There’s a big difference between noise floor and resolution. The noise floor listening to Voyager is pretty high compared to the signal, but we have enough resolution to receive the signal in the presence of many orders of magnitude more noise.
Here’s a low level (-120dB) 1k Sine (I don’t remember which release):
Most of the lower frequency crap is my computer not the DAC and the 8k is from the USB cable not the DAC. There’s a little 2k harmonic distortion, but not bad for -120dