Why? We’re talking about headroom, not levels… I believe there’s a misunderstanding due to imprecise wording in the documentation. When the OT2 documentation says:
Balanced audio inputs: Input level: +17 dBu peak Audio input impedance: 21 kΩ
that doesn’t mean that we need to input a +17dBu signal (5.48 volts RMS) to get a full-scale digital recording. What is omitted is the nominal input level, which we can logically suppose to be +4dBu like any studio grade balanced input should be.
In real life, this means that when we have our source VU meter (say it’s a mixer) indicate 0 VU, the source voltage will be +4dBu ( 1.23 volts RMS) and we can safely assume that this will result in a 0dBFS recording without clipping, which is what I experience.
The input circuit in the OT MkII is designed to handle a higher voltage, up to 13dB overshoot, but the gain must be turned down to prevent overloading the converters. When we turn the gain down, the leds follow. So, as I see it, the input amplifier is a DCA and sits before the converters (to be sure we should take a look at the actual circuit board). Otherwise a "gain " parameter on the inputs wouldn’t make much sense.
But that’s all guessing and not very helpful IMHO, really. The important part is matching source and input levels, the internal gain structure is pretty decently engineered with enough headroom for resonant filters, EQ’s and other goodies. All I need to know when sampling on my MkII is that when my digital mixer’s output level meter says -18dBFS I’m hitting the +4dBu ceiling at the analog outputs, and that means I’ll get 0dBFS samples with no clipping, given I have the OT settings to default. Your reference may vary, but it’s the same principle. And it’s exactly the same for the MkI, except that it has less headroom.
EDIT:
Sorry, but his doesn’t make sense at all. There’s no digital input on the OT, we’re in the analog domain. Much, much confusion…