Audio & MIDI > MIDI Interfaces

19" Steinberg SMP24 + SMPII (1987?) for use with PRO 24 the ATARI ST

(1/2) > >>

2 ports in / 4 ports out?

the Steinberg SMP24 has:
2 midi inputs
4 midi outputs
"steingberg midi processing 2 / 4" i think this explains the name of the device
as noted in the diagram it connects to the ATARI ST via a centronics cable?
im guessing into the "modem" connector?

may 1987 article


M.idi for

this device was named SMP-24 because it was a hardware counterpart to Steinbergs flagship product at the time which was "Steinberg Pro-24" sequencer for the ATARI ST. remember this device predates CUBASE entirely!!! but was made to be compatible with Cubase.

--- Quote ---Mark Badger discovers that Steinberg's first 'hardware' release is designed to do more than just sync their Pro-24 sequencer to tape...

Multitrack tape recording, electronic drumming, MIDI sequencing, sound and effect synthesis, are all electronic 'event management' techniques which have become common tools for music creation.

The sheer variety of options available to the modern musician/engineer can make controlling all the possibilities a mind-boggling task. While progress within any single discipline can proceed along logical and straightforward lines, when it comes to utilising the benefits of two or more different systems, and causing them to work together in a single production, the problems can mount exponentially.

The difficulties are compounded by the differing specifications, standards, and capabilities of the many electronic devices which modern musicians employ. Witness the four differing types of 'absolute' timecode (a code which is defined by seconds, minutes, and hours), the multitude of 5 volt 'clock' standards (a timed pulse governed by tempo), and the way in which different manufacturers have 'interpreted' the MIDI specification.

Increasingly, musicians have been looking to computer technology to assist them with control over all these techniques, and to provide them with a friendly environment from which this control can be exercised.


In 1986, and probably '87 too, the computer that was finding its way into many pro and home studios was the Atari ST. Not because it's the only computer for the job, but because a German software house, Steinberg Research, had written a program (the Pro-24) which elevated computer control of MIDI events from mere 'sequencing' to full-blown 'orchestration', and pointed the way towards powerful, and relatively economic, music work stations.

The heavy professional use of the Pro-24 program has brought to light limitations of the Atari ST computer (on which it runs), international synchronisation standards (or the lack of them), and the MIDI standard in general.

It has also focused attention on the problems the electronic musician can encounter when they try to translate their studio work into a live stage performance. Changing patches on all their synths, dealing with incompatible codes and signals, or - god forbid! - pulling out a MIDI lead in order to swap master keyboards between numbers, can lead to all sorts of difficulties.

Responding to the needs of their Pro-24 users, Steinberg have now ventured into hardware design and production. They've brought out a multi-purpose 'black box' (literally) which offers a great deal of assistance for the musician attempting to grapple with event management (the facilities offered by MIDI, SMPTE/EBU, and Clock sync), by solving many of the interfacing difficulties outlined above. They call it the 'SMP-24' - SMPTE/MIDI Processor.


The SMP-24 takes the form of a single 19" rack-mount unit, about eight inches deep. On the back there are two MIDI Ins, four MIDI Outs, SMPTE In/Out, a footswitch socket, a Centronics port (for connecting the SMP-24 to an Atari ST computer), fuse socket and Eurostyle power connector.

The fascia of the SMP-24 is covered with black vinyl, on which are eight printed squares that serve to distinguish the membrane switches beneath from their surroundings.

Programming and data display is carried out via a four-inch wide, eight character, alphanumeric LED display. This is very bright and can be easily read at a distance and from any angle (a general benefit of LEDs as opposed to LCDs). The remaining small LEDs indicate what the internals are up to and which level of operation is being displayed at a particular point in time.


The SMP-24 operates on two linked, but independently programmable, levels: it's a MIDI processor, as well as a SMPTE/EBU-MIDI-Clock synchroniser. Operations can be further separated depending on whether or not it is connected to an Atari ST micro with Steinberg's Pro-24 software loaded and running.

