumental sound (timbre) can be created and then played back when triggered from an electronic version of the musical score. Analogously, visualize a player piano, playing music mechanically from a paper-roll version of the score, and imagine the piano had several sounds it could be told to play.
Music synthesis often tries to emulate real instrumental timbres, creating a ‘band in a box’. For example, an electric piano sound is a synthesized acoustic piano sound. The advantage is clear; one could dispense with a real band (union rates put a pro band beyond reach of hobbyists!).
Realistic instrument timbres can be generated by sampling sound from an actual instrument, or by modeling an instrument’s sound from first principles of physics and acoustics.The quality of the timbral reproduction is determined by the finesse with which the acoustic instrument sound was sampled or modeled.
Yet the power of music synthesis goes way beyond emulation of actual instruments. Given the wide variety of tools available, only one’s imagination limits the possibilities. Synthesized instruments, capable of dynamic timbre shifts, can create great swirling masses of sound, generated by the variety of modulation sources available within synthesis tools.
A synthesizer is a sound generating module for creating such custom instrumental timbres. Originally, synthesizers were analog hardware devices using electrical circuits: voltage-controlled oscillators, filters, and amplifiers. Their outputs were analog sound, represented by varied electrical voltages in a cable. Gradually, hardware modules became digitized, becoming small, special purpose, sound-generating computers.
Ultimately they morphed into software applications running in general purpose computers and producing digital sound. Today’s computer software audio processing tools require a digital audio signal; the analog signals of older synthesizers must be digitized using an Analog to Digital converter (A/D or ADC).
A raw musical performance consists of multiple audio signals, called tracks, typically one track for each synthesized instrument, and/or one track for each microphone used in recording a live performance. Tracks are merged and distributed horizontally in the sound stage (panned) via a track mixing process, typically resulting in a two track result (stereo final mix).
For a typical song, live vocal/instrumental tracks are recorded by microphones interfaced to a computer. Backing tracks (e.g strings, percussion, vocal harmonization) are synthesized on the computer, then merged (mixed) with the live performance audio. There are innumerable variations on this theme, accounting for most all of the popular music heard today, from entirely live to entirely electronic, and all in between.
A DAW is a software based digital audio workstation capable of creating and processing the various audio tracks comprising a musical performance. A DAW’s processing repertoire includes sound creation and storage (parameter sets determining specific timbres); performance creation and storage (the set of raw performance tracks); track mixing, editing, enhancing, recording, and scoring.
Track aural enhancement is performed by applying effects (FX). FX is the generic term for various signal processing functions, such as equalization (EQ), chorus, delay, and reverb. A track prior to FX application is called ‘dry’; after FX, the track is ‘wet’. FX can be applied to individual tracks, and to the master tracks resulting from the final mix.
MIDI (Musical Instrument Digital Interface)
Musicians since the mid-80s have had available a computerized representation of a performance, the MIDI file, alongside the centuries-old human-readable representation, the musical score. A MIDI file (think of a piano roll, or a sound disk in a music box) is a computer file of general instructions regarding the pitch, duration, and volume of the musical notes, as well as other coloration to the sounds being played. The streamed instructions in a MIDI file are understood by all MIDI-capable music playback devices.
The DAW may be used to generate a MIDI file through its user interface. But those with even rudimentary keyboard skills will likely choose a keyboard MIDI controller to generate a performance. Typically, this will be an external, digitized keyboard with complete MIDI capability, which acts as the source of commands specifying MIDI notes and dynamic effects.
The MIDI controller will typically provide a way for the performer to dynamically color the sounds produced in response to a MIDI stream. Such dynamic coloration is initiated by wheels, sliders, and other control interfaces offered by the controller. These continuous controllers produce commands to the sound modulation functions of a synthesizer, causing the affected notes to have dynamic changes in emphasis, pitch, and timbre.
