If you're one of those people who enjoys staying on top of the latest audio gossip, then you're bound to have come across the topic of analog summing. Like any good digital-versus-analog debate, the topic has ardent supporters on either side of the issue as well as a vast majority in the middle who are either confused or indifferent. Essentially, the concept of analog summing is the idea of mixing digital tracks through some type of analog mixer or dedicated summing device. But before getting too deep into the subject, a little back story is in order.
When all-in-computer recording (digital audio workstations) arrived, it changed everything. For the first time on the project-studio level, music could be recorded, edited and mixed quickly and cheaply using an off-the-shelf home computer. And since that time, bedroom musicians have grown accustomed to (and even dependent on) many of the luxuries afforded by such systems. Sound quality, however, has remained a bone of contention. The issue is less controversial now than in the early days — when analog purists and digital enthusiasts engaged in heated debate about what was to become of the future of audio recording — but it hasn't gone away.
Because a high standard of sound reproduction was attained years ago using quality analog gear, engineers who respect technical perfection are loath to give up any degree of audio quality just for the sake of digital ease. Thus, after years of rising bit and sample rates, endless revisions of software niceties like plug-ins and dither, and better A/D/A converters, along came the idea that computer recordists could better their systems by employing the process of analog summing. In this context, that means taking individual digital tracks and stereo submixes (or stems), running them out through separate D/A converters and mixing them back together in the analog realm before recording the master.
The idea is that less math is performed in the computer (which, apart from summing of tracks, is dependent on factors such as whether the audio is output at unity gain and whether plug-ins are being used); therefore, less is being done to alter the recorded tracks. And to its believers, the employment of analog circuitry is more of a known quantity than the “digital voodoo” that is the proprietary product of a secretive software engineer. Proponents of digital summing will point to many mixes that were done entirely in the box and sound good. The results of analog summing are arguably incremental no matter who you talk to. In-the-red rockers and crunked-out laptop kids are probably going to forever remain nonplussed at the whole debate. But for those in search of the holy grail of sound, every step in the right direction — no matter how small it may seem — is an important one.
The idea that completely in-the-box summing had its limitations is nothing new to digital audio users. For several years now, it has been fairly common practice to send separate audio tracks or groups out through individual D/A converters and then mix the converted audio in an analog console. And, of course, the better the console, converters, outboard processors, cables, gain staging and so on that you use, the better the results will be.
Like many others, I currently perform this type of mixing from my Apple Mac G4. I use an RME Hammerfall 9652 to run 24 channels of ADAT Lightpipe I/O to a Tascam MX-2424 equipped with the ADAT and analog option cards. This allows me to then run the channels out analog to my Soundcraft Ghost console. I also have a patch-bay setup, which lets me use my analog compressors and other outboard gear. This isn't the same as the analog-summing devices purveyed by Dangerous and Roll Music. In my case, I'm running the recorded audio through a big chunk of the board's circuitry (thus degrading the sound to a degree), but on the other hand, it presents me with a good degree of flexibility. Apart from being able to use the board's EQ, I can, for instance, bus the drum tracks out as a stereo pair to a compressor before it hits the console, or ride an analog fader (rather than a digital one using a mouse).
As for digital automation, I'll often work on rough mixes entirely in the digital domain. After getting the automation right, I'll then bus stems out to the console through various analog outboard gear and set relative levels on the Ghost. Ultimately, the flexibility is the key, and later on if I decide to mix digitally and compare the two, the bulk of the work is already done.
The device that arguably got the ball rolling in the realm of strict analog summing is the 2-Bus from Dangerous Music. Designed by former mastering gear manufacturer Chris Muth — who runs Dangerous Music Studios in New York City with his partner, Bob Muller — the 2-Bus provides transparent analog summing with a makeup gain stage using the balanced connections and minimal circuit pathways to perform the summing. The 2-Bus is just one component in a Dangerous system that also includes a monitor section, a talkback and metering system and an 8-channel stereo line mixer. Dangerous also makes the 2-Bus LT, a less-expensive version that uses DB-25 connectors.
The way Muth sees it, the progression toward analog summing tends to go like this: People first get rid of their consoles because they want more room and they can get any flavor of mic pre they want. For the most part, they'd rather not deal with analog equipment if they can avoid it, and they really want the editing and the features that a DAW can give them. “So they get a DAW system and ditch the rest, and then they start to find the limitations of the DAW, which mostly have to do with timing problems and sonic issues,” he says. “Some people say that they don't have any sonic issues at all with digital, and I say that's great. But I do.”
The 2-Bus itself began life when a friend of Muth's made the switch to digital and didn't like what he was hearing. “The issue is, when you bring your faders down to — 30 or — 40 dB in the computer, you're throwing away a whole bunch of information,” Muth says. “We built a system where you could run six or eight faders at — 6 or — 10.” Muth built the 2-Bus prototype back in the days of 16-bit digital. “As the DAW systems got better and they had higher resolution inside the box, a lot of people predicted that we wouldn't need to have these external summing boxes anymore,” he says. “Fair enough, but I thought that if the digital stuff got better, the math is going to stay the same and an analog summing bus is going to sound even better.”
But, he contends, that's not all that an analog summer can do to help the sound. Another important contention for the rivals of the digital-summing debate is that latency occurs when internal processing is introduced, resulting in phase shifting or outright mistiming of tracks. Muth is adamant that analog summing helps get rid of latency problems. “People may point to their automatic latency controls, but I've watched people mix this way, and it takes them 15 minutes to figure what latency is on which channel,” he says. “And things change again when they make changes to the processing. With a 2-Bus, I can say, ‘I have a stem mix, and I want to compress it.’ I patch in a pair of compressors, and in about 15 seconds, I have it dialed in — all without delays.”
