Gain question

[DThompson55]

We've been very happy over the last 2+ years with the X32. We're using it for church sound, and have just a couple questions regarding gain.

We always set gain such that we set it to peak at -18db (the first yellow LED) and average -30db, and this seems to be fine and still give us enough headroom that we rarely have to push faders above 0. It seems low to me, but I figure as long as it works, it's probably good. I wanted to get some validation on those settings.

The 2nd thing is that we send all channels out USB prefader/premute, to yet another computer. That computer runs something called vMix which lets us combine audio from the X32 with video from a pair of PZT cameras and live stream out over the internet. I don't expect you to know anything about vMix, but what I've found there is that I have to add 12db gain in vMix to every channel coming in from the X32 in order to get any headroom, otherwise the faders on vMix are maxed out and I cannot get decent volume on the live stream, where I shoot for -18db LUFS.

So the question is, is the X32 reducing the signal out by 12db when sending over USB? Or is it sending my actual signal strength and vMix probably just needs to add an additional 12db?

[craigf]

If you're peaking at -18dB you're just wasting headroom, in theory at the expense of signal-to-noise and clarity. As long as you've got headroom for transients without clipping you could set your gains a bit higher. I'd be aiming for a "performance level" signal to be averaging at about -18dBFS, and as long as peaks are under 0 you're fine. If you 18dB of headroom should be plenty for live church sound.

If you do actually need 30dB of dynamic headroom that sounds like a very dynamic mix. That amount of dynamic headroom is heading into orchestral music territory, and usually venues are built with acoustics in mind to accommodate. Most churches aren't. I'm kind of jealous. Of course if you do adjust your gains you'll need to adjust your dynamic thresholds and other gain stages later in the signal chain to compensate.

If you're tapping signals straight off the preamps then chances are the X32 isn't doing anything to the signal before it sends it out over USB (the exception being if your gain is set to +45dB, then there's preamp digital trim shenanigans.) At most it'll be a 2dB difference due to the preamp stepping, but not 12dB (unless there is channel digital trim involved and you're tapping post that.) Chances are that your software is just built to work with stronger signals.

[DThompson55]

"If you're peaking at -18dB you're just wasting headroom"
- I concur. And other than our piano player who has a big dynamic range nothing should very dynamic so we shouldn't need all that headroom on most of it. We'll play with that.

"Chances are that your software is just built to work with stronger signals."
- That's possible. I'll try a running a DAW on that machine and see how that behaves. Whether the levels are low on that or not, gives me it's another data point. Behringer's X32 ASIO ought to be transparent, but I'll check that a well.

Good catch on the +45db thing.

[GaryH]

@DThompson55 Card routing (which is really card out/usb out routing) defaults to 'local' 1-32 as its sources. Local tap by default is just past the preamp and can't be changed. Using out 1-16, aux out or p16 out categories instead for the Card routing choice, gives you other tap point options. It is unclear to me if you were doing that already for vMix via USB? Anyway, any choice other than post fader can be a bit to quite a bit lower in level to a DAW (and possibly vMix depending on how that was implemented) Many prefer using a Matrix or 2 for the live streaming source and adjust those post fader levels to taste. maybe you are doing all that, just thought I'd mention it.

[DThompson55]

@GaryH - Correct. I misstated earlier. We actually have inputs 1-16 -> p16 -> card 16/16. It's the p16 that is routed prefader.

We originally used a matrix to send to vMix but found it was unwieldy trying to adjust both the room mix and the stream mix on the X32 at the same time. So now we send all the inputs to vMix prefader+premute, and a 2nd sound tech makes streaming audio decisions there.

This said, I think my questions have been answered. I'm convinced the X32 is sending levels as indicated on the gain LEDs. So I'll just boost them on the vMix side.

I understand we are wasting headroom on the X32 with our gains adjusted to peak at -18db, so we'll play with that to see what improvements we can make.