As a stand-alone MIDI processor, the SMP-24 acts as an 'intelligent' MIDI buffer, enabling many of the benefits to be gained from using a MIDI sequencer to control large arrays of MIDI devices.

Asa stand-alone synchronisation processor, it will read and generate most of the common timecode standards, and serve as a master time controller/interface, translating these codes as necessary.


There are 63 'MIDI Programs' available, each acting as a memory store for whichever of the many MIDI processing options are required to be in operation. These MIDI Programs can be called up remotely by program change commands from a master keyboard, or by pushing the appropriate buttons on the SMP-24 itself.

This program change facility is controlled by a 'master' program, which is active whenever the unit is on, called the 'SMP.OO'. This controls at which input and on what MIDI channel the SMP-24 should expect to receive these program change commands.

Other 'macro' parameters which affect the overall operation, and which are also accessed from the SMP.OO program memory, deal with the footswitch operation (program up, start/stop, or display change); the setting of the MIDI in/out speed (the standard MIDI baud rate can be doubled to 63 kHz), whether the onboard metronome is activated or not (one of the first things I edited, as it's a particularly penetrating 'beep'); and whether the LED display uses numbers or letters to indicate which parameter is currently being edited.

The 'SMP.OO' program also allows the user to utilise 'running status' - a form of MIDI transmission which ignores unnecessary 'status bytes'. Some MIDI messages are needed only by older synths, in order to tell them things like "Yes, you should be on but not playing any notes at the moment." More modern MIDI instruments can figure this sort of thing out for themselves, so they can dispense with 'status bytes' which just clutter the MIDI lines and slow transmission.


With two independent MIDI Ins and four independent MIDI Outs, virtually any patching or connecting requirement can be met by the SMP-24. The status of these connections, and the source and state of the signals appearing at them, are programmed into what Steinberg call a 'Configuration Program' memory. You can store 63 of these.

Each of the 63 'MIDI Programs' mentioned earlier can utilise up to four of the 63 'Configuration Programs' for its operation. It can also send up to 20 specified program change commands, with total flexibility over which of the four MIDI Outs, and on what channel, each is sent.

By layering the operations of the SMP-24's MIDI processing between these two types of program, Steinberg have ensured that the SMP-24 can be flexibly altered in order to deal with virtually any MIDI difficulty. (Like Yamaha codes detuning Casio synths!)

Each configuration allows you to specify which of the two MIDI In signals will be processed and which channel(s) are to be effected. Filters can be activated which prevent the SMP-24 retransmitting such data as note-on, note-off, aftertouch, pitch bend, pressure and system exclusive/timing information etc, and there are two user-defineable controller information filters to cater for special needs.

Pitch and velocity transposition to a specified degree can be invoked for each of the four Outs, as well as high and low split points.

By providing these options as configuration parameters to be controlled, along with other configurations, by a 'higher level' program (the 'MIDI Program'), complex but effective MIDI information matrices can be constructed.

So to recap, the SMP-24 offers the user the ability to join MIDI signals together (merging), to filter unwanted or incorrect data (some MIDI-equipped units send out all sorts of garbage) on selected channels and outputs, to cause two events to happen at once (as opposed to an instant after one another - remember, MIDI works by serial data transmission), and to split 'unsplittable' MIDI keyboards. Not bad for starters!


As I said, the SMP-24 is really two devices in one box - a MIDI processor and a synchroniser - and it can perform both jobs at once.

As a master synchroniser, the SMP-24 can generate and read all forms of SMPTE/EBU 'absolute' timecodes (24, 25,30, and 30 drop-frame) accurate to the nearest 'bit' (there's 80 of these for each 'frame').

SMPTE is an abbreviation of the Society of Motion Picture and Television Engineers (who invented the digital code); EBU, the European Broadcasting Union. While these groups have agreed on what a timecode should contain, they have not found a common ground, which necessitates provision of facilities for working with the four common types of timecode on any synchronisation device which is going to see action on both sides of the Atlantic.

