Home -> Magazines -> Issues -> Articles in this issue -> View
Steinberg Software Page | |
Article from Sound On Sound, July 1988 |
The third instalment of our column devoted entirely to Steinberg software. Written and compiled exclusively for SOS by Steinberg's experts, it deals this month with MIDI clocks and how to keep time!
Welcome to a new experiment: the Steinberg Software Page. Each month in Sound On Sound we will bring you the latest news of current and forthcoming versions of your software, some 'in-house' tips and tricks on its use, and will discuss computer-based MIDI sequencing systems in general. It is our hope that everyone using a computer to assist them with their musical productions will find our subjects of interest, whatever software they use. Read on and respond!
The Steinberg Help Line has reported that there seems to be an unusually high number of you calling in to ask fairly basic questions about time and its relationship with all that is MIDI. With this is mind, we're going to devote most of this month's column to time-related MIDI issues. Most of what we'll be discussing will apply to any sequencing set-up.
MIDI CLOCK: A MIDI Clock is a System timing message (everything in your set-up will 'listen' to it, you select those which will respond) sent out by a drum machine, sequencer, or synchroniser 24 times per beat. Its rate per second is directly related to whatever the current tempo is. At a tempo of 120 beats per minute, for instance, all the MIDI devices you have connected in a MIDI chain will be processing the arrival of such a clock signal every 0.02 seconds. You cannot record this clock on a regular tape machine.
STANDARD CLOCK: These range from FSK to 5 volt PPQN (pulses per quarter note) types, just so long as their frequency is well within a tape machine's frequency bandwidth. In practice most FSK or PPQN codes run at a rate of only 48 or 96 clocks per beat.
CLOCKS GENERAL: Both types of Clock message, MIDI and standard, have a severe limitation: they only contain one piece of information that tells a device to 'advance one step'. If you start the clock running at any point other than the very beginning, the slave device will not know where to begin. You always have to start at the top.
SONG POSITION POINTER (SPP): Yet another MIDI System message. It is almost always broadcast by the clock source, usually in response to an external trigger. You could almost include Start, Stop and Continue in the list of SPP messages, but in fact they are defined separately. Song Pointers are calculated from the number of beats that have elapsed since the start of the song and are broadcast fractionally before a Continue message. They are only sent once, being utilised to inform all the connected machinery at what point to begin playing, if not the Start.
CONTINUE: Most drum machines broadcast a continual stream of MIDI Clocks, whereas most software sequencers wait until you've started them before sending MIDI Clock data. The Continue message is sent in place of one of these clocks after the Song Pointer has been sent. There is no requirement for an SPP to be sent before a Continue message will be recognised. It is the proper combination of these three messages - SPP, Continue, and MIDI Clock — which enable any MIDI sequencing device to be told to start playing at somewhere other than the beginning of the song.
SMPTE (pronounced 'simpty'): This is also a time-code which you can record on all sorts of tape. The name is an acronym for the Society of Motion Picture & Television Engineers, an American body who established a standard for an absolute linear timecode using a format of five pairs of numbers. The numbers relate to Hours, Minutes, Seconds, Frames, Bits.
In the USA, the frames are divided into 24, 30, and what is called 30-Drop (used with colour TV synchronisation). The standards are different here in Europe, and so is the name of the group who define them. We use 24 or 25 frames per second, as defined by the European Broadcasting Union (EBU). You will notice that 24 frames per second is a sub-division for both groups. This is because celluloid film runs at the same speed in both continents. Whatever the Frame rate, there are 80 Bits (binary digits) in each frame.
In the music industry people tend to talk about SMPTE generically, lumping all of these codes under the heading of 'SMPTE' (though they tend to be using EBU not SMPTE!).
Encoded within the SMPTE timecode there is a continual progression of timing information, the data itself being defined by 'real time'. Using the appropriate equipment this timecode can be decoded and used to synchronise a variety of machines. Usually the same machine which decodes the information performs the encoding in the first place. You might run into difficulties if this is not the case.
