The new Reality version 1.5 has lots of new features, including new synthesis types, and will now run on any DirectSound compatible soundcard.
This month sees a positive flurry of new software, ranging from full‑price releases to freeware. Martin Walker sifts through the latest offerings.
The long awaited version 3.6 update to Cubase is finally here, and it's free to existing users. Although there are a couple of download sites, the fastest for UK users is probably www.steinberg.net/news/cubase36.... Be aware, however, that there are three different update files, depending on whether you have VST, VST Score, or VST Audio, and you might run into difficulties if you attempt to update with the wrong file.
The new version of Cubase VST Audio has some extra options. It now becomes Cubase VST/24 3.6, and is capable of 24‑bit, 96kHz operation with suitable hardware. 16‑bit and 24‑bit files can be used together freely, and you can switch between 16‑ and 24‑bit recording at any time. VST/24 also provides extensive support for the Yamaha DSP Factory, as I described in my review of the DSPF in the October issue. One or two DS2416 cards are supported, and each card can also be used with one or two AX44 expansion units — an extra PDF‑format manual is provided with all the details. Owners of the DSP Factory card, or indeed anyone who wants to change up to 24‑bit operation can move sideways from VST to VST/24 for £149, and from VST Score to VST/24 for £119 (both prices include VAT).
Other new features are common to all three versions. Since this is a free interim update, the bulk of the new goodies which I found in the Macintosh Cubase 4.0 version are not yet implemented — there are no screensets, MIDI Track mixer, or Marker Track. However, what new features are available are implemented in arguably the most important area — audio:
- Maximum number of audio channels is increased from 32 to 64 (96 for VST/24).
- Individual channels now have eight effects sends rather than four (but there are still four channel Inserts and Master effects).
- Eight stereo Group channels have been added for submixing, with their own mixer.
- Rebirth version 2.0 can be integrated using the new Rewire option (see the SOS review of the new version in the Nov '98 issue).
- There are new non‑destructive Audio Crossfades, as well as Fades In and Out.
In addition, the ASIO Multimedia driver now has an extensive test routine built in, which launches automatically after the update, and whenever any audio settings are changed. Many people's problems with Cubase VST have been largely due to factors outside its control, such as soundcard drivers and the Windows Multimedia system. Steinberg are obviously taking this aspect a lot more seriously, and have even included a six‑page manual explaining what the utility does, and what to try if it fails. Although previous versions of VST included a test to find the optimum buffer size, it tested only individual ports. The important distinction is that the new Audio System Test opens every currently activated input and output port simultaneously, and then simulates data flow for the whole system. The test is passed only if all ports can transfer data correctly with no loss of sync from the reference port.
If you have problems with lost data, you can selectively disable a few hardware input or output ports, and can try the test with different combinations of MIDI ports activated in the Setup MME utility. This shouldweed out those machines that can't really cope with a large number of simultaneous hardware inputs and outputs. It will be interesting to see how successful this is out in the field — there will be a lot of happy users if MIDI + Audio timing problems can be detected before people start recording. Before you decide to upgrade to the new version of Cubase, however, be warned that many people are experiencing teething troubles — see the 'Cubase Problems' box on page 236.
More Real Than Ever
Partition Magic 4.0 (see page 234) now runs directly from Windows 95/98 and NT4, and is even quicker and easier to use.
As I mentioned last month, Seer Systems' Reality software synth has just reached version 1.5, and the upgrade is free to all existing owners, who can download it from Seer's web site (www.seersystems.com). I've now received the new boxed version, which doesn't use a dongle at all — apparently, existing users hated it. Reality 1.5 now has DirectSound support, which releases it from its previous restriction to SoundBlaster cards. Any soundcard with DirectSound drivers can be used to run the new version of Reality, and if you can find a multi‑channel one, you should be able to devote a single stereo pair to software synthesis and leave the other channels for normal audio playback.
Another fundamental but welcome change is that the software now includes SoundFont support, which instantly expands its sound library. SoundFonts are slowly but surely gaining recognition as a standard way of specifying the WAV files, loop points, filters, envelopes and LFO settings that comprise one or more complete 'instruments'. Although originally used only by SoundBlaster‑compatible cards, they are a quick and easy way to get a multisampled sound that can then be treated through the powerful Reality engine, complete with a host of filter and synthesis types.
Several new synthesis options have been added including DeltaString (a plucked string model), and an additional 1‑into‑3 topology for FM synthesis, for fatter sounds. The carrier in FM patches can now also be a sampled sound, for even wider flexibility. Another tiny but welcome addition is that the space bar can now be used to trigger sounds as well as MIDI, which is helpful when sitting in front of the PC.
There is a new MIDI page that allows Patch, Volume, and Pan to be set for each of the 16 multitimbral channels, and includes individual channel Activity indicators. The maximum polyphony has been increased from 64 to 128 voices (though I suspect that this will be academic for all but cutting‑edge PC owners), and a GM mode has been added to restrict MIDI channel 10 to drum sets. Audio level and CPU meters now show real‑time values, and a MIDI sequence player allows standard MIDI files to be imported and played back (ideal for stand‑alone demos). The included demos prove once and for all that Reality is perfectly capable of simultaneously generating every sound for a complete multitimbral song in real time. You can even grab a complete Reality performance as a WAV file using two new Audio Capture Modes, either using the Stop/Start controls of your sequencer, or when Reality's output levels reach a predetermined Threshold value.
