Programming Sound Cards/Programming Sound Cards.iso/sound_38/lecture.doc

Text File | 1995-01-01 | 39.0 KB | 623 lines

    Sequencers, Midi and Music Software

    A "Sequencer" is defined as: a circuit, device, or software which records
    and stores performance information so that it may reproduce the
    performance; a sequence recorder. 
                                    the Rock Synthesizer Manual

    I. Sequencer History
    a).Analog sequencers
    Sequencers have been popular in electronic music since Dr. Robert Moog
    introduced the System 55 synthesizer in the late 60's. This type of
    sequencer was known as the analog sequencer, due to the setting of a
    number of continuously variable(analog) poten tiometers to derive pitch.

    In Dr. Moogs' original design, a bank of 24 analog potentiometers could be
    tuned individually, each producing a control voltage that was then patched
    to the sound producing circuitry of the synthesizer. A variable clock
    would step the sequencer from one potentiometer to the next, starting over
    again once it reached the last position. These voltages, when patched to
    the inputs of his voltage controlled oscillators, would produce varying
    pitches. A repeating sequence of notes resulted and hence the device came
    to be called a sequencer. This original sequencer had twenty-four pots, so
    you could have a maximum of twenty-four notes or eight three-note chords
    sequence over and over again. An example of an eight note sequence is the
    droning pattern heard in i nk Floyds' "Dark Side of the Moon" recording of
    some years back. Although other companies introduced stand alone versions
    of analog sequencers about that time as well, Moogs' design is usually
    thought of when the term is mentioned.

    b). Digital sequencers 
    About 1975, Oberheim electronics introduced the DS-2, a "Digital"
    sequencer capable of remembering one-hundred and forty-four notes.
    Although it was only a monophonic device, the attraction to this type of
    sequencer was that notes could be entered in both "real time" or "step
    time" from the synthesizer keyboard. This was of great interest to
    composers who would prefer to play notes in from the piano keyboard,
    rather than having to painstakingly "tune in" one note at a time with a
    potentiometer. This unit had the capacity for three sequences that could
    either be played back independently or chained together to play in turn,
    but there was no means of permanent storage for anything loaded into the
    device. Since all information was stored in a "volatile" R A M memory,
    everything was erased when power was removed.

    Shortly thereafter, Roland, a young company known mostly for their drum
    machines, introduced a digital sequencer known as the MC-8 Microcomposer.
    This was the first of many automated devices from the Roland company and
    was considered to be clearly ahead of it's time. This device allowed
    precise entry, editing and playback of a multiple of parameters including;
    control voltage, step time, gate time, portamento and volume or modulation
    settings. It was one of the first devices to implement tape
    storage/backup-up functions, which is on nearly every
    synthesizer/sequencer on the market today.   Although it could be accessed
    from an external synthesizer for entry, manual step time entry was the
    preferred method. It had an extremely large memory for that perio in time,
    1200 notes, with an optional 16k upgrade that allowed over 5000 notes.
    Designed for use with Rolands' modular studio synthesizer, the MC-8 was a
    giant step forward for electronic music. An example of the precise step
    entry method with the later version MC-4 is the complex synthesizer solo
    of Totos' hit song, "Rosanna".  

    Using the control voltage scheme, polyphony was difficult to achieve
    without the use of a modular synthesizer containing several oscillators.
    Given even that, composers could enter only one single note passage at a
    time. 

    c.) System Common
    What was needed was a system with which a composer could play polyphonic
    parts into the sequencer from the synthesizer and have it play back
    verbatim. Determined to solve this problem, Oberheim Electronics
    introduced in 1979 a new sequencer designed to w ork exclusively with
    their OB-8 synthesizer, the DSX. The DSX allowed eight note polyphony and
    full synchronization with their new drum machine, the DMX. This was a
    great breakthrough,  and together these units would allow a composer to
    generate fluid a d complex musical passages that could be edited, looped,
    quantised, cut and pasted and heard with percussion. Soon thereafter,
    Sequential Circuits followed suit and introduced a polyphonic digital
    sequencer for their Prophet V and Prophet 10 synthesizers. This unit was
    the first to utilize a disk drive for storage of song files. a

    II. MIDI
    Manufacturers were now slowly responding to performers' requests to find
    ways allowing instruments of one brand to communicate key information with
    others of different brands. The control voltage technique was
    unsatisfactory because a separate communication line was required for
    each notes' information. Also, there was no universally satisfactory
    standard for this data transmission among the various manufacturers. 

