Home -> Magazines -> Issues -> Articles in this issue -> View
Studer Dyaxis | |
Digital Audio Production SystemArticle from Sound On Sound, May 1990 |
Today's high-end tools for the professional are tomorrows affordable tools for the musician in the street. David Mellor explores a Macintosh-based digital recording and editing system from a company best known for their analogue tape recorders.
Studer is a company which is very highly regarded for its analogue tape recorders, from simple stereo to sophisticated 24-track. Studer machines are, by the way, expensive - the Rolls Royces of the recording industry. In addition to their analogue range, Studer now produce digital reel-to-reel recorders, obviously making full use of their experience in tape transport mechanisms. But although reel-to-reel recorders, and cassette-based digital recorders, have their advantages, hard disk systems have such potential that it would be unwise for a major company to ignore them. Studer, rather than develop their own system from scratch, bought a Californian company already active in the field, Integrated Media Systems, and formed a fresh company called Studer Editech. One could speculate from this that Studer probably intend to use the R&D and manufacturing base they have acquired to develop high-end digital systems to complement their other high-end equipment. But for now, we are seeing a development of an existing system which happens to be in - for Studer - a low price bracket.
This first product, the Dyaxis Digital Audio Production System, is a two channel recorder and editor which uses an Apple Macintosh computer as its user interface, and also for some of the audio processing. Now this can't be just any old Mac, I am sorry to inform all those MacPlus and SE owners out there. Although the basic Macs will work, they can only cope with sampling rates up to 32kHz, which equates to an audio bandwidth of something under 16kHz. Depending on your needs, this may be OK - after all, FM radio only extends up to 15kHz and that can sound pretty reasonable in a good reception area. But if you want to edit recordings made on DAT at 44.1 or 48kHz, then you will require the services of a Mac SE30, IIex, IIci or other model fitted with the faster Motorola 68030 microprocessor.
The hardware part of Dyaxis consists of a smallish rack containing an audio processor, hard disk drives, and an optional timecode reader/generator. There are two versions of the processor: the analogue-only input/output version, and the digital version which has interfaces for all the digital formats you are likely to come across. The hard disk drives are available in 105Mb (Megabyte) or 320Mb capacities, and the system supports up to six drives. Audio storage therefore ranges from five minutes of stereo to 75 minutes of stereo, depending on the fatness of your wallet. The biggest version would be necessary if you wanted to hold a complete full-length CD's worth of material on hard disk at one time.
Software-wise, the user is given the power, via the Mac, of digital recording, playback, sync to timecode, waveform editing, Sound File mixing, crossfading, and other processing such as envelope shaping. A further option, not discussed here, is a DSP (Digital Signal Processing) card which allows real-time digital equalisation, dynamic level and pan control, on-screen metering, time scaling, and sampling frequency conversion.
Before looking in more detail at the Dyaxis working environment, it is useful to know the various file types it employs. A Sound File is an original recording made onto the Dyaxis hard disk, or a mix of two or more files, or a new file created by copying a portion of another sound file. Sound Files have big notices on them saying 'Do Not Touch'. In fact, you can't touch them other than to create them and delete them totally. This, of course, is the beauty of hard disk working - you never alter the original material. It is always held safe and inviolate on its own patch of disk space. Compare this with editing a quarter-inch master tape where one wrong cut and you're... probably out of a job!
To edit the material contained in a Sound File, you need a View File. A View File contains no audio, but has edit decision information about how to play back the audio contained in a Sound File. For example: 'play the first 30 seconds, skip 10 seconds and start playing again without a gap'. A View File is automatically created when you open a Sound File, and the information it contains is displayed in a View Window. For example, if you recorded a snippet of audio and called the Sound File thus created 'Snippet', then when you look at this, via a View Window on the relevant View File, you will find the View File is called 'Snippet.view'. Sounds sensible so far. This is just the first View File that will be opened on this one segment of audio. There will probably be several, and it is up to you to dream up meaningful names after the first one.
If you wish to see a representation of the audio on the Macintosh screen, as you probably will, then another file is necessary. This is a Waveform File. In my example it would be called 'Snippet.env' automatically. A Waveform File is needed because it would take too long and occupy too much of the Mac's power to draw every waveform point on the screen. The Waveform File is therefore a downsampled version of the Sound File and is much shorter, in terms of bytes, and easier to deal with.
When you have completed work on a few View Files - perhaps edited versions of several tracks or simply 'topped and tailed' the tracks, removing those odd noises that always accumulate on the tape just before the beginning and just after the end - you need a Mix File. A Mix File contains data on how to play several View Files in sequence, specifying any additional processing such as crossfades, stereo panning, and changes in level from track to track. Once again this file contains no audio, just instructions on how to play back data from your original, and still unaltered, Sound Files. A Mix File needs a new name, which might be 'Album.mix'.
The Mix File, at this point - ie. after you have created it but before auditioning - might contain playback instructions that cannot be performed in real time from the Sound File data. A stereo crossfade, for example, needs four playback channels and Dyaxis only has two. A new file is therefore created which contains the processed data, and this is called a Mix Output File. The Mix Output File follows the instructions contained in the Mix File and is updated, if necessary, every time you request a playback. In my example, it would be called simply 'Album', without the '.mix' suffix. The Mix Output File is actually identical in form to a Sound File - it contains real sound data. If you wished, you could open a View File on it and start a whole new session of editing. Remember that, although it sounds as if there is a lot of processing and general trickery going on, this all takes place in the digital domain and is not subject to the same quality losses as equivalent analogue procedures would be.
To round off this section, let me recap on file types:
Recording isn't quite as simple as recording directly onto a tape or DAT recorder. There are several options to be set from the Dyaxis Record menu (Figure 1) such as file name, recording time, sampling rate etc. The sampling rate can be between 7kHz, for telephone line quality, to 48kHz. Four convenient rates are given their own boxes in the Record dialogue box, but you can type in your own rate from a choice of 125 alternatives. Recording Mode is simply stereo or mono. Obviously, you can record twice as much mono information in a given disk area than you can stereo. Emphasis can be off or on ('pre-emphasis' being where high frequencies are boosted during record and reduced during replay. It can be applied to most recording systems, including DAT, and reduces noise at the expense of a smaller margin of high frequency headroom).
Direct digital copying requires the user to select the appropriate interface (Figure 2). There are three sets of digital inputs which can be configured to a variety of formats:
S/PDIF - the Sony/Philips digital interface used on DAT and Compact Disc players, uses a single phono connector to carry two channels of digital audio. Sony F1 format tapes can be played back on a Sony PCM601 machine into Dyaxis using this input.
S/DIF - the interface used on the Sony 1610 series digital recorders.
AES - a balanced, professional, version of S/PDIF working at higher voltage levels and using XLR connectors. On Dyaxis, a 'D' connector is used instead of an XLR to save panel space.
Yamaha - the Yamaha stereo interface is used to connect to a DMP7D digital mixing console (the DMP7D is the digital input version of the DMP7). A mono Yamaha format also exists, but this is not supported by Dyaxis.
PD - yet another interface, used by Mitsubishi and Otari on their ProDigi machines.
As recording progresses it is possible to mark sections for later editing by clicking the mouse button. Later on, when you open a View Window on the recording - the Sound File - there will be a series of markers along the bottom of the window, showing where you clicked.
Opening the View Window, you are presented with a blank screen except for the Event Band along the bottom (where your markers will be located), a small toolbox on the right, and timing information along the top. To see the waveform you need to create a Waveform File, which may take some time if it's a big Sound File, but it's worth the wait for ease of editing later. There is an alternative to creating a large Waveform File, which is to use the 'Always Display' option. With this selected, a complete Waveform File for the whole Sound File will not be calculated, but Dyaxis will calculate just the section of audio currently held on the screen. This results in a small but constant delay each time the screen is redrawn, such as when you zoom in.
The Waveform Display shows the amplitude of the Sound File, and it's easy to locate good editing points just by looking at it. As you zoom closer and closer in, showing smaller and smaller time intervals, the display changes to the waveform itself. A view of one sample per pixel (picture element) is possible, so removing audio clicks should be a doddle.
Finding places to cut is done with two cursors, which can be dragged out with the mouse from the left and right sides of the View Window. The cursors can be roughly positioned with the mouse, then fine trimmed with the left and right arrow keys on the Mac keyboard. Markers that were inserted during recording, or subsequently, assist quick cursor positioning. The cursors can be moved in 10 millisecond or one millisecond increments; if you are in Fine Locate mode or viewing at the individual sample level of resolution, then the increments are 10 samples or one sample. Needless to say, if you can't find the right edit point with single sample resolution, then it's no use blaming the equipment! Figure 3 shows a View Window on a Sound File called 'Test Track', together with its waveform display. A section is highlighted in reverse video for editing.
Playback of edit decisions is easy: the Macintosh's 'p' key plays the segment between the cursors; the 'b' key plays the part before the left cursor; the 'a' key plays the part after the right cursor.
It's interesting to compare this type of editing with old-fashioned quarter-inch hacking and chopping. On tape, you have to acquire the skill to know that you are marking and cutting in the right place. If you bodge it, you might be in trouble. One way of listening accurately to edits that don't seem to turn out quite right is to splice in a short length of leader tape between the two sections being joined. By doing that, you can hear up to the edit, and as a separate event listen after the edit. That little trick usually shows up exactly what the problems are. On Dyaxis, you can have full confirmation that your edits are spot on very easily. Therefore, mistakes should become almost impossible.
A neat trick that Dyaxis offers for times when you are cutting out sections is to play across the cut - called 'Cut Play', selected from the menu bar. In this playback mode, you hear everything that is displayed apart from the segment you have marked out. This function doesn't have a single key command dedicated to it but the system I used for this review had a Mac utility called QuicKeys which allows single key commands to be set up for almost anything. It seemed sensible to set up the 'c' key to perform Cut Play.
Figure 4 shows the Fine Locate display, where the waveforms just before and just after the cut points are shown. Here, the discontinuity in the waveform (shown by the mouse pointer arrow) produced audible glitching which was cured by slightly shifting one of the points.
Editing in the View Window allows you to work on only one Sound File at a time. You can cut out some of the material and play through what is left, but you can't re-order it. That is the job of the Mix Window.
The Mix Window (Figure 5) consists of five columns and a number of rows, each row representing one segment of audio. The first column gives the name of the segment, eg. 'Intro.view'. These are the View Files you have created. The second column shows horizontal bands representing the length of the segment and its position in time in the Mix File. Column three allows a segment to be muted, and column four is a gain control which allows the level of the segment as a whole to be adjusted. The last column contains a pan pot. Once again, this window has a pair of cursors, used to play back sections of material as in the View Window.
View Files can be placed in the Mix File by 'Paste To Mix', which inserts the segment at the start of the mix; 'Abut To Mix', which places it after any segments which are already in the mix; 'Crossfade To Mix', which does an Abut, then moves the segment back in time by a preset amount so that it will fade in as the previous element fades out.
With a few segments - View Files - in the Mix Window, perhaps verses, chorus, middle eights etc of a song, you can start shuffling them about. Moving the segments is achieved with the Clipboard. Segments can be Cut to the Clipboard or Copied. 'Cut' means that you move the segment itself to the Clipboard, leaving nothing behind; 'Copy' means that the segment is copied and that copy placed in the Clipboard. The contents of the Clipboard can be Pasted to any point in the mix.
Joining the segments together as butt edits is straightforward and they just play one after the other with no gaps. If they are to be crossfaded, or their level changed, then there is work for Dyaxis to do.
As I remarked earlier, Dyaxis is a two channel device. Also, without the DSP card, it has to perform any processing out of real time. These two facts mean that if you, say, wanted to crossfade between two segments, Dyaxis is going to have to add them together digitally and create a new file (a Mix Output File) before it can play. Figure 6 shows the Mix Window display of a crossfade between two segments.
Calculating the output file fully every time you wanted to audition a crossfade would be a time-consuming business. Fortunately, there is a 'Fast Mix' function which only computes the section you have changed, and only for the part between the cursors that you have requested Dyaxis to play. During the computing process, there is a handy display of how long the task is expected to take and you can abort the procedure if it looks like you have bitten off more than you would like to chew. Fast Mix files are short-lived affairs, for auditioning purposes only. When all is as you want it, it is necessary to revert to standard mixing. Playing back one more time will create a finished Mix Output File which you can dump to DAT as your finished edited master.
Well yes it does, and I wouldn't have expected a company like Studer to get involved in a product if it didn't. Obviously the system has its limitations, principally one of speed - the computer needs to recalculate virtually every time you make a change to a mix. There are some detail minus points, such as there being no indication that you have added together too much audio resulting in digital clipping. There is also no scrolling of the display during Mix Window playback. After a few segments, you are off the screen and lose the benefits of being able to see what you are doing. There are a couple of other niggles, but at the low price - compared to many other hard disk systems - you have to consider what it can do.
What Dyaxis can do is help you make perfect edits in stereo material. There is limitless precision in finding the right edit points, and very easy auditioning of those points. The crossfade function makes it possible, given a little time and experimentation, to join virtually anything to anything. The variable gain and the envelope shaping make it straightforward to match segments of audio with otherwise incompatible levels. In short, Dyaxis is probably the perfect entry point to the world of digital editing.
FURTHER INFORMATION
Dyaxis DA320 (320Mb, 32 channel minutes) £8900 +VAT.
Dyaxis DD1200 (1200Mb, 120 channel minutes) £24,000 +VAT.
These prices vary according to exchange rate and do not include the cost of the Macintosh computer.
FWO Bauch Ltd, (Contact Details).
IMS Dyaxis - 16-bit Digital Recorder
(MT Aug 87)
Browse category: Hard Disk Recorder > Integrated Media Systems
Gear in this article:
Review by David Mellor
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!