Are you a DSP junkie who's always trying to squeeze just a bit more processing power out of your computer? Well, Soundscape's Mixtreme multichannel audio

Are you a DSP junkie who's always trying to squeeze just a bit more processing power out of your computer? Well, Soundscape's Mixtreme multichannel audio card may be the solution you're looking for. Mixtreme combines up to 16 channels of digital audio I/O with a Motorola 56301 processor that runs a user-configurable 16-bus mixer. The software offers one or more 2-band parametric EQs per channel and support for proprietary Soundscape plug-ins. Mixtreme's mixer and its family of plug-ins are actually borrowed from Soundscape's SSHDR-1 hard disk recording system, which has a well-established reputation for quality. In fact, the Mixtreme card can be used with the SSHDR-1 for added I/O and DSP power.

Mixtreme's onboard audio I/O consists of two 8-channel digital TDIF ports along with word-clock/Superclock In and Out. For anyone with a TDIF-compatible digital mixer or MDM, Mixtreme is a great value right out of the box. For other users, Soundscape offers a variety of interfaces to suit many needs. The optional hardware ranges from a simple S/PDIF daughterboard to the high-end SS8IO-1 with balanced-analog and ADAT Lightpipe I/O.

MIXING IT UPThe heart of Mixtreme is its software mixer. It runs entirely on the Motorola processor to prevent any strain on the host computer's CPU. Because of this, the mixer is quite responsive, and changing settings while recording or mixing doesn't bog down your system. The mixer is also completely configurable, so you can have only what you need onscreen, thereby reducing screen clutter and DSP consumption. Of course, it's unlikely that you'll run out of processing power any time soon. The hardware provides enough juice for almost any combination and number of mixer controls you might want, although the effects plug-ins can eat up processing capacity.

For this review, I tested version 1.05 of the Mixtreme software. By press time, Soundscape had released version 1.06, which incorporates surround mixing and panning to support the Dolby Surround Encoder/Decoder plug-in.

Soundscape's buzzword for Mixtreme is low latency, in part because the system's DSP-based mixer nearly eliminates record-monitoring latency-even when using effects. Furthermore, Soundscape has included ASIO 2.0 and GigaSampler drivers (in addition to the standard multimedia drivers), for which it claims extremely low latency with the appropriate applications. It's worth noting that these are all multiclient drivers, making Mixtreme's resources available simultaneously to multiple applications. For example, you can run a sequencer such as Steinberg Cubase VST or Emagic Logic Audio while generating sounds from a software synth such as NemeSys GigaSampler or Seer Systems Reality.

Mixtreme's mixer is a blank slate on which you can map just about any sort of routing arrangement you can imagine. A handy toolbar provides one-click access to all of the important commands, such as creating, moving, and deleting channel strips, aux sends, faders, and meters, or assigning inputs and outputs to various elements. By clicking on an additional tool button you can change the mouse mode from Edit to Control. At first, I wished I could right-click to edit and left-click to control, but Soundscape has implemented right-clicking in a different but quite useful way, as you'll see. Instead, each of the tool buttons has a hotkey equivalent.

Soundscape has done a nice job of designing an efficient mixing interface, squeezing a lot of information into a small space while providing enough room for precise control of all parameters. Moreover, Mixtreme makes good use of a mouse's five basic functions. Drag a fader in the normal way, and it moves as you'd expect; right-click and drag a fader, and all faders in its fader group move with it. Similarly, you left-click a solo button to solo that channel, and right-click to solo all channels in the same solo group. The mixer provides up to 99 solo groups and 99 fader groups, which you assign by clicking in the small boxes labeled SG and FG, respectively. As is the case with most of Mixtreme's value boxes, right-clicking a group assignment box increments the value, left-clicking decrements the value, and double-clicking opens a dialog box into which you can type the group number.

Double-clicking on any effect opens its edit window, double-clicking on a mixer element opens a dialog box (when one is appropriate, as in pre- or postfader selection for a send), and the Pan control even provides a tiny "center" button. If you must mix with a mouse, this is a pretty good way to do it.

