Zen and the Art of Mixing is a great read. Mixerman (AKA Eric Sarafin) has a knack for writing about a subject that normally defies the written word. However, there was one part of the book that really got my goat when I read it, and here I am, six months later, my goat still got. It can all be boiled down to this quote:
I can’t tell you why. I can’t tell you how. I can’t even prove what I’m about to tell you, and I can assure you that the DAW manufacturers, particularly Digidesign, will not only reject this claim but will actively try to persuade you otherwise through flawed white papers that most of you can’t understand and bogus comparisons that most of you wouldn’t know are bogus.
All DAWs bog down at the 2-bus.
For a book that proclaims, through its back cover blurb from Ken Scott, to teach “the Art [sic] of great mixing, not the pseudoscience,” that’s an awfully pseudoscientific claim.
Am I here to stake a claim in the never-ending summing debate? Heck no. Rather, I’m here to ask a simple question, directly related to Mixerman’s obstinate assertion: If what he says is true, why is it that analog summed mixes (or perhaps we should say mixed-with-an-analog-summing-box mixes) sound better than those done ITB? We, as a species, can put a man on the frigging moon. So why can none of the seven billion people on earth tell me why, using empirical evidence, analog summing sounds better to the human brain than digital? Is it the distortion? The re-amping? The AD/DA passes? Inherent limitations of digital audio? If so, what exactly are those limitations? Phrases and words like “bog down” and “choked” are utterly meaningless. I want science, people.
Audio engineers (and by “engineers,” I mean real engineers—not fake ones, like me), software developers, NASA scientists of the world: I call upon you to settle this debate once and for all. And not in an internet forum—in a honest-to-goodness lab. We can even break out the white coats, 1960’s Abbey Road style. Who’s with me?
9 Responses to “Zen and the Art of Analog Summing”
You know that feeling you get when you pick up a vintage guitar? It doesn’t have to be seriously “broken in,” in fact it could very well be that it had hardly ever been played. Maybe it sat in a closet for the last 30 or 40 years, who could even know. The point is, it just feels different. (Pretend “different” is italicized since that is not an option here.) It feels different in a really good way. Suddenly it seems more effortless to pull a perfect passage from the neck and strings. Who could say exactly why – is it the quality of manufacturing – a past age of more abundant tone woods? Or maybe the number of windings on the pickups and the age of the magnets? No one can say exactly why. It just is. It’s fundamentally made of different stuff. This is like that. I put forth the humble hypothesis that, just because it cannot, or has not been quantified, doesn’t mean it isn’t real…
Ultimately, if you can get moved harder by {THIS} processing over {THAT}, well, how can you make any self respecting argument to the contrary? You choose {THIS} and then you feel so good. So good 100%? Eerrrr maybe (dare i think it) so good 110%? Quality over quantity? Quality in this example gives us quantity (of goodness feeling). weird…
Anyway, if box summing makes you feel alright then just do it. BUT if it’s just the easy way – well you better explore those options. It seems to me at least, that the world might benefit from more people DOING SOMETHING with a lot of effort, rather than BEING SOMEONE with a lot of effort. (this is not directed at you sir, just a general observation) To site your example, we didn’t get to the moon by taking the easy way now did we?
USA! USA! USA!
Thanks for your comment. I completely agree that the feel of an instrument, or tool, or in this case, summing bus, has tons of value in and of itself. A lot of the gear choices I make as an instrumentalist and recordist (and romantic at heart) are based on feel/vibe/aesthetics/je-ne-sais-quoi-ness.
Still, my inner scientist wants to know at least a modicum of cold hard facts. Given the amount of time and energy and money being spent on this debate and its solutions, it’s crazy that nobody has yet taken the time to get to the bottom of it by sitting down and figuring out exactly (or at least roughly) what’s going on under the hoods of these various summing schemes.
Reminds me of when I read How to Brew, and was surprised to learn that for all that we humans know about beer, there’s still a lot about the chemistry of the beverage that we just don’t know. And we’re talking beer! Not, you know, the 66 moons of Jupiter. And while sometimes it’s best to just open up a window and indulge in the mysterious and tangy vibe of wild yeast, sometimes it pays to be a little more deliberate about things.
You hit the nail on the head, especially pointing out the faults in Mixerman’s unfounded opinions. It’s trivial to prove that modern 32-bit digital summing is highly accurate. Yet even when shown that proof, guys like MM cling to their belief system anyway.
Years ago, most recording engineers actually understood the science of audio. Back in the 1960s and 1970s gear manufacturers aimed for a flat frequency response and low distortion, and they routinely included performance data in their ads and brochures. But now that high quality audio gear is affordable, some parts of the audio industry has become dumbed down. Today, unfounded opinions substitute for scientific evidence. People who may well be excellent *mixing* engineers wrongly assume they also understand the science, even when they don’t.
“Heck no. Rather, I’m here to ask a simple question, directly related to Mixerman’s obstinate assertion: If what he says is true, why is it that analog summed mixes (or perhaps we should say mixed-with-an-analog-summing-box mixes) sound better than those done ITB?”
This might be easier to answer if you didn’t have DAW developer’s heads in the sand.
I also lay out exactly how you should test the summing box, because the test must be process driven, not results driven. Yes, your results will improve, but there is no possible way to actually perform a scientific results driven test for this.
Mixerman
(Yes it’s me. I’ll post a link to this from my FB page to verify.)
Well one engineer once told me this: Becouse 1+1 is 2 and not 10 :))))
btw. digidesign made their mixbus analog-like with heat plugin couple years ago and gues what it sounds better. In this video talks Dave Hill its creator about its principles. There is lot info which might be considered as an answer to your question http://www.youtube.com/watch?v=SyqPZcFRkRY
My take on the ordeal http://rhythminmind.net/1313/?p=2402
I agree with MM. I have done my own tests. I still use a console for the front end and mixing. I would record the master buss on the console for my “mixdown”. For my test I would create a mix in the box and “bounce” it to disk. Then I would take the mix and spread it out over 8 channels of the console and mix it down to the DAW in a stereo track. Ten bring em both up in a fresh session and A/B them..The truth is right there..the mix summed thru the console sounds smoother, rounder, bigger and wider. These summing boxs do the same thing except without all the knobs and faders. I’m sold.
Simply put; Reality contains imperfection. There is, and can never be, a perfect representation of imperfections.
When you pass audio through a console or analog summing bus, what you are actually doing is real. There are electrical impulses pushing and pulling on the transformers, capacitors, wires, resistors etc… The variables are too unpredictable.
We live in a physical, mechanical and analog world. Digital will always be educated guesses, but as it strives for perfection, it becomes less real.
“Today, unfounded opinions substitute for scientific evidence. People who may well be excellent *mixing* engineers wrongly assume they also understand the science, even when they don’t.”
Of course, when given the opportunity to run a “scientific test” on GS, you altered the files so as to get the results you wanted. Not once, but twice.
Yeah. That’s scientific.
Mixerman