MIDI THEORY
Components of a sound wave:
- Molecules at rest (normal pressure)
- Compression (maximum pressure)
- Rarefaction (minimum pressure)
The characteristics of a waveform:
- Frequency: Number of times per second a waveform completes it’s cycle. Frequency is measured in cycles per second or hertz (HZ). Ex: Kilo- 1000, 1khz- 1000hz, 15khz- 15000hz
- Cycle: (sound) A complete change from normal pressure to high pressure to low pressure and back to normal again. (signal) A complete change from zero volts to positive to negative and back to zero.
- Octaves: An octave is the doubling or halving of frequency. From A1 to A2 is exactly double the value, so low B on a bass would be 43hz while the octave would be 86hz. Humans have a 10 octave range. On a synthesizer, frequency is controlled by the tuning of the oscillator (wave form generator) and by the keyboard or other oscillator. An Oscillator is a waveform generator found in mixers and used for tape alignment.
- Velocity: The speed at which a waveform travels through a medium. 1khz tone at sea level 70 degrees F travels at 1130 feet per second. Lows travel slightly faster than highs.
- Wavelength: The distance it takes for a waveform to complete it’s cycle. Wavelengths decrease as the frequencies rise. Low frequencies have longer waves and higher frequencies have shorter waves.
- Amplitude: The amount of change in sound or signal intensity. For sound- the amount from normal pressure to peak hi or low pressure. For signal- the amount from zero volts to peak positive or negative. Devices that can control amplitude are amps and faders.
- Harmonic Content: Also known as tone color or timbre. The relative number and amplitude of overtones contained in a waveform. Sine waves are simple waveforms containing no a single frequency and no overtones while Complex waveforms are a combination of tones.
- Fundamental Frequencies: Usually the lowest and loudest frequency contained in a musical tone. Used to identify pitch.
- Overtones: Any frequency contained in a musical tone that is higher than the tone’s fundamental. Changes the perception of tone.
- Harmonics: Special overtones whose frequency is a whole number multiple of the fundamental. As a performer plays louder, the amplitudes increase more than the fundamental frequency does, causing a tonal change. More harmonics make the sound brighter and louder.
- Envelope: A description or control of how a waveform’s characteristics (frequency, tone, amplitude) change over time. A synthesizer’s are controlled by an envelope generator. Amplitude envelope can be controlled via compressors, limiters, expanders, and gates.
- Phase: The time relationship of a waveform to a known time source. Phase is measured in degrees. Absolute phase when compared to a zero degree reference point, and relative phase when compared to another waveform. Waveforms that are in phase (0degrees difference) at any given time will reinforce each other. Waveforms that are out of phase at any given time will cause partial or complete cancellation of one another. Phase can be controlled by phase reversal switches on your pre-amp or moving the mikes in relation to each other from the sound source.
Samplers: A digital recording and playback device that records to ram and has real-time control of pitch.
Random access memory (ram): Allows data to be both written and read from it’s memory store. Allows almost instant access. The amount of ram available determines available recording time along with it’s sample rate.
Pitch can be changed almost instantly, but this change is another characteristic of the sound. The duration or length of the sound is affected. The compression or expansion changes the sound duration without changing the pitch.
Multi-sampling: Sampler can record and playback several samples and each sample may be assigned a key or group of keys.
*The combined length of all samples cannot exceed the total amount of ram. Ram is volatile memory. It erases upon powering down. Some samplers have internal hard drives for storage but even these may fail. For this reason external storage is preferable.
Memory: Since any sampler’s sampling time is limited, we may preserve memory by looping. This plays all or part of a sound repeatedly, conserving memory yet producing long sustained tones or rhythms.
A sample from a copywritten source should first be cleared. A sample clearing house determines who owns the rights to the sample and can request permission for it’s use. Generally a fee and royalties are requested , the amount being determined by the popularity of the sampled performer, the length of the sample, and how often it is used in the composition. Some artists and film companies will refuse permission if a sample is altered beyond recognition and in some cases it constitutes fair use and is not liable to a lawsuit.
Truncate or trim: Removing an unwanted part from a sample’s beginning or end.
Fly-in: Sampling a performance from multitrack tape and recording that performance in a different part of the song than where it was originally performed. This allows a performance to be duplicated and saves time. It also allows re-arranging of the song on the fly.
Drum triggering or replacement: Drum track on tape sent to trigger input on sampler (or an alternative drum triggering device). If adjusted properly, each time the drum on tape plays, the sampler acts as though its key was struck and plays a selected drum sound. This can than be recorded on another track. Excessive leakage on drum tracks can cause false triggering. This technique will not work on overhead tracks, nor on an instrument playing different sounds such as open and closed high hats or different pitches. The sampler cannot determine which sound to play as it operates on threshold ratio. It may also not trigger accurately on snare rolls or subtle snare work. Some sampler trigger inputs respond too slowly, thus producing an audible delay such as a flam or doubled hit when combined with the original track, or the leakage from the original drum on other tracks such as the overheads. This will also affect the song’s timing. For this reason it is best to use a faster responding device. You could also delay the other tracks by same amount of the triggered sound. Another technique would be to dump the sound into a computer based audio editor and change the start position of the track to be early by the appropriate amount of time.