My only big complaint with the interface is its lack of Copy and Undo functions. Once you have a channel strip set up the way you want, it would be great if you could generate seven more just like it by pressing Control-C and Control-V a few times. Instead, you have to create all seven copies from scratch. Even more annoying is the inability to undo actions. The Create tool has a hair trigger, and I found myself selecting the wrong thing more often than I care to admit. In the long run, you'll probably end up creating a handful of standard configurations that you'll use repeatedly with only minor edits, so you'll miss these functions only at first. Still, you may miss them a lot.

GO CONFIGURETo see just how configurable Mixtreme's mixer is, let's look at a couple of examples. Fig. 1 shows my setup for basic tracking of solo parts. I separated the input into two different channel strips so I could control the input level separately from the monitor level. Channel 1 has two prefader aux sends: one sets the level of the incoming audio going to bus 5-6 for the headphone mix, and the other sets the level of incoming audio going to my digital audio sequencer via bus 3-4. Channel 1's main fader controls the level going to bus 1-2 for the control room mix. Channel 2 is the destination of the record-level send and contains just a Track module corresponding to the input I've selected in my sequencer. Channel 3 mixes the output track from my sequencer with the live audio, then sends it to physical outputs 3 and 4 for the headphone mix.

Granted, it's a pretty rudimentary arrangement, but it took only a few minutes to create it from scratch, and it represents a big improvement over the control that many sound cards provide. I've already thought of improvements I could make to this setup, and that's really the point: the mixer is what you make of it, no more, no less.

Fig. 2 shows a much more elaborate setup. To test Mixtreme's limits, I decided to entertain the classic engineer's nightmare of a finicky four-piece horn section with each player wanting more of himself or herself in the headphones. The classic technique for this "more me" scenario involves patience, diplomacy, and occasionally firearms, but Mixtreme can indeed provide an alternative. In channels 5 through 8, I've set up four stereo buses, one for each player's headphone mix. The input channels, 1 through 4, each have five prefader aux sends. Notice that the level of the lead trumpet's send to his or her own headphone bus (bus 1-2) is centered and maxed out, while the level of the send to everyone else's headphone buses is panned off-center at a more moderate level. Each player's input strip is set up the same way, putting them dead center and way above the din in their own personal mix.

As convoluted as this setup seems, it didn't even begin to tax Mixtreme's available DSP. I remembered that most players don't want to hear themselves dry in their headphones, so I added some reverb to fill out the sound. This entailed adding one more bus (bus 9-10), with a send from each input channel. I inserted Wave Mechanics Reverb into the reverb-only bus and set its output to 100 percent wet. Then I set up sends to each individual headphone mix, so I could give the players as much or as little of the reverb as each wanted. After all this, I was still using only 70 percent of Mixtreme's DSP capacity and 80 percent of its memory.

EFFECTIVE MIXINGI had the opportunity to check out some of the effects written for Mixtreme's DSP. While none of the plugins are included with the basic Mixtreme package, it's important to consider them as part of the whole system. Few users will be content to let all that excess processing power sit idle.

Overall, the effects are excellent, although some of them can eat up a lot of Mixtreme's available resources. The Wave Mechanics Reverb, for example, required about 43 percent of the DSP's processing power and 32 percent of its memory. In most mixer setups you'll have the resources available to run one copy of this effect, but running two would leave room for little else. To keep this in perspective, though, ask yourself how many top-quality hardware reverbs you could buy for the same money and how many spare aux buses you have available for them on your console. Besides, if you really need one more reverb, you can easily set up Mixtreme with an aux send/return to an external effect.

What do you get for that 43 percent? Quite a bit, I'd say. Reverb was written for the SSHDR-1, and it's a well-developed plug-in with plenty of power and control. You can vary the decay time of high and low frequencies relative to the middle band, and you can even control the crossover frequencies. It sounds great and includes a good variety of presets.

Soundscape's Audio Toolbox provides a useful set of common effects, including a chorus/flanger, a two-tap delay, and a dynamics processor. Their DSP needs are modest-all under 10 percent, with the exception of a long (up to 1,088 ms) delay, which requires more than 75 percent of Mixtreme's memory. The interfaces for the three devices are simple but functional.

