STEINBERG

MERGING MIDI DATA

Audio signals, in the semi-pro world at least, are fairly forgiving. If you need to merge two signals, Y-leads and splitters do an adequate job. The results are variable, as the input and output impedances really need to be taken into account, but it generally works, and you can always tweak the gain controls here and there to compensate.

In the world of MIDI such things are totally out of the question. MIDI data is a stream of digital information whose order is very important. You can't just stick bits in here and there, or arithmetically add two signals together — you will need a MIDI merger. These devices are really quite complex in operation — a merger needs to be able to listen to and understand two streams of data, decide what needs to be sent through immediately, and what must be buffered 'till later and only sent through when there is time. Moreover the unit must be able to understand the sense of the data, and let complete 'packets' of related data pass through as a whole. MIDI mergers are not Y-cables!

They can get you out of a spot when the MTC output from a synchroniser must be merged with new MIDI from your master keyboard when you record in sync, but they are still a compromise because of the 'time buffering effects'. Much better is the solution of a multi-input system, where all the data arrives at the sequencer's input together. If you do need to buy a MIDI merger, it is vital to try before you buy — and try it in a realistic situation. Some mergers are definitely better than others.

SMPII INTERFACE CARD

The tempting prices of laptop and notebook PCs (IBM-compatible) make them look like good platforms for Cubase Windows, but there are some important issues to consider. Laptops have no expansion bus available, so standard MIDI interface cards using the ISA bus cannot be fitted. There are a growing number of serial port MIDI interfaces around, but be warned that many of them come with a disclaimer stating that the manufacturer cannot guarantee operation in all conditions. The fact that Steinberg supplies drivers for for this type of interface doesn't mean that we have magically made it all work.

The internal architecture of some PCs means serial port MIDI interfaces will not work properly. If you take your sequencing seriously you need a reliable interface. Basically that means one that fits into the ISA-EISA expansion slots or the parallel port. Steinberg support standard MPU cards as a 1-in/1-out MIDI interface solution, but also support other multi-input/output cards. Two of the most notable are the Steinberg SMPII and the MOTU MIDI Time Piece. Both of these units require an interface card that fits into an expansion slot before the unit can be fitted.

SCORE PRINTING ON THE MAC

Steinberg have just released Cubase Score for the Apple Macintosh. This allows for full page-based score printing, in addition to the standard score editing facilities. All the features of V3.0 for the Atari are included, such as the ability to graphically move any object to any user preferred position, just to make it look right on paper, without having to quantise or edit the music in any way.

One feature that isn't present in the Atari version is the ability to print to a PostScript file. This means that the 'printed' output will be saved as a file on disk that can be taken to a typesetting bureau and printed to any resolution — not just the 300dpi (dots per inches) that is available to most users. Individual pages can also be saved as Encapsulated PostScript (EPS) files. These can be imported into other Mac DTP packages, and therefore allow sections of your music to be incorporated into other documents. But beware — both these file types are not 'pictures' of the screen, but instead complex lists of instructions for how to recreate the entire page on another system. The list of instructions needs all the information that the original needed. So, without the fonts supplied with Cubase, other programs will print the file with a substitute font. In most-cases this will be Courier 40pt — not quite what you wanted, and perhaps a little too avant garde for most people's tastes.

DR. T's KCS

MORE ON CONTROL SEQUENCES

Last month we started looking at control sequences and how they can be used to govern the playback of other sequences. Astute readers will have noticed that in last month's examples we were starting sequences with the so-called Secondary Start format rather than the Primary Start variety. Let's look our very first example again, but written now in Primary Start format:

Seq 1, Song 1a

When using Primary Start commands, the value under NOTE is the pitch transposition and that under VEL is the velocity transpose; when using Secondary Start commands, the value under VEL is the pitch transposition in both cases DUR is the number of times to play the sequence. The reason for the two different formats stems from KCS's early days on the Commodore 64. The Secondary Start format was added to the ST version because, primarily, of the increased number of sequences available for data storage. It would seem that there is no way to do a velocity transpose when using Secondary Start commands but they may be achieved thus:

Seq 1, Song 3

This would start sequence 2 and immediately reduce all velocities in the sequence by 25. While this may initially seem a more cumbersome method, it does have the advantage of standing out visually in the edit list. ST users should in all circumstances use the Secondary Start format; this is because SongEdit's Convert Control Sequence function requires all start instructions to be such to ensure the conversion process works correctly. If you create a control sequence with SongEdit, you will notice that it uses this format, and wherever a velocity transpose has been performed on a Section that the conversion process embeds VT events in the appropriate place.

Next month we will look at some more advanced applications for control sequences and, in particular, explore some of the possibilities of indeterminacy in controlling the program flow.

MUTING SEQUENCES FROM THE KEYBOARD

Users of KCS on the ST often ask why it is that sequences 2 and 3 cannot be muted from the keyboard using [Shift] 2 and [Shift] 3. The reason is that the standard American keyboard is laid out slightly differently to the UK, so consequently the mute function for these two sequences appears not to work. Instead, use [Shift] ' (@) for sequence 2 and # (no shift key) for sequence 3. Incidentally, the same problem occurs in Copyist when trying to define the four control points for slurs; these would normally be [Shift] 1 through [Shift] 4 — using the keys noted above will cure the problem.