Gain Staging the X32
Using the Behringer X32 as a line-level digital interface
You'll see the comments all the time from friendly and unfriendly users of the Behringer X32 digital mixer. Some will say, the preamps are very transparent. Others say that they would NEVER use it for studio recording. But nobody would tell me what I wanted to know: What if it is used just as a line level device? Well folks, you've come to the right place for the answers.
For those of you who have experience with my long-winded articles, I'll cut to the chase. The essentials are as follows:
The X32 preamps at 0dB gain are very clean.
Keeping the X32 A/D converters a -18dBFS or less is good policy.
The Echo AudioFire 12, Pre8, and Mackie BlackJack at 0dB can work as clean line level interfaces.
Click on the sidebar images to get a larger pop-up window
So here we go. First up, [Figure-1]. In that picture, we see the topology of my testing setup. I first created a perfect sine wave in OS X Yosemite using Adobe Audition. The sine wave is 992 Hz to avoid having exact multiples of my 48khz sampling rate being reflected back when we ultimately do compressor and limiter testing for inharmonic foldowns. For our measurement purposes, 992 Hz is close enough to 1khz to allow 2H, 3H, 4H (2nd harmonic, 3rd harmonic, 4th harmonic, etc.) values to be casually interpreted as 2khz, 3khz, etc., without concern for scientific accuracy.
The sine wave was driven from the DAW session through the Analog-1 output of the X32 (labeled as Aux-1). One immediate discovery was the tremendously clean output of the D/A converters of the X32. Every device under test showed a noise floor down near the theoretical minimums related to Johnson noise at room temperature, and the external devices under test sometimes had mere wisps of harmonic products breaking through -120dBFS. We'll get to that soon.
If you let your Internet imagination run wild, you'll imagine that the sweet tones of your latest retro analog tube and transformer wonder will be crushed and destroyed in a hail of distortion as shown in [Figure-2]. What actually happened in that figure is that the -1.7dBFS peak sine wave does odd things in the context of an A/D converter. There's something called system gain or performance gain because converters are actually 6.02N+1.76dB in their behavior. You can read all about it on page-29 of this document here [http://www.rohde-schwarz.de/file_5613/1MA98_4E.pdf ].
When I reduced the Analog-1 output level from -1.7dB to -1.8dB, everything went much better, and the expected profile of a well-behaved solid state preamp emerged as shown in [Figure-3]. The real beauty of the X32 as a line level device emerges when we get to [Figure-5 and Figure-6] the images that show how the distortion products H2, H3, H4, and H5 are dramatically lower than before, and in very usable levels when we have reached the target values of -12dBFS and -18dBFS.
At -18dBFS, the 3rd harmonic distortion level is -108dBFS. That's 90dB below the fundamental level. Expressed in Hi-Fi terms, it's 0.0031% distortion. The non-musical 5H is about at the same level. To put this in sound pressure level terms, the way I use my studio monitors puts a sine wave from a mono channel in my ears at 80dB SPL when a sine wave in the DAW has a -18dBFS level. (Pink noise at -18dBFS checks in 6dB lower at 74dB SPL.) The ~3Khz 3rd harmonic at -108dBFS would never be heard, as it's 25dB below the noise floor of my studio, and it's actually in negative SPL territory with that calibration; clocking in at a physically impossible -10dB SPL. I'll explain more fully in another article.
In upcoming installments, I'll show you how the other line level interfaces measure up, and how the various preamps look in the same FFT display.
It's gonna be a long series...
- Ted Gary of TedLand
May 29, 2016