Two effects plug-ins from TC Works are also available for Mixtreme. TC Reverb is everything you'd expect from TC Works; it lets you choose room shape and size and control decay across three frequency ranges. TC Reverb eats up only 29 percent of Mixtreme's processing power but consumes 52 percent of its memory, so you can't use more than one instance of it. TC Dynamizer is a 3-band dynamics processor based on TC Electronic's Finalizer. This is the sort of processor that you're most likely to use after mixing to a stereo output file, so its power needs are less consequential. However, because it uses 40 percent of Mixtreme's processing and 31 percent of its memory, you might be able to squeeze it into your mixer's master output bus if you really want to.

INS AND OUTSMixtreme comes with no analog I/O, so if you don't have a TDIF-compatible device (such as a digital mixer or Tascam DA-series MDM) to use as a front end, you'll need one of Soundscape's optional interfaces. The most basic of these is the S/PDIF daughterboard, which uses the RCA connectors that would otherwise provide word clock or Superclock I/O to Mixtreme. Of course, this still doesn't get you all the way to analog-you'll need an S/PDIF-compatible external D/A converter, as well.

The SS8IO-3 provides a direct route to and from the analog domain, with eight channels of 20-bit I/O on RCA connectors. This half-rackspace unit attaches to one of Mixtreme's two TDIF ports by a 6-foot ribbon cable, which provides a reasonable amount of placement flexibility. The front panel features two-segment LEDs for each channel, indicating levels of -30 dB and -3 dB. The front panel also provides selectors for sample rate and clock speed. (A 2-channel version of the SS8IO-3, called the iBox 2-Line, was released too late for this review.)

The SS8IO-2 comes in a nearly identical unit but offers eight channels of ADAT Lightpipe I/O. Like its sibling, it has sample-rate and clock selectors on the front panel. Soundscape has designed the unit to pass the full 24 bits that Lightpipe supports, even though current ADATs are only 20-bit devices.

If you have more industrial-strength needs and your budget is sufficient, the SS8IO-1 is the way to go. Its eight channels of analog I/O are provided on balanced XLR connectors, and the front panel sports ten-segment LED meters. The unit also provides a Lightpipe connection for swapping bits with ADATs, and it offers word-clock/Superclock support like the others. According to Soundscape's Web site, the SS8IO-1 is being replaced by a 24-bit model, probably by the time you read this. The new model is called the iBox 8-XLR/24.

I tested both the SS8IO-2 and SS8IO-3, and it's hard to find fault with either. The converters on the SS8IO-3 sounded great both coming and going. The SS8IO-2 was a snap to set up for ADAT transfers. No muss, no fuss, just audio in and audio out. What more could you ask for?

COMPLAINTS DEPARTMENTMixtreme is a powerful system, and it performs its tasks quite well. Unfortunately, its talents don't currently include MIDI control. That's unfortunate because you can't save and recall mixer snapshots or automate mixer moves from your sequencer. External control surfaces could be a great part of a Mixtreme system, but for the time being they're useless. (Soundscape has informed us that it has already written MTC sync and MIDI-controlled mixer automation into the next software release. An update should be available by the time you read this.)

Although it supports word clock and Superclock for synchronization, Mixtreme lacks the ADAT nine-pin sync connection necessary to lay tracks back to tape with sample-accurate positioning. If it had MIDI I/O (or if you have a separate MIDI interface), you could use MIDI Time Code (MTC), but of course that is not sample accurate.

The documentation for Mixtreme is marred somewhat by excessive "marketingspeak" and sloppy editing, but Soundscape has, nevertheless, made a good effort to offer thorough and accessible content. In particular, I was pleased to find that the user manual explains its mixer structure and the workings of Audio Toolbox and Reverb. The documentation for the SS8IO-2 and SS8IO-3 provides only essential technical details. It would have been nice, especially for the SS8IO-2, to have some examples of setups and applications.

All in all, Mixtreme is a great choice for anyone who needs to get 8 or 16 tracks of audio into and out of a computer. It's at its best as a complete system-the plug-ins add a lot of value if they fit your budget-but even the basic setup has plenty of appeal. With low-latency monitoring, a "have it your way" digital mixer, lots of I/O, and room for expansion, it could be a great foundation for a personal or project studio.

Brian Smithers is a musician, conductor, and arranger at Walt Disney World. His Web site,, covers making music with notebook computers.