True. Most are designed to behave a certain way based on 24-bit amplitude, or rather, some general representation of analog 0. But the part that confuses people unnecessarily is a universal truth about floating point, e.g., you can trim it anywhere to any level you want. Besides, any expensuve analog emulating plug worth it’s salt should probably have an input gain knob, if not, trim the gain on the plug before it. None of it really rises to the level of being critical because it’s all infinitely adjustable all over the place… at any stage.
The PT guys made an argument that you’d be sending to hot to hardware verbs or simlilar, which is quite true if you’re up near digital 0, true but funny and out of context, why? Because sends themselves are simple gain controls, send whatever level your hardware needs, starting from dead silence. They oddly seemed to be arguing from a position that these signals are somehow stuck at unity and can’t be changed. They can, everywhere amd anywhere.
Seriously, the plug I’d use to connect hardware to my daw has trim / gain on both ends of it (send and return) so I could be running +6 on the track and still trim it down there, without worrying about internal gain staging at all. It’s all just pure digital gain/trim, nothing that will ever stop a show or damage the audio at all.
This really has been a very educational thread. I think it clears things up in a real-world, practical way that all of the other forums and discussions I have seen really haven’t. I now really understand the value of 32 bit float as well. Thank you guus so much for your great contributions!
Regarding plugins, obviously they all have their sweet spot. Most have overload indicators as well. I think that was kind of a given, I was really only interested in the internal mixing structure of Cubase itself, and my mind is at ease. Just out of curiosity, does anyone know if this all applies to Logic as well? I have had to mix some stuff on a Logic system at another studio recently, and I assume it has the same internal float processing as Cubase? (Completely off topic, Logic’s track / channel naming scheme is the most horrifying thing I have ever dealt with. So glad I use Cubase as my main rig. I literally almost pulled my hair out.)
One of these days I’m gonna mix a 24 track project with every single one of my mixer channel meters permanently pinned, just to maybe make this point for once and all, no meters bouncing anywhere except when things actually stop playing, and post it on YouTube, just for the heck of it.
I’d start it by adding max (24db) of clip gain to all the clips. Now that would be really, really “lopsided” gain staging. That would be one really, really odd looking project, but it would sound fine.
I’d only really have to trim levels just before comps and limiters since their thresholds won’t have settings that high. Then add +10 / +15 of output gain and keep going.
The master fader would likely be at -50 or so… and it would sound perfectly ok.
Start from the end and work back, ie have masters, groups/fx out’s and VST instrument outputs all at 0 and make adjustments in the VST instruments accordingly, in addition to note volumes and controllers.
This will preserve headroom, as well as dynamic range.
Basically there should be no problem with what you ask.
FWIW I like having the metering and faders working around the optimal point so as to have a good range on the faders.
Also gives a good overview in the mixer of whats what!
How do you know when you’re clipping a software plugin (Say a UAD 1176 compressor) that has no output meters? The reason I ask is that one of the selling points of analog based plugins is that they have a “sweet spot”. Or, is that nonsense?
Secondly, why does a Cubase user need a trim/gain plugin before sending to hardware effects? Doesn’t the Gain Knob on each Cubase channel suffice?