    Digitally controlled oscillators were just beginning to appear in units
    such as the Roland Jupiter 8 and the Korg Poly Six. Since key information
    was transmitted in the form of a number instead of a voltage, more
    efficient handling of the data was possible. If all manufacturers were to
    agree on the type, rate and size of the transmitted data for each key
    depression, theoretically any keyboard could trigger any other. So, in
    1983, with mutual cooperation between Roland, Sequential Circuits and
    Yamaha, the MIDI specification was born. The Midi spec provides for the
    transmission of a number of performance parameters including: note-on,
    note-off, velocity, aftertouch(pressure), and controller data, such as:
    pitch bend, modulation wheel, sustain pedal and others yet to be defined.
    The original intention was merely to decide on a standard that would
    permit transmission of data from one synthesizer to the next, regardless
    of brand. Also provided for was a channel system to permit remote
    selection of instruments. It was soon realized, however, that since this
    data was in a form that could easily be handled by a computer, great
    amounts of musical information could be stored and manipulated with a
    microprocessor and RAM memory. 

    III. Midi Recorders/Dedicated Computers

    The term "sequencer"  was now rapidly becoming out of date, since users
    were no longer restricted to a small amount of notes repeating over and
    over in a pattern. This new type of device would be dubbed a "Midi
    recorder". Manufacturers rapidly began to design hardware sequencers(or
    dedicated computers) to accompany their synthesizers, encouraging the user
    to begin forming a Midi instrument network. One of the first MIDI
    recorder/dedicated computers to appear was the Roland MSQ-700. This unit
    could store over 6500 notes for up to 16 channels in its internal memory.
    It had eight multi-channel tracks which could be recorded in real or step
    time. These tracks could be linked together in up to 73 repetitions in
    order to form a song. Also by Roland was the MSQ-100, a lesser expensive
    but very capable Midi recorder. Since that time Yamaha and Korg have
    introduced hardware sequencers. A tape storage back-up interface was also
    provided for these units to permanently store song data for later
    retrieval. The advantage of a dedicated computer such as this is the
    convenience of portability and ease of use. Also of mention is the
    inherent ruggedness of a 'made for portability' unit. The disadvantage of
    these devices is the limitation of  expandability o f memory and
    functions.

    IV. Computers and Midi 

    The idea of using a micro or personal computer for the task of recording,
    playback and editing of Midi events has recently become very much in vogue
    due to the general public awareness of the personal computers' value in
    other tasks. Musicians cannot rationally ignore the vast potential the
    personal computer offers and sees the opportunity to combine a variety of
    applications in one machine. Personal computers come in several different
    types and with a varying amount of user memory with each. In order to make
    an intelligent decision on which computer to purchase a user must decide
    what software is currently available for the tasks to be performed and
    what price point will be within his/her means.

    Among the least expensive, but well suited to Midi applications is he
    Commodore 64 and 128 computers. While still lacking the power of more
    expensive units, this computer has quite an admirable collection of Midi
    software available for it. Although 64k of memory was the maximum
    available for many personal computers of the seventies, the Midi data
    format only allows about 5000 events to be recorded into this amount. Next
    in apparent popularity was the Apple II computer. Able to display an 80
    column screen and with expansion slots for adding accessories, this
    computer was a big seller to many schools and individuals interested in
    personal computing. An equal amount of Midi software is available for the
    Apple II as is the Commodore. Although the Apple II is a much more
    expensive computer, the Midi software for it remains on about the same
    level as the Commodore. Moving up through the ranks and into the eighties,
    the introduction of the Apple Macintosh showed much promise as a musicians
    computer as well as a business machine. Even so, software for Midi and
    business applications was slow in coming for the Mac. Present day
    applications have proven that the Mac is indeed one of the best for use
    with Midi.

    The IBM-PC had established itself in the early eighties as the machine of
    choice for inexpensive business computers and was appearing on the desks
    and in the homes of thousands of middle class information workers. This
    computer is only now(1986) beginning to show promise as a musician's
    computer, largely due to the fact that an enormous amount of work-alike
    "clones" are appearing on the market at a price that is well within the
    budgets of serious music hobbyists and performing musicians. Since (1.)
    the bulk of modern business software has been geared for use on the IBM
    and compatibles, and (2.) the wealth of peripheral equipment easily
    attached to the IBM is expanding at at an alarming rate, music software
    developers have naturally turned to this unit as a "standard" for which a
    very large market is already established. 