That last comment is paramount to Muth. “No matter what anybody says, if it goes back into an A/D, it's not going to be in time anymore,” he says. “So you wind up having to slip tracks around, and it's a pain. Whereas with a patch bay and a compressor, you go click, click, click and turn a few knobs, and you're ready to go.”
ROLL MUSIC FOLCROM
When Justin Morse of Roll Music, a recording-studio owner and boutique manufacturer in Minneapolis, made the first prototype for what eventually became the RMS216 Folcrom, it was because someone called him wanting a summing device. Morse took the minimalistic approach, building a completely passive device that didn't even require external power. In the Folcrom, the summed channels require makeup gain from a mic pre, which in itself sets the unit apart from other similar devices.
“I thought that you shouldn't have to pay for all these knobs on the mixer if you don't need them, because you can control everything in the computer,” he says. “The other thing is that there's no reason to have your audio go through all this extra circuitry that it doesn't need to go through. Dan [Kennedy of Great River Electronics] had the idea of getting the makeup gain from a mic preamp, since he makes them.”
Morse started working on the design; it originally used transformers on both the inputs and outputs. “I started realizing that I could get rid of those, too,” he says. “With a symmetrical, balanced, floating resistive summing network, there's no reason for any of that stuff. So the design just kind of worked itself out really elegantly; we just kept on removing parts from the circuit concept until there was almost nothing left. That's kind of how I've always approached circuit design: to use as few components as possible to leave the most basic, simple circuit that I could get at, which worked out really well with the Folcrom.”
Morse cites another benefit of the less-is-more approach: “The whole idea is that when you create a mix in a computer — which, of course, has the ability to recall it exactly every time — it's counterproductive to have anything on the external mixer that could not be set precisely and recalled exactly. So the Folcrom exists with a fairly narrowly defined use. It's for people who like mixing in the box but don't want the sound of the box.”
And what is the “sound of the box” according to Morse? He believes that there are two schools of thought on the matter, at least among people who find fault with it. “Some people say there's something wrong with the algorithms; others say it's perfect in the digital domain, yet they don't like it, so they want to introduce a little ‘warmth’ or ‘euphonic distortion,’” he says. “I suspect it's the former line of reasoning: that there's something still wrong with digital summing. Now, I have heard people say that this whole business must be a lot of marketing hokum because everybody knows that it's just math and a simple addition of bits, and there's no way it can't be perfect in the digital domain, but I think people have been underestimating [certain people's] ability to screw up the math for a long time.”
In the course of designing the Folcrom, Morse performed a series of comparative listening tests, taking every precaution to make sure that the final mixes were as close as possible and that the reference level between the two was the same within a tenth of a decibel. The results, according to him, weren't even close. “The detractors of summing say that we're simply adding distortion and noise to the mix,” Morse says. “And that definitely happens; there is distortion and noise added when you run audio into an analog mixer. But it becomes very difficult to listen to one and then the other and say, ‘Well, this one's missing the distortion that would have been caused by the internal summing,’ because there's other distortion there. It's really hard to say whether there's masking going on.
“I suspect that in the analog mixes, it's primarily lower-order, even-order harmonic distortion,” Morse continues, “whereas in the digital mixes, it's higher-order things that we're less used to hearing historically, and that bothers us more. [Maybe] those things are still there; they're just covered up more by the good stuff.”
SPECK LILO AND XTRAMIX
Some companies build products that exceed initial expectations. Speck Electronics' Xtramix, for example, has been in production for about 12 years and was originally intended as a keyboard mixer, but over time, its purpose has changed. According to Speck's owner and chief designer, Vince Poulos, “The consumer seems to set the demand and the application for this type of product. The Xtramix started as a keyboard mixer, but in the last couple of years, most of the units sold are being used as a back end to a workstation. The new aspect of the Xtramix's use has to do with people's feeling that mixing within a workstation just isn't cutting it for them. I have to tell you that I'm not a complete believer that that's the case. But some people say it is, and I'm not going to disagree or argue with them.”
Speck sold the Xtramix for several years, and now comes the LiLo, which is a mixing console stripped of most everything one typically expects in such a device. He stresses that transparent reproduction of audio is the key feature in a mixing platform. “Because I'd been making mixers for a long time, I started getting e-mails from people asking for a minimalist mixer,” he says. “I gave it some thought and put a concept drawing on our Website to solicit ideas from people. I got a lot of great ideas, and what kept coming back is that people wanted a clean, neutral mixing platform that didn't have anything in it — a feature-free mixer, if you will.”
Poulos says that he wanted to have a mixer that could not only handle external mic pres and EQs — the gear that users have already been collecting — but also have the ability to use it with their workstations. So the LiLo has two functions: It allows you to connect 16 channels of converters and simultaneously plug in all of your favorite mic pres and EQs and blend it all together.
A BRIEF SUMMATION
It appears that, judging by these manufacturers' opinions, there is still plenty of research and arguing to be done before anyone comes up with a definitive analysis of analog versus digital summing, let alone a “winner.” Regardless, designers have and will continue to develop niche products that provide additional high-quality analog-mixing flexibility to computer recordists. And at the same time, DAW manufacturers will continue to fine-tune their products and develop new technologies.
How you like to work may ultimately be the deciding factor in whether you wish to incorporate analog summing in your system. But if you already have enough mic pres to get the audio into your computer in the first place — and could do without the expense, hassle and space necessary to own an analog console — summing devices provide an elegant way of getting your audio back out of the box after you've gotten it in. Whether you're looking to avoid excess processing and impart the sound of circuitry or just have a new pathway to patch in your analog gear, analog summers provide more choices and greater potential during mixdown.