The SMP-24 can also output, and sync to, 5 volt 'clock' timecodes (which are timed relative to tempo). It can handle any clock pulse up to 192 pulses per quarter note as well as DIN sync, with its separate start/stop and clock signals. For extra accuracy, the SMP-24 reads both 'sides' (the up and down edges) of the clock pulse (usually a square wave), which effectively doubles the degree of locking - especially useful for slow tempo stuff.

While it's doing either, or both, of the above activities, the SMP-24 can be translating all this sync information into MIDI song pointer and clock data which it can transmit out of any chosen MIDI Out port to connected peripherals.


These sync parameter options are accessed from within one of the 64 'Synchronisation Program' memories. Sync memory 64 is reserved for use by Steinberg's Pro-24 sequencing software though, and cannot be programmed directly from the SMP-24.

By activating one of the other 63 programs, you can specify whether the Tap button on the fascia of the SMP-24, the footswitch, the Clock In socket or either of the two MIDI Ins will provide the trigger source, and what sort of 'clocks-per-beat' rate the SMP-24 should expect.

If you wish, the SMP-24 can calculate tapped tempos, input manually via the Tap button on its front panel or the footswitch, in relation to a previously recorded timecode - a useful feature when there's no regular indication of beat that can be linked to the clock (like a repeating hi-hat sound or snare). As the first tap is taken to be the 'start' point, this facility can also be used to find beginnings. The Owner's Manual states that the next software update will include provision to lock onto MIDI note events as clock sources, in addition to the present options.

The Sync Program parameters include specification of the clock output rate, beat count, and at which of any or all of the four MIDI Outs this data will appear. This is also where the type of SMPTE/EBU timecode that the SMP-24 will generate or read is set, and the start and cue times (to 'bit' accuracy) are specified.

Start times can also be set in a special mode called, appropriately, 'Time Edit'. This mode displays the time value as an eight digit figure on the LED display and allows programming by 'zone' (hours, minutes, seconds, and frames) rather than by specifying the multiples of each unit thereof on separate displays. Eight digits mean you can't display the 'bit' figures in this mode, so to get at this setting you have to return to the Sync Program.

Each of the Sync Programs looks to something called the 'SMP-24 Song', which amounts to a tempo change table. It is here that the results of any 'tapped tempo' are stored, and a different tempo can be set for each quarter-note of a song, lasting up to 1024 bars.

If you intend adding some 'breathing' to chorus and verse sections and you get to a point where the tempo isn't going to change, the remaining bars can be 'filled' with a specified tempo - a facility that can save you a lot of programming time.

The entire contents of the SMP-24's 191 memory locations (63 MIDI Programs, 63 Configuration Programs, 63 Sync Programs, the 'SMP-24 Song', and the 'SMP.00' Program) can be saved to, and loaded from, audio tape via the SMPTE connection. This allows you to save individual set-ups together with the song on tape or as a back-up on cassette while on tour, and can prevent a lot of grief.


Having covered what the SMP-24 represents in terms of a stand-alone MIDI/SMPTE processor, we can now look at how it works in conjunction with the Atari ST computer and Steinberg's Pro-24 MIDI sequencer program [reviewed SOS July 86].

When connected via the Centronics ports on the back of the respective machines, most of the SMP-24 controls become integrated with the Pro-24 environment, providing a very comprehensive 'MIDI studio' control station and allowing the Atari computer to lock-in to the 'outside world'.

The locking itself is extremely accurate because the computer's direct connection to the SMP-24 allows timing information to be passed between the two units at four times the standard MIDI clock rate of 24 pulses per quarter note.

The SMP-24 vastly improves the Atari's MIDI implementation. While the built-in MIDI interface makes the Atari ST ideal for running MIDI software (which can make up for the greater cost of the ST in relation to other computers), as users load their MIDI systems with vast amounts of data (something encouraged by the power of the software), speed of transmission suffers. This delay is particularly pronounced when using the 'absolute gating' provided by digital techniques.