Leash lines: Snake that has one cable per channel.
MIDI: Musical Instrument Digital Interface. A means of communication between midi capable devices. Midi’s initial purpose was to allow one keyboard to control others.
Digital Interface: A connection for the exchange of data.
Basic Midi commands:
- Note-on: Which key is played, each note given its own code. There are 128 midi notes.
- Note-off: Which key is released.
- All notes off: Sends note off to all devices, all keys. Also called panic button.
- Velocity: When the key is struck quicker, the sound becomes louder and brighter. Works on most instruments.
- Aftertouch: While holding key, pressing harder will change sound. Can affect pitch bend, modulation, panning, filtering, etc.
Midi Controller Messages:
A midi device may have up to 128 functions controlled by midi. Some control messages are designated for specific functions, some are left to the manufacturer to define.
Two Types of Midi Controllers:
- On-Off: Generating one of two possible commands. Ex: Sustain Pedal.
- Continuous: Generating 128 sends or variations. Ex: modulation, breath, pitch bend.
Program Change: Activates a preset or system setup (there are 128 available) on any midi capable equipment, including effects units.
Sysex or system exclusive: Data only understood by one particular device or devices of the same manufacture. Ex: Program and patch changes.
Midi Clock/Song Pointer/MTC:
Midi Machine Control: Allows midi devices to control machines such as a tape deck. Used for sync of multitrack recorders such as the Adats and BRC.
Midi data carried is carried on cable using 5 pin din connector, although on the 3 center pins are used.
Midi Ports: Connecting ports on midi capable equipment.
- In: Receives data
- Out: Transmits data
- Thru: Passes data received at the devices in port.
The midi thru port causes a delay of two microseconds, or two millionths of a second. This is not enough of a delay to be heard. Any audible delay can be blamed on the receiving device’s processor responding slowly, not midi itself.
Midi connectors do not carry audio signal. They carry control data to other midi devices. Midi systems( synths, modules) must have audio outputs connected to the mixer in order to be heard.
Midi modules, sound modules, rack modules, synths or samplers with no built in means of control must be controlled thru midi. Midi controllers may be devices other than keyboard controllers.
Midi data is transmitted on 16 channels over a single cable. Each midi device may be set to receive data on any of the 16 channels.
Master controllers may transmit on a single channel or have yhier keyboard split up to transmit on several channels at once.
Sequence: A chain of events. Music performances are sequences.
Midi Sequencer: Records, plays back and edits or reorganizes midi events. It does not record audio. Midi data is to the sequencer like sheet music is for the performer.
Sequencer and hard disk recorder: Some MIDI sequencers are coupled with another function, a hard disk recorder that can record audio, but a device that is a dedicated MIDI sequencer cannot.
Prior to midi sequencers, all performers had to be in the studio to record their projects.
Today a performance may be sequenced outside of the studio.
Midi sequencers are multi-tracks. If we want a track’s data o be sent to a different MIDI instrument, that track must be assigned to transmit on a specific midi channel and the instrument intended to receive the data must be set to the corresponding midi channel.
Multi-timbral: An instrument that can play more than one program at a time with each program being assigned it’d own midi channel. Limited in how many notes it can play, simultaneously limited to number of available voices or polyphony (oscillators). Some complex sounds require more than one voice per note.
Hardware sequencer: A computer made to function only as a midi sequencer. Generally, it has dedicated controls for sequencing and a built in MIDI interface.
Software sequencers: A program that allows a specific type of computer to function as a MIDI sequencer.
MIDI Workstations: A multi-timbral synth or sampler with a built in sequencer. Very popular due to affordability.
Workstations compared to software:
Workstations:
Smaller and less fragile
Dedicated controls
Limitations in memory
Limitations in waveform editing
May not require software updates
Sequencer only
Built in midi interface
Smaller display
Software:
Larger unless laptop
Uses keyboard and mouse
Expandable memory
More detailed editing of waveforms
May be upgradeable via updates
May run other programs
Interface must be added
Larger display
Synths or samplers can be added to computer as software or hardware.
Tempo: Rate or speed of performance. Internal clock regulates tempo.
Metronome or click: An audible indication of tempo or beats per minute (BPM).
Sequencers arrange music by bar (measure), beat and clock tick. A clock tick is a smaller division than a beat.
Unlike a tape recorder, we may overdub on the same track which we previously recorded. No magnetic particles are being used to record a waveform, so we can simply add more data without erasing previously existing data.
Looping or cycling: Continuous play/record between two points.
Quantization: Shifts midi event to closest available memory position. Available positions are selected by user in the form of rhythmic value. This is useful to correct timing errors, but can remove the human “feel”.
Sequences may be saved as MIDI files and transferred to another sequencer regardless of manaufacturer.