PRO-24 & SMPTE: Steinberg build two hardware devices which are intended for use when synchronising the Pro-24 sequencer to an external 'master'. The Timelock provides a basic, but effective, level of synchronisation to all forms of SMPTE code for Pro-24, via the Atari's Centronics port (the printer socket). The SMP24 provides more sophisticated synchronisation which is capable of responding appropriately to a number of different timing information sources, also via the Centronics port. Those of you without such machines will find that there are two menu options which remain in grey on your computer's screen: 'Out-Channel' and 'SMPTE'. These menu options are used to provide software control of the Steinberg hardware functions. In order to achieve SMPTE sync, you will need some other device which: (1) Is capable of reading and writing a SMPTE timecode; (2) Is capable of having a start time entered into it, along with more detailed data about how the Tempo values will progress throughout the piece you are hoping to synchronise.
INCOMPATIBILITIES: I spoke earlier of reading and writing the same SMPTE timecode with two different machines. Due to the nature of 8-bit electronics and the sort of processor speeds necessary to attain accuracy for the clock signals, all synchronisers perform 'rounding off' operations on the results of their tempo versus time calculations. You will very probably find that two machines from different manufacturers put the same SMPTE codes at two completely different bar positions for the machines synchronised by SPPs, Continues, and MIDI Clocks (the necessary ingredients for synchronising any MIDI sequencer part-way into a song). The SMP24 can cope with the calculations of the SBX80 and SRC synchronisers because Steinberg studied the way these machines perform their timing calculations.
GETTING OUT OF A HOLE: If you always record a 'click-track' as well as your SMPTE code, then whatever way the machine you were using was calculating your tempo/time data, you should be able to determine the proper tempo mapping for the machine you are using now. Provided the synchroniser has the facility to read both the SMPTE timecode and the click-track, resolving the two as the track proceeds. This facility can be found on the SMP24. You don't really have to record the click-track until you have finished completely with your session, as you are only archiving the tempo.
TIMING ERRORS: As you can readily see from all of the above information, there are ample message types and facilities in the MIDI world to perform highly accurate synchronisation, yet within this accuracy lies a 'sleeping lion'. Any device you have plugged into your system which does not understand MIDI Clocks (synths and drum slaves for instance) cannot play on the beat. Why? Because the MIDI Clock occurs on the beat and the device will be busy ignoring it, introducing a delay of about one millisecond.
Always take this into account when examining the timing of your music. Don't send MIDI Clock anywhere it is not absolutely necessary, the same with MTC. For super-metronomic, mega-tight grooves which you record on tape, you may also find it judicious to record the kick and snare drum individually on separate passes. Whenever they coincide, due to quantisation the note on the lower numbered track gets transmission priority, an inbuilt delay of another millisecond if you are using Running Status, more if not! You will find that a multiple address MIDI Out extension for your Atari, if your sequencer developer offers one which is compatible, can do a lot to help minimise these MIDI dilemmas. These are not due to limitations of your Atari or anyone's software, they are caused by the serial nature of MIDI message transmission.
With Pro-24, try saving each song as a mega-mixdown single pattern, onto no channel. Start a new song which will become your 'performance memory'. Load your mega-patterns in one at a time, moving the Locator Positions and the Destination Counter to start each song/mega-pattern at a new position (remember that the counter goes up to 999). Store the Locator Positions into the Function Key memories as you go. Then just two simple keypresses,
The MIDI Out socket on Atari ST computers is non-standard. Use only the middle three pin connections of the regular 5-pin MIDI plug, or treat yourself to a good quality Klotz MIDI lead.
There is no onscreen indication of the Atari keyboard's CAPS LOCK key having been pressed when running Pro-24 or indeed any other Steinberg program. Many software functions change their behaviour when this key has been pressed, so if things don't act as you would expect them to, check this out.
This page has been written and compiled for SOS by Evenlode Soundworks, the UK distributors of Steinberg software.
On The Cards - Sound card feature |
Sound Bites - Production Tips & Techniques |
Interfacing The Past (Part 1) |
Amiga Notes |
Augmenting the 3750's Memory |
Dance Music Sequencing Techniques |
![]() Amiga Notes |
![]() Amiga Notes |
![]() Software Support |
Brave New World - The World Of CD-i |
![]() PC Notes |
Browse by Topic:
Feature
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!
New issues that have been donated or scanned for us this month.
All donations and support are gratefully appreciated - thank you.
Do you have any of these magazine issues?
If so, and you can donate, lend or scan them to help complete our archive, please get in touch via the Contribute page - thanks!