    Two new entries in the marketplace are showing much promise; the Commodore
    Amiga and the Atari 520 and 1040ST. These computers are of the 32 bit
    variety like the Macintosh, and are in a very affordable price range. The
    software development for these machines is being approached very
    cautiously as each companys' future is dependant on their success.
    Developers are wary of investing the thousands of hours necessary to
    develop Midi software for a machine that has an uncertain future. As of
    this writing, to my knowledge, two companies have music software on the
    market for the ST and only one for the Amiga. This is very disappointing,
    especially in light of the fact that the Atari actually comes equipped
    with a Midi interface as standard equipment. Both machines share the same
    main processor as the Apple Macintosh for which there is a very nice
    software catalog to choose from. 

V. Music Software

    Even though a special purpose computer designed exclusively for Midi
    recording tends to be easier to operate, the flexibility and open
    architecture of a personal computer could far outweigh that single
    advantage. With the ability to provide full screen graphic displays of
    Midi events and even musical notation for editing, not to mention the
    storage capacity, the personal computer is a natural choice for
    professional and novice composers alike.

    Even so, software writers who understand the musician's needs are very
    scarce, so it is only natural to see the development of powerful, easy to
    use music software as a slow process, growing more sophisticated as the
    writers and the market mature. a

    a.) Software sequencers

    One of first companies to produce music software was Passport Designs.
    Their first offering grew from the experience gained with their non-Midi
    Soundchaser synthesizer/software system they had developed for the Apple
    II in the early eighties. Passports' first package was called MIDI 4, and
    was developed for the Apple //+,//e and Commodore 64.  This software
    allowed 4 channels of polyphonic Midi recording/playback, but with a bare
    minimum of editing capability. Each track could be looped continuously
    and manually punched-in and out for recording, but single note editing was
    yet to be implemented. Tracks could be merged together, but the resulting
    track could only be sent out on one Midi channel. This meant that only a
    total of 4 Midi instruments could be played at a time, a serious
    restriction in that Midi allowed for up to 16 discreet channels. Although
    not as full featured as many musicians would hope for, the feedback that
    Passport and other would-be Midi developers received from this pioneer
    product prompted more research into the possibilities of more powerful
    music software.
    
    Another company that had introduced music software about that time was
    Cherry Lane Technologies, previously known for their large catalog of
    sheet music and books. The packages they were offering were independently
    developed but nationally distributed by Cherry Lane. Among these was
    Texture, by Roger Powell. Roger had gained a lot of experience in the
    sequencer field by working with the band Utopia and it's leader Todd
    Rundgren.  Realizing that most songs he played were comprised of
    reoccurring verses and choruses, Roger wrote this program to implement
    programmed "links" or "phrases" that could be chained together to form an
    entire song. This was a very useful feature and allowed more music to be
    played back from less available RAM memory. To elaborate on this method of
    song composition, software writers borrowed Rolands' idea (used originally
    on the TR-808 drum 'computer') of saving memory by "calling" a previously
    recorded sequence instead of having the computer remember all of the notes
    used in each chorus or verse as it progressed, effectively having the
    original sequence repeat when that part of the song came around again. In
    Rogers' program, each of these "links" could contain a number of Midi
    channels, so entire songs could be broken down into their basic
    component parts; i.e. Intro, Verse, Chorus, Turnaround, Bridge, Ending,
    etc. This method allowed more effective use of computer memory and saved
    redundant passages from having to be re-recorded.