Other drawbacks of the Atari's single MIDI In and Out connections arise when trying to cater for extensive and complex studio and stage set-ups or when attempting to record two MIDI instruments at the same time.

Each of the 24 tracks on the Pro-24 sequencer can be assigned to one of the SMP-24's four MIDI Outs. As each track can play information on any or all of the 16 MIDI channels available, this facility allows extensive and flexible control of up to 64 discrete MIDI channels.

The same techniques that allow a track to play more than one channel can be used to record more than one player, so Pro-24/SMP-24 users can now hold MIDI 'jam' (improvisation) sessions - a facility previously available only to select users of MIDI software running on Apple Macintosh computers (notably Southworth Music Systems' Jam Box, which has four MIDI Ins!).

Incidentally,the SMP-24 will not support the Pro-24's MIDI dump facility, though this option is still available from the Atari's built-in MIDI port.


Control over the SMP-24 facilities is accessed from within the Pro-24 program by pulling down the 'MIDI' menu with the mouse and selecting the 'SMPTE' option.

A 'window' then opens on the Atari screen to reveal controls for start time and cue point setting, SMPTE type, and which of three possible modes of operation the computer is in - Free-Run, Cue-Start, or Tape-Sync.

Locking the program to tape could hardly be easier. A code is recorded on whatever tape/video machine you are synchronising your Pro-24 work to, and then the start time (or up to 16 cue points) is set. The only other necessary operation is to programme and activate the Pro-24's 'master track', information from which is sent to the SMP-24's tempo change table.

These settings become part of the song file and are stored with it to disk when the song is saved (the Atari has a built-in 3.5 inch disk drive, remember). This solves the problem of having to find the correct beginning or tempo whenever a particular song is loaded.

The computer will then follow the tape, jumping to whatever position is indicated by the SMP-24's SMPTE readout, tempo table, and start time setting, whenever it hears the timecode.


The computer takes about three beats (1.5 seconds) to find the correct bar and start playing. While this is hardly instantaneous it's certainly fast enough, and dispenses with the need for 'drop-in' guide tracks when multitracking (necessary if you are using clock pulses for synchronisation and you don't want to start from the top every time).

Curiously, the speed with which the computer finds the required song position is no faster than if the Pro-24 were chasing MIDI song pointers, with the SMP-24 MIDI Out connected to the Atari's regular MIDI In.

What does improve is the tracking between the song pointers as the tape is running. By utilising the direct connection provided by the Centronics link, I got noticeably less flanging of a drum machine's ride cymbal sound when playing the machine back via MIDI along with a recording of the samething. (Ideally, if the two were perfectly in-sync, flanging would not occur.)

I compared both results with the Pro-24 sequencer chasing song pointers from a Fostex 4050 synchroniser. The flanging effect produced was about the same depth as with the Pro-24 following a similar output from the SMP-24.

There have been some less than complimentary reports rumoured about the SMP-24's performance. This is partly because the early models that arrived in this country were unfinished, and were prone to locking 'up' rather than 'in'. Steinberg wisely restricted early sales of the SMP-24 to existing Pro-24 users who were prepared to be 'computer cowboys' and 'rough-ride' the unit whilst the bugs were identified and ironed out.

In the short time that I have had my hands on this latest SMP-24, I have done my best to subject it to hard-working conditions. I have done three lengthy (14, 17, and 19 hour) SMPTE-reading experiments, writing songs with the computer and my 8-track locked in marathon music sessions. The system performed continuously, and without hiccup, throughout, and the result is some monster-sounding tapes!


As I said in the beginning of this review, Steinberg Research moved into the professional music market with their Pro-24 software, a program which takes excellent advantage of the display and 'human interface' benefits to be gained from using a computer for the control of MIDI events. Using the SMP-24 highlights the same advantages; it could hardly be easier to set up and operate from the computer. However, without that user-friendly and powerful environment, as a standalone device, making the SMP-24 work according to one's needs can present a daunting prospect.

