The Sum of All Fears
by Mitch Gallagher, for EQ Magazine / July 2003

Over the years there's been quiet muttering in dark recesses of studios . . . whispered questions about the sound quality of DAWs . . . fearful queries about that dread topic: summing.

Far from a topic to be discussed under cloak of darkness, summing is simply the process of combining (mixing) signals. From the dawn of the DAW age, there's been question: Does summing in a computer really sound as good as that done by a mixing console?

There are a ton of factors that figure in, one being how much math is done in the DAW. Every action, whether mixing two signals, changing a channel's level, panning a channel, whatever, requires calculations that may impact the quality. According to some, if you simply bring two signals in, mix them at unity gain, and send them back out, there's going to be impact on the signal due to the way in which computers handle dither and binary addition.

This has made some engineers leery of mixing "in the box." Some swore they could tell when a song was mixed in, say, Pro Tools, by the sound quality alone. To counter this, some pros send individual channels (or submixes of channels) to an external mixer to create the final stereo mix. DAW supporters, on the other hand, have countered that when combining digital signals, the mix process shouldn't result in a "sound" — two signals combined in Nuendo should sound like two combined in any other DAW. It is, after all, just two numbers being added together.

Doing a comparison of the sound of DAWs is a daunting task. The problem is eliminating variables when you're dealing with disparate platforms. Lynn Fusion of 3D Audio in Franklin, TN, has undertaken the first stage of this; his results are now available on CD. (You may be familiar with Lynn's mic and mic preamp comparison CDs.)

Called the Awesome DAW Sum Sampler, the 2-CD set contains two types of comparison tests, both done with 24-bit, 48 kHz sound files. The first test has 24 tracks being mixed with all faders (including the master) set to unity gain — you should hear only the summing, with no other math applied. In the second test, the individual faders are pulled down 3 dB, while the master is up 3 dB. The same test was done on 29 platforms, including 15 DAWs, five digital mixers (Panasonic. Yamaha, TASCAM, and Sony DMX and Oxford), and five analog mixers (SSL, Trident, Yamaha, Manley, and Dangerous 2-Bus)

Awesome DAW Sum contains the stereo output files from each platform. The files are unidentified; to learn which platform is which file, you go to another page on the 3D Audio site. You're on the honor system, so guess before you look at the answers.

Before listening to the tests, I had my own ideas about DAW summing. While certain that DAW summing resulted in sonic changes, I had my doubts that you could really hear significant differences between platforms — enough that you could tell one platform from another in blind listening (given identical conditions). It was also my belief that the difference in DAW "sounds" wasn't near as important as, say, converter, mic, or preamp selection.

Sitting down to compare the files isn't easy. To reduce ear fatigue, I listened in many 15-20 minute sessions over a week. At first, I was hard pressed to hear differences. But with repeated listening and a lot of A/Bing, I started to pick out things. Eventually I was able to fairly reliably pick out certain files when listening blind — the analog mixes were easiest to pick out, although they weren't necessarily the best sounding. Interestingly, when A/Bing between two digital files, I was sometimes able to pick one from the other. For example, after much listening, I was able to blindly tell the Sony DMX-100 mix from the Panasonic DA7. With the DAWs, it was harder to "hear" one reliably in blind A/Bing — in fact, a number are numerically identical.

One thing this exercise teaches you: how to listen and evaluate sound. I learned to focus intently on frequency bands in the mix, working my way from bottom to top; next I would focus on stereo image, then clarity, and so on.

So which platform "won" for me? Drum roll . . . none. The differences simply weren't that pronounced — which says something. However, keep in mind that we're talking about only the most basic aspect of mixing in a DAW. If you factor in multiple fader and pan moves, plug-ins, and other processing, the differences would likely be more audible.

I'll sum (sorry) it up like this: In the end what matters is the results your studio produces, not what you use to get those results. As long as you're happy with the sound quality, then all is well. But educating yourself about your tools is always beneficial. With regard to DAW summing, Awesome DAW Sum is a good step in forwarding your knowledge.