    To elaborate on this method of song composition, software writers
    borrowed Rolands' idea (used originally on the TR-808 drum 'computer') of
    saving memory by "calling" a previously recorded sequence instead of
    having the computer remember all of the notes used in each chorus or verse
    as it progressed, effectively having the original sequence repeat when
    that part of the song came around again. In Rogers' program, each of these
    "links" could contain a number of Midi channels, so entire songs could be
    broken down into their basic component parts; i.e. Intro, Verse, Chorus,
    Turnaround, Bridge, Ending, etc. This method allowed more effective use of
    computer memory and saved redundant passages from having to be
    re-recorded. Other companies such as Syntech, Dr.T's, Sonus, and others
    have produced excellent software for the Apple and Commodore line using
    the 'smart-looping' techniques. On the other hand, computers with very
    large memory capacity like the IBM and Macintosh didn't have to worry
    about economy in the storage of notes and use a 'straight-line' recording
    method. If a previously recorded section needed to be repeated, the user
    would "copy" that section and "paste" in in the track at the appropriate
    time. An example of a very user-friendly program that uses this method is
    Voyetra Technologies' "Sequencer Plus" program for the IBM. It displays
    the command options in a below-screen menu. Using the first letter of a
    command as the command itself proved to be a very intuitive and
    non-redundant method of issuing the various instructions. The many
    variables in this program are all assigned by either the plus and minus
    keys or wit the greater-than or less-than keys. This program continues to
    be one of most popular programs for the IBM today, due to its ease of use.
    a

    b.) Sequencer entry methods 

    Passports' Midi 4, Roger Powells' Texture sequencers and others relied on
    real-time entry of notes. You either had to have your parts well rehearsed
    before you began or deliberately slow down the tempo of the computers'
    metronome in order to play parts in at a more relaxed pace. The other
    popular type of entry method is "step-time" entry and is more commonly
    used in hardware sequencers. Step-time recording does offer much more
    control over the entering of very difficult to play passages, allowing the
    electronic musician to enter passages that would be impossible for any
    human to play. Step time entry does not rely on timing to be set by the
    length of time a note was held down, rather the timing is pre-set and only
    note position is needed to be entered by the performer. This allows non
    keyboard players to easily construct sequences of perfect timing that
    sound very smooth and even whe played back at normal speed. Each type of
    entry method has its merits.     


    c.) Notation Software  

    Professional Composer, for the Macintosh is a non-Midi music notation
    program. It depends on input from the mouse pointing and clicking each
    note on a grand staff. This program is the most complete for music writing
    at this time. It is capable is scorin g the most complex musical passages,
    providing all of the general and many exotic music notation symbols, as
    well as allowing the user to create his own. Full composers' scores can be
    written in any format and single parts can be extracted in each
    instruments' natural key and range. Text in any Macintosh font may be
    inserted and even background graphics such as pictures can be inserted.
    Although the music entered can be played back either from the Macintosh'
    built-in speaker or external audio out, the qality of sound is very
    limited and restricted to only four notes at time. Rolands' MPS on the
    other hand is a Midi based program and includes both sequencing and
    notation capabilities. MPS was met with much resistance due to its method
    of user interface. MPS is controlled exclusively by the IBM's ten function
    keys, labeled F1 through F10. This has proven to be a very unfriendly and
    un-intuitive method of user interface because in one menu a given  F key
    would perform one task, and on the next menu, the same F key would perform
    an entirely different function. Musicians apparently desire a single
    letter command for each function that will remain constant throughout the
    program to perform the various tasks of recording, editing, and playback.
    This program also faltered in it's notation aspects in that it could only
    display a single measure of music at a time. 

    Other integrated software for the IBM includes Jim Miller's Personal
    Composer program. It is the most popular notation software for the IBM and
    has a fairly good sequencer built-in. After music or computer keyboard
    entry, the full-page music notation may be viewed and edited. Users may
    also design their own symbols in addition to the full complement of
    symbols provided. Although the music will not "scroll" across the screen
    as it is being played, page up and down commands may be used to follow
    along with the score.



    We should now distinguish between the three different types of music
    software we have touched on up to this point. 

         1.) Sequencing software
         2.) Notation software
         3.) Integrated software 

    1.) Sequencing software generally allows the user the most  flexibility in
    the type of input and in the editing of parts of the program. 

    2.) Notation software is designed to provide automated print-outs of music
    notation entered from the computer keyboard or mouse. 

    3.) Integrated programs, such as Jim Millers' Personal Composer and
    Rolands' MPS integrate portions of each type into one program. Since a
    great amount of code is required in these programs in order to do both
    jobs, compromises were made to both the sequencing and notation sections
    resulting in functional limitations. A way around this problem is to
    design each program to store their files in the same format so that each
    can read the others'.

