I was able to turn it down and turn down the signal coming in somewhat. No difference in terms of distortion.
This behavior would be the same in both scenarios I describe above. The only visual indication of the input stage of the recorder being overloaded while recording might be a subtle reduced "activity" of meter movement. You can still turn it down more, lowering the indicated peak levels on the meters, but it will not prevent the distortion which has already occured.
A difficult to convey aspect of this is that a setting lower than the "do no go below" input setting
can work without distortion if the input signal is low enough to not overload the input stage. Yet such a setting would produce a recording with very low levels so there is no good reason to use such a setting. The problem is when the input signal is high enough that you have to turn it down, and it overloads things prior to the "turn it down point".
Ed describes the correct way to run the recorder for a hot line-in signal, by reducing the level before it reaches the recorder, in his case using in-line attenuators.
The source must’ve been coming from somewhere else.
^
Not necessarily for the reason I describe above (this is tricky as described)
We generally go through this same learning curve at TS with each small recorder until the "don't go below" setting has been determined by practice. Upon which it becomes general knowledge for that recorder.
Source of distortion could very well could something prior to the recorder, but the sceario I describe has not been ruled out definitely.