MIDI Drum Machines:
- Dedicated rhythm sequencer with internal drum sounds.
- Require little sound memory.
- May have touch pads.
- May be programmed from a MIDI controller instead of pads.
- Not appropriate for sequencing of other sounds.
- Unnecessary if we have a sequencer and synth sampler with drum sounds or a midi workstation.
General MIDI:
A system of standard features found on any general midi device. Specifies how instruments receiving midi data should be setup.
Any GM instrument has 128 programs.
Each program number contains the same sound on all GM instruments.
Midi channel 10 is always a drum kit.
Specific drum sounds can be assigned to specific midi note numbers.
In this way, anyone exchanging midi files in GM mode would hear the same sounds from their midi instruments as the composer without having to change their synth’s program locations and drum note assignment.
Synchronization: Causing to or more devices to run at the same rate.
Midi clock: Midi data used for sync. Contains start, stop and tempo data. 24 clock pulses per quarter note.
Song position pointer: Midi data indicating location within a song to the closest 16th note value.
FSK: Only midi clock that is shifted down in it’s frequency to become audio signal.
Devices that are typically synced include:
- Sequencer to sequencer: To allow sequencers programmed into each other to play at the same rate and tempo using midi clock.
- Sequencer to multitrack: To overdub sequenced performances
- Multitrack to multitrack: When we need more tracks.
Multitracks and sequencers to video: To provide music, dialog and sounf effects that are in sync with the picture.
Console automation to multitrack or video: Automation of our mix allows a computer to “learn” what we do with mixer controls, the playback then executing our moves.
SMPTE: Society of Motion Picture and Television Engineers.
SMPTE time code: A system used for numbering of frames of video. May be used as location and timing information (video frames occur at regular intervals) to allow synchronization of systems other than video.
In the U.S. video is used to run at 30fps. 30fps x 60sec/min = 1800 frames/min
1800fpm x 60min/hour = 108.000 frames/hour
Time code is counted in hours, minutes, seconds and frames.
First frame number available: 00:00:00:00
Next frame: 00:00:00:01
Last frame of first second: 00:00:00:29
Next frame: 00:00:01:00
Last frame of first hour: 00:01:59:59:29
Next frame: 00:59:59:29
The highest SMPTE number possible is: 23:59:59:29, after that it’s 00:00:00:00
Six types of time code:
- 24 fps - film
- 25 fps - European Broadcasting Union (EBU) time code used for Phase Alternating Line (PAL) video systems. France uses SECAM, also 25fps.
- 30 fps - Also called 30 non-drop frame (NDF or ND). U.S. black and white video. No longer in use except for film to video transfer and pull up of audio synced to video that must be transferred back to film.
- 30 drop frame (DF): Same purpose as 30 ND, used occasionally on long program material for TV broadcast.
- 29.97 fps: Also called 29.97 non drop. U.S. color video NTSC (National Television Standards Committee).
- 29.97 drop frame: Same purpose as 29.97 non drop, only used on long program material.
When color was developed, the U.S. system had a technical problem. The addition of the color signal caused noise in the audio. By slowing the frame rate down by one tenth of a percent, the noise was eliminated. The elapsed time shown on time code was now 1/10 shorter than real time.
Drop Frame:
Developed to make 29.97 fps time code match real time. Drop frame drops the frame numbers “00” and “01” for the first two frames of each minute, except minutes “00,10,20,30,40, and 50”
No frames are left unnumbered, just two numbers in the sequence are skipped allowing time code and real time to be the same at the end of one hour.
If syncing audio to any material already time coded such as film, video or other audio, use the same frame rate as the time code on that material.
If syncing multiple audio devices together, none of whivh have been previously recorded and time coded, you may choose any frame rate as long as you use the same time code on all the devices.
It is common to use 30nd
Time code generator: You must use a device called a time code generator to create time code. It may be part of the tape recorder, mixer or if the appropriate midi interface is used, part of the sequencer.
Copying SMPTE: SMPTE should not be copied from one tape to another, as it’s quality is degraded and may not be read properly. In this case, generate new time code.
Re-shaper: fixes damaged time code.
SMPTE can be recorded in two forms:
- LTC: Longitudinal Time Code: Recorded on a track generally for the length of the tape. Used in audio and video production.
- VTC: Vertical Interval Time Code: Type of time code that may be used in video, located in the video signal’s vertical blanking interval as a series of dots.
MTC: Midi Time Code: SMPTE hours, minutes, seconds and frames converted into midi data.
MIDI Clock: Some sequencers or drum machines can only sync to midi clock. We must have a converter to change SMPTE into midi clock. MIDI clock may be problematic in some instances.
Reasons to sync sequencers and multitasks:
- To overdub: When our midi instruments are producing more sounds than they have separate outputs and we want to record them on separate tracks.
- To change sounds from any recorded performance.
- To use virtual midi tracks. Tracks not recorded to multitrack but played in sync with the multitrack and recorded live with performers or during mixdown.