    A successful example of this concept is used by Mark of the Unicorn. They
    have split the functions completely apart and use a separate piece of
    software to handle each job.  The sequencer companion to Professional
    Composer, named Performer, is a sequencer with amazing editing and note
    management routines, allowing over five hundred tracks containing up
    150,000 events of Midi information to be edited and played back to the
    resolution of a single 128th note.  Complete Midi event editing is
    possible with this program. After composition and editing is complete, the
    tracks can be re-arranged to the format of a conductors' score and then
    saved to disk in a form that can subsequently be read by the notation
    softw are. This two step method will probably be successfully merged into
    one program as the price of memory becomes cheaper and processors become
    faster. 

    d.) Midi interfaces and Syncronization

    Through necessity, Passport developed the first Midi interface. These
    first interfaces worked with the Apple // series and Commodore and
    utilized one Midi in, one Midi out, and a 5-pin sync jack for use with
    pre-Midi drum machines. This jack provided a 24 pulse-per-quarter-note
    (PPQ) output or would receive same if the software was set to recognize
    it. Another type of interface was required for Texture, Roland's Apple
    software, and all IBM software. The interface, known as the MPU-401, is
    manufactured by Roland for the Apple IIe, Commodore 64 and IBM personal
    computers. It is termed a "smart" processor because it handles all of the
    Midi data within itself and frees the computers' processor of these duties
    in order to increase computer speed and allow more graphics. This
    interface provides one midi input, two midi outputs and offers tape sync
    as a standard feature. 

    Syncronization with a rhythm device is important if the music sequenced is
    to have any electronic percussion score within it. Since only non-Midi
    drum machines such as the Roland TR-606, TR-808, Oberheim DMX, Emu
    Drumulator and the LinnDrum existed at this time, some method of
    syncronizing their clocks with a Midi sequencer was necessary. An
    important point to consider here was that a still un-standardized PPQ
    clock format existed among these manufacturers. Midi sync of 24 PPQ was
    being implemented on the newer machines, but Sync clock formats of
    24,48,96 and higher PPQ formats were used in these drum machines. Reliable
    methods of syncronizing these machines with the computer constantly
    baffled pioneers of Midi software and hardware. It bears mention here
    that now prominent manufacturers such as J.L Cooper and Garfield
    Electronics began their companies as "black box" interface designers for
    the custom applications of touring bands attempting to solve such
    interface problems. 
        An optional port was now being offered by Passport
    on their interfaces known as "tape sync". By taking the audio output from
    this jack and recording it onto tape, a sync "tone" was created. This
    tone, when routed back into the interface could control the sequencers'
    start-stop functions and maintain syncronization over time. With a
    multi-track tape recorder, a user could compose and edit tracks on the
    sequencer, and sequentially transfer them one by one to the recorder in
    perfect sync. The advantage to this method was in that an owner of only
    one synthesizer could produce a tape with the sounds of many synthesizers.
    This method is used quite frequently in recording studios to drive banks
    of synthesizers in real time, while saving tape tracks for non-midi
    instruments such as vocals and guitars. Although this interface has become
    the standard Midi interface for the IBM series, very few manufacturers for
    the smaller computers have adopted that interface, due to it's high cost
    compared to the Passport interface.

    e.) Voice Librarian software

        One of the functions of Midi synthesizers that can easily be taken
    over by computer is that of the digital storage of their parameter
    settings, or "patches". By virtue of a portion of the Midi specification
    known as "System Exclusive messages", each manu facturer is allowed to
    specify the codes by which each model of their synthesizer can communicate
    with another of the same brand and model. This was originally intended to
    allow an owner of two or more of one model synthesizer to operate the
    controls of one or more of the the same model remotely, i.e., change
    control lever "a" and it will change control lever "a" on all of that type
    connected in the chain, Midi being the networking buss. This is a nice
    idea, but few musicians own duplicate instruments in their setup.

         What it has proven useful for however, is in the bulk transfer of
    parameter settings of a particular model to another of the same model.
    What the original engineers of the Midi spec didn't expect though, was the
    use of a computer to emulate the commands of that second instrument in
    order to direct the first to dump it's parameter setting onto a floppy
    disk instead of into another synthesizer. With the proper software, a
    computer owner could effectively bypass the use of a cartridge or tape
    backup system to store an unlimited number of parameter settings (sounds).
    This has become a very important and cost effective use of the computer in
    a Midi setup.