Along with most of the other 1U high, rack-mountable gizmos available at present, the physical constraints of having to fit all the buttons and the display onto only 33 square inches of space has necessitated a few compromises when it comes to 'ease of operation'.

Steinberg's decision to go for visibility, in the form of their large LED display with only four-character descriptions of parameters, has meant that this unit can at times present some cryptic programming options (hieroglyphic even, as letters such as 'g' get bumped up!), and the compromise bites deep into the unit's friendliness.

These two physical problems are not helped by the manual, which must rate as one of the most slack German-English translations I've ever seen, with 'oder's and 'modi's turning up everywhere [It can't be as bad as the original PPG Wave manual, surely? - Ed]. This is a surprise as their Pro-24 manual is certainly adequate, if not good.

(Note: Steinberg tell me that the SMP-24 manual is "under review" and that purchasers will get the updated manual as part of the service. They will also receive free updates of the SMP-24 software, rather than having to 'buy in' to an update scheme. The first SMP-24 update should add 'onboard' sequencing to the SMP-24's already considerable utility.)

Luckily, most Germans are pretty straightforward and the people at Steinberg Research HQ are no exception. After a day's use, you fall in with their line of thinking and using the SMP-24 begins to become understandable and fairly logical.

Steinberg are currently hard at work, writing software that will run on the Commodore C64/128 and the Atari ST, in an effort to make the SMP-24 easier to use without the benefits of the Pro-24 operating environment. This should vastly improve the 'user-interface'for those of us with such hardware.

Those without, however, will have to continue to struggle with the eight-character display, at least until operating the SMP-24 becomes second nature. This is a process which, given the present manual, may take some time! I'd recommend at least a couple of hours of 'fun and experimentation' before settling down to earning some money with it.


Whilst the MIDI options are not as extensive as those of Axxess Unlimited's 'MIDI Mapper' (say), nor its syncing options as wide-ranging as the Friendchip SRC, the Steinberg SMP-24 nevertheless offers users a number of attractive and powerful features, whether or not they own an Atari ST and Pro-24.

These features cannot be found linked together anywhere else in the manner that the SMP-24 accomplishes (as far as I am aware), and as such it is a tool which can offer extensive and useful control in many situations where you would normally resort to interfacing two, or even three, MIDI devices.

This ability to 'multi-task' goes some way towards mitigating for the unit's rather high price (£799 plus VAT) - though SMPTE synchronisation never comes cheap! The Bokse SM-9 weighs in at £693, and the Fostex 4050 and 4030 at £783 and £1300 respectively (all plus VAT).

While these units offer comparable synchronisation options, none offer the SMP-24's range of MIDI facilities.

These factors, as well as Steinberg's reputation for good support (they offer a telephone 'hotline' service) and continuous product development (as anyone who has watched the Pro-24 software grow can attest), ensure that their SMP-24 SMPTE/MIDI Processor will receive a lot of use in all types of studios.

The software nature of the product means that new synchronisation standards (such as MIDI-SMPTE or MTC) can be easily accommodated through updates of the program chips which drive it - insurance that, once spent, your investment is not overtaken by events (!).

With the proviso that you should be prepared for an awful manual, and thus a brain-bashing first few hours, I heartily recommend this device to anyone who thinks they may gain some benefit from its use. Anyone, that is, who is struggling at the interface between the many different systems of event control that are currently, and even soon to be, in use.

(Incidentally, the manual is actually better than the one Fostex supply for their 4050 synchroniser, so don't be too put off by my warning.)

My thanks to Chris Collings at Raper & Wayman (Contact Details), who supplied me with the Fostex 4050 to use as a 'yardstick'.

SMP-24 costs £799 ex VAT (£918.85 inc VAT).
--- End quote ---

photo of the SMPII unit with interface card which looks like its for IBM/PC ISA slot

midi ins + outs : 2 in 4 out (16 x 4 = 64 channels?)


[0] Message Index

[#] Next page


Go to full version