The MIDI controller typically also includes an internal synthesizer, so it can play its own generated MIDI commands as they are produced by the keys. Optionally, it can save a MIDI stream to internal memory for later playback, in which case it is called a sequencing controller, or sequencer. This is handy for a lightweight live performance without computer connection, where one can sequence the accompaniment ahead of time, then play live with accompaniment. For a DAW to address external synthesizers and MIDI controllers, it must have a MIDI interface via which MIDI messages can be sent and received.
The MIDI command set further offers instructions, called system exclusive (sysex) messages, specific to each synthesis engine, specifying values for parameters that uniquely generate each of its sounds, and commanding utility functions such as the loading and backing-up of the sound parameter sets.
General MIDI (GM) is a late-90s extension of the MIDI standard, specifying minimal capabilities for compliant synthesis engines, and standardizing available timbre types to a set of 16 chromatic (tunable) timbre descriptions plus a set of percussion timbres. Within a timbre type are defined eight specific instrument timbres. The advantage of GM is that a GM MIDI file can be sent to any GM-compliant synthesizer and a musical result will obtain, likely not far from the conception of the composer.
Those of us with pre-GM synthesizers (all in this studio are) need to categorize all our instrument sounds so we can easily find what we are looking for. Since GM, I use the GM classifications to categorize my instruments.
There is one main computer, which does duty as media server, DAW, photo and video editor, and the usual utilitarian functions. It hosts my digital music library (1000+ CD-quality albums), the source of most of my music listening enjoyment (excepting 25 or so multi-track, ultra-fidelity audio albums that I still play from the disc).
Our only piano is a 76-note keyboard MIDI controller/synthesizer/sequencer. It is used both as MIDI-controller and for keyboard practice/performance. Although the internal acoustic piano sound is quite good, I prefer the acoustic piano sounds provided by Owen’s generous gift of Synthogy Ivory, a set of quality piano samples together with a software sample player. It is triggered by MIDI output from the keyboard to the DAW.
I use the Bösendorfer Imperial Grand samples, and usually listen during practice via headphones connected to the computer (DAW). On the computer, I bring up eith the Ivory stand-alone sample player, or Apple Mainstage, which I have configured to use the Ivory samples in a Practice Piano setting. Once a sample player is running, I have a Bösendorfer at my disposal, albeit with a shortened synth-action keyboard.
I also play various other sounds to keep practice interesting. If I get better at playing, there are full 88-key, weighted-action, touch-sensitive MIDI keyboard controllers that would more accurately reproduce the experience of sitting at an acoustic piano. But right now, all my music fits into the 6+ octaves reachable by the synth keyboard.
The keyboard is pressure sensitive, offering zoned aftertouch, a 4-channel variant of channel aftertouch. Aftertouch allows the performer to change timbral response to notes, based on the pressure exerted on the keys after the key is struck. Further continuous controller functions can be provided by three modulation wheels, for introducing dynamic functions such as pitchbend, tremolo, and vibrato, and any other effects to which they are assignable in the MIDI setup.
The keyboard controller is a quality synthesizer in its own right, one of 9 stereo and 1 mono hardware synthesizers in the studio (dating from mid-80s to mid-90s, preceding the age of S/W synths and general MIDI). A comprehensive set of synthesis techniques are provided by these 10 modules: additive, subtractive, FM, in most cases combined with sample playback.
Studio Functionality and Audio Routing
ADAT is a digital communication format that packages eight digitized audio tracks into a single standard digital cable (but called a lightpipe when using ADAT packaging). Each synth feeds its audio output into one of the two 8 channel A/D converters (20-bit, 48kHz, 128x oversampling, 96dB SNR and dynamic range). Via a lightpipe connection, each ADC unit sends its 8 channels to the digital audio I/F (24-bit, 48kHz), and then via firewire to the computer.
Two of the synths have sampling capabilities. Six of the synths have onboard FX units for enriching their individual outputs; one of these has its FX unit accessible by up to 4 insert channels, to add FX to the outputs of any modules without their own FX units. Since investing in DAW software, the individual module outputs are all sent dry to the DAW, which offers a comprehensive set of digital FX for applying to channels individually, and then as master FX applied to the final mix. But the onboard synth FX are available if there is some effect best done there.