         Among the first to utilize this feature was Computers and Music, Inc.
    They researched the codes necessary to request system exclusive dumps from
    the Yamaha DX-7 and wrote the program now licensed to Passport Designs,
    Inc. The success of this program has encouraged them to write similar
    programs for other popular synthesizers. On the horizon are programs
    designed to contain the codes for all current synthesizers that will write
    in file formats to be compatible with all popular personal computers. This
    will solve the problem of owners of different computers attempting to
    share data from like synthesizers, but using different librarian software.
    Currently, librarians exist for all personal computers, with only one
    sharing a common file format, this being SYSEX by Key Clique. the drawback
    of this program is that is only acts as a 'bulk dump' utility, not
    allowing individual patches to be transferred between banks.

         The use of a system exclusive software based librarian has several
    advantages over the common method of cassette of cartridge based storage
    methods. First, the transfer of information is instantaneous, avoiding the
    hassle of level setting to a recorder, which in itself can be a major
    endevor. Second, (especially evident in a cartridge system) the amount of
    patches storable on the medium verses cost is substantially reduced with a
    floppy or hard disk destination verses a limited cartridge destination.
    Virtually unlimited resources are availa ble on the hard drive when
    compared to the very limited capacity of a standard RAM cartridge. The
    only advantage to the cartridge method is in it's portability. This idea
    is voided if a software based sequencer is to be employed in the same
    system.

         Third, the inherent full screen display of all of the voices in a
    given bank is of great advantage, allowing you to view as much as 64
    sounds at a time in their prospective order. By virtue of the computers'
    memory, several banks can be viewed at a time, allowing a mixing and
    matching scenario between the various banks of the synthesizer, through
    which you may create custom banks of selected sounds on a one-by-one
    basis. Although most modern synthesizers have the ability to send their
    voice parameter information through Midi, it is still very confusing to
    the receiving computer if all synthesizers in the system transmit the same
    command at once and on the same channel. In a large system, these commands
    will often be issued on a particular channel, accessing a single
    synthesizer module at a time.

    f.) Patch Editing software

        Patch editors are another animal indeed, as these programs are
    designed to allow the user full access to every parameter of their
    synthesizer from the computer, using system exclusive messages. This
    software has made the programming of synthesizers such the Yamaha DX-7
    much easier as you can work with a full screen display of the information
    instead of a the small window provided on the synthesizer. An excellent
    example of this type of software is Opcode System's DX/TX Editor. A
    computer patch editor of this type will free the fledgling synthesizer
    programmer from the tedious chore of single lever digital editing system
    of todays' synthesizer.

         To elaborate on that concept, once upon a time, a long time ago
    (about three years now), synthesizers had a variable lever or knob for
    every control function on the machine. Due to technological advances and
    price cutting pressures, the number of actual control devices was cut to
    one, since this one control could be 'assigned' to any function
    individually. The user would select function '88', "sustain level", for
    instance, and by changing the level of that control, the sustain level
    would vary accordigly. Next, the user would perhaps select function '72',
    modulation level, and continue to use the same control to change the
    various functions of the entire synthesizer until the desired sound was
    realized.  While achieving the goal of simplicity in design and
    manufacture, this method has produced the unpopular side effect known as
    "user-unfriendliness" and has put a damper on the enthusiasm of many a
    prospective synth programmer. Voice editors have helped spark the desire
    of experimentation to synth programmers and have instilled the sense of
    being "in control" to new users of the otherwise confusing music
    synthesizer.

    g.) Sound sampling software

         This is the newest and possibly most exciting development in music
    software since Midi was conceived. For the first time, it has become
    possible to plot and graph an actual sound, study it, change any portion,
    and to hear the result to confirm the expected result. One of the most
    difficult of all senses to understand, the acoustic phenomena, can now be
    broken down into it's most basic parts, analyzed, changed and
    reconstructed into a predictable whole.

          The software I am referring to is the Sound Designer series from
    Digidesign, Inc. It is written to be used with the Apple Macintosh
    computer in conjunction with digital sound sampling instruments by E-mu,
    Sequential Circuits and Ensoniq, respectably. I n effect, the software
    does not create the actual sound, it is the means by which the user can
    examine and modify an already "sampled" or "captured" sound. These
    instruments make a limited digital recording of an actual sound that is
    then stored in on-bard memory. Through the magic of Midi, this digital
    information can be transferred to the computer for detailed display.
    Although very complicated to look at in it's raw form, the soundwave can
    be reduced to a single vibration within thousands. The level to which
    these programs can display sound data was previous only able to be
    approximated by other analytical means. This advancement means that
    studies in the nature of sound and the subtleties that differentiate one
    sound from another can be concretely plotted, with meaningful results.