Finally, Seer Systems have come up with SeerMusic — a freely downloadable web browser plug‑in that plays streamed Reality‑based audio via any DirectSound compatible soundcard. Owners of Reality can create music to post on a web site, whereupon anyone who has the plug‑in can hear it in real time. Posting music files on web sites requires a little JavaScript, but the required scripting lines are included in the Help file and can simply be pasted into your web page's source code, with the filename changed to point to your own SeerMusic file.
Since the clever stuff is in the browser plug‑in, and a standard bank set is provided which already contains hundreds of patches (including samples), downloading the music itself takes up only a fairly low bandwidth. However, this doesn't compromise sound quality, which remains solely dependent on your soundcard.
Although Reality will run on a Pentium 166MHz MMX if pushed, you really need a more powerful machine to run a sequencer alongside. Its main strength is the wide range of synthesis types available — I'm currently using it to add FM bells, gongs, and pianos to my music, but this is only the tip of its sonic iceberg. The full retail version is now £299.99 including VAT (for once this looks good for UK users against the $495 US price). Reality may look rather expensive compared with Native Instruments' Generator (£170) and Bitheadz' Retro AS1 (£150), but neither of these can match its wide range of synthesis types and physical modelling capabilities. Contact Cimple Solutions (+44 (0)181 904 4141) for further details.
New And Improved
Coagula in action (see page 236) — it takes a few minutes to construct images like this, but wait until you hear them!Well, they're finally here. Event have posted their PC ASIO drivers (version 4.0) for the Layla, Gina, and Darla soundcards on their web site, which include a new enhanced Echo Console. To download, point your browsers at www.event1.com.
A new version (4.0) of Partition Magic is now available (around £50 for the full version, and £29 to upgrade from version 3.0). The biggest change is that the utility can now format, partition, re‑size, and change cluster sizes directly from Windows 95, 98, or NT 4.0, without having to reboot into DOS as before. You can also change freely between FAT16 and FAT32, and FAT32 partitions over 8Gb can now be created, re‑sized and moved. Apart from making the reviewer's job a bit easier (you can now easily grab a screenshot!), the utility is much quicker to use, and is made easier by the addition of Wizards which can automate most common tasks with step‑by‑step instructions. Various changes can be previewed on multiple drives, and the Apply button then carries them all out. If the changes are to drives other than the one running Partition Magic, they will be carried out immediately, but otherwise the system still has to reboot to do this.
A new supplied Boot Magic utility replaces the previous Boot Manager, which required its own separate partition to run. Boot Magic doesn't suffer from this restriction, allows one of several operating systems to be chosen from a graphic selection screen during startup, and is ideal for those of you who want to have both Windows 95 and Windows 98 on your PC. Partition Magic 4.0 can be obtained through all the usual PC software suppliers.
Cubase Problems
Since upgrading to the new Cubase VST 3.6 on my PC, I've discovered various teething troubles. The most dramatic is an apparent crash that may happen on some PCs when you first launch the program. I say apparent because Steinberg UK told me that if you leave Cubase long enough it should recover, and next time the freeze happened to me I timed it — after 45 seconds it carried on initialising correctly as if nothing had happened. A public beta version of the MROS32.DLL file is available from the Steinberg ftp site which claims to cure this problem, but it leaves your MIDI outputs incorrectly labelled in the Arrange page — mine read 81, 82 and so on rather than the names I gave them, though this is only cosmetic. Another cosmetic defect with both versions of this file is that if you switch on Follow Song using the drop‑down menu option, it leaves a corrupted cursor on screen.
More seriously, audio playback won't start at all unless you have a MIDI output device activated using the Setup MME utility (audio‑only users can add a dummy device such as Hubi's Loop). After scouring the newsgroups to see if anyone had discovered anything else, I came across the very informative Timo's Cubase VST PC Homepage, which now carries a full up‑to‑date list of 3.6 bugs reported by users (you can download the MROS32.DLL beta file from this site as well). Take a look at www.geocities.com/SiliconValley/... ‑‑ you may decide whether you want the dust to settle a bit before upgrading.
Coagula: Painting With Sound
Ever since attending the Frankurt Musikmesse early last year, I've been envious of Mac owners who could run Metasynth, a graphics package that allows you to render its images as audio samples. I'm now happy to report that PC owners can try out something along similar lines. Coagula is described by its author Rasmus Ekman as an image synthesizer, and is currently available as a basic freeware demo. Even at this embryonic stage, however, it still manages to generate sounds that are totally different from those produced by any other form of synthesis. As you might expect, this is resolutely non‑real‑time, and a typical five‑second sample takes a few seconds to render (depending on how complex the image is). It works by adding together the output of a huge number of sine‑wave oscillators — the position from bottom (low) to top (high) of pixels in the image determines the frequencies used, and they way in which they change over time is mapped from left to right. For instance, a single horizontal line would produce a steady tone, while a set of dotted horizontal lines would produce a rhythmic chord. Red pixels control the left channel, and green ones the right.
Once the basic image has been drawn, you can use filters (BMP files) to modify the image, and in fact you can save your images to use them as filters. You can render the current image as a WAV file at any time, and within a couple of minutes I seemed to have resurrected the BBC Radiophonic Workshop! With a SoundFont‑based synth (such as the SoundBlaster series, or Seer Systems Reality), I could then turn these already complex aural landscapes into a polyphonic synth. Wonderful! At only 70K, everyone should download this immediately from hem.passagen.se/rasmuse/Coagula.htm, and support this talented programmer when the full version comes out. If you want an even bigger incentive, Rasmus has also written GranuLab (there is a link to this from the Coagula page), which lets you try out granular synthesis.