The digital audio I/F has a 32 channel input capacity, but only 16 will ever be used. We now live in the software age; the hardware studio has been in archival mode since year 2K. There are two retro-analog software synths running in the DAW (modernized and with general MIDI), bringing the studio total to 12 sound-generating modules. It is enough.
Given that there are potentially 19 simultaneous analog channels to be squeezed into a 16 channel digital interface, there is an analog line mixer thrown into the plan. ( I could also have purchased another 8-ch D/A unit, but I had the mixer already.) Those channels weakest in output volume are routed to the mixer, their sound is boosted to be compatible with the hotter signals in the rack, then mixed down into a stereo output before digitizing. The stereo output could also be routed as inserts to the accessible FX unit, which would mix the resulting wet signals with its own output.
Importance of Clocking Digital Interfaces
Internal clocks are needed to keep digital audio gear operating synchronously. All digital interfaces must employ bit clocks to guarantee signal integrity (no bits lost). The overall system uses sample clocks (word clocks) to coordinate digital sample streams arriving from different sources. One device with a word clock must be designated as the system master clock, and all other devices are synced (slaved) to it.
Normally, given equal quality of all available clock sources, one would want the A/D converter to be the system master clock. But I use two of them, so one would need to be slaved anyway. Further, the digital audio I/F cannot be slaved correctly at 48kHz due to a design glitch. But since this unit has the best clock implementation (JetPLL), little is lost by designating it as the studio master clock. This digital audio I/F sends its clock signal back to the ADCs, slaving them to its master clock.
The firewire I/F to the DAW employs no word clock and so operates independently of the system master clock. It packetizes the samples in chunks, so that the receiving device can unbuffer each sample chunk, forwarding each sample from the buffer according to its own clocked sample rate (called a reclocking interface).
A specific timbre within a synthesizer is determined by a set of parameter values unique to that synthesis engine. The sound parameter editor is a software program (app) that models and provides user access to sound-generating parameters, via MIDI sysex messages. (It is possible to program sounds for synthesizers directly from their front LCD displays and buttons, but resulting frustration and psychic pain generally acts as a powerful dissuader, particularly in complex synthesis modes such as FM.)
Because hardware synths are of the past, no effort at marketing general sound parameter editors is evident these days, other than those built into DAWs for their internal use. So we cling to our tools from the past. Here, we use the SoundDiver application (PowerPC only) to create new sounds for the ten hardware synths, to catalog all available sound parameter sets stored in these modules, and to back-up all user-created sound parameter sets. This all runs on a G4 PowerBook under Leopard OS, using the USB2 link to the studio MIDI I/Fs.
Sound parameter editing is sometimes useful to have available when laying down and mixing audio tracks. But studio hardware and software limitations make it the usual use case here to do such editing as a background task, independent of DAW operation. This has been sufficient for my needs.
In addition to sound parameter editing within the various connected synthesizers, there is another more elemental sound editing function available in DAW software, that of modifying the actual audio waveform within a sample or track, named a sample editor or audio file editor. When I first assembled a studio two decades ago, Alchemy and Sound Designer were the sample editors of choice; they still lurk on my OS9 machine. In recent years, the freeware Audacity and the DAW-bundled SoundTrack Pro were the go-to editors in the studio. The feature is now bundled internally in the DAW software.
Recording Live Performance
Live recording done outside the studio is supported on a PowerBook running GarageBand. Two high-quality cardioid condenser microphones (flat response 20Hz-20kHz) are connected to a 2-port phantom power supply. With appropriate adapters and cabling, the mics can be connected in quick and dirty fashion directly to the PowerBook microphone input mini-jack, a 16-bit interface. For any purpose beyond hobbyist demo recordings, a portable firewire interface (24-bit, 96kHz) seems the best bet for maintaining the fidelity of the mics.