    h.)Sound Modeling

          The latest development in the field of on-screen sound research is a
    new program called "SoftSynth", also by Digidesign. This program is
    designed to be a tool by which a user can develop new sounds from scratch
    using the additive synthesis method. This method of synthesis utilizes a
    number of sine waves with different frequencies and amplitudes added
    collectively to form a composite waveform. An existing instrument which
    uses this method of sound generation is the Hammond organ with drawbars.
    Each draw bar represents a different harmonic, and when each of the
    various harmonics are added together in different degrees many complex
    waves are possible, some of which resemble other instruments. The
    SoftSynth program for the Macintosh, however, allows much more complex
    additions of sine waves, as the user is allowed to specify the center
    frequency and amplitude envelope of each harmonic, up to 32. Since it is
    possible to plot any harmonics' amplitude through time, Fourier analysis
    may be studied in real time and manipulated in order to better understand
    how sound waves interact to produce sound of a given timbre.

     VI. Live performance

         Live performance utilizing Midi has to be approached carefully. Let
    us consider what we are attempting to replace by casting a Midi sound
    module in the place of a musician. When a musician plays, attention is
    focused to that person. There is action throu ghout the performance and
    the audience can generally identify the sounds they hear with the musician
    creating it. In the case of the Midi band, however, much of that action is
    missing. Even though the execution and technique may be perfectly
    programmed , the slight human flaws that are inevitably included as part
    of the live artists' performance are missing. That's what's wrong if
    you've tried this approach and have gotten less than a standing ovation.
    You cannot create a precise performance, have it executed by Midi modules
    and expect to bring the house down when played. You have created no more
    than an expensive "juke box" in this solo approach.

          What can be done however, is to accompany the real performers on
    stage with selected Midi modules and effects. I cannot overemphasize the
    possibility of having a well orchestrated performance augmented by one or
    more Midi instruments as live musicians fill out the rhythm section . This
    is the application of Midi that was originally intended. As we are
    continually amazed by the proliferation of Midi controlled devices, we
    must  decide which ones can be used expand our own abilities rather than
    trying to replace them. 

    VII. Midi effects

         Since Midi information is used to direct a network of instruments as
    to which note to play, how long, which sound to switch to, etc., then it
    seemed feasible to include a Midi jack on effects devices in order to have
    them automatically switch through their functions also. Even though
    effects are not instruments in themselves, when used with musical
    instruments and/or voices they can enhance an otherwise dull performance
    or bring realism to the most lifeless synthesizer sound. Examples of
    effects that can be effectively Midi'd are digital delays, digital
    reverbs, mixing and lighting consoles. Through creative use of these
    "instruments", effects never possible by human adjustment are easily
    attainable and repeatable.

    VIII. Education

         Not only does Midi make the realization of music infinitely easier to
    produce after its original conception, it also allows beginners to learn
    music and progress at their own rate without the constant supervision and
    criticism of a human teacher. Modern, interactive, music software has
    begun to win the affection of youngsters eager to learn music but are
    apprehensive of having yet another teacher standing over them. Music will
    always be difficult to master, but no one says it shouldn't be fun to
    learn at the same time. With the aid of the computer, very  logical and
    thorough programs of music training can be taught, repeatably,
    consistently and with every bit as much interaction as could be realized
    with a human teacher.

          The Future May consider the advent of Midi control to be the start
    of a revolution in the musical community. Midi allows a single performer
    to easily realize the sound of anything from a small ensemble to an entire
    orchestra without the expense and human emotions usually associated with
    groups of players. The only limiting factor in the midi composers' pursuit
    of his music is imagination (and his budget). Musicians and the concept of
    group playing will never become obsolete with Midi, as is often feared.
    Nothing can match the feelings  of spontaneity between performing
    musicians, nor can machines ever exude the warmth and feeling that a
    talented player can coax from his instrument. Midi is a tool, a
    multifaceted one. It must be used to enhance the talent that works with
    it, not replace it. Through creative use of the many wonderful things that
    are now possible with this digital standard, music can move on to the next
    generation of expression, allowing us to experience musical textures that
    before could only be imagined, and further on to musical frontiers that
    today could not even be conceived. 

                                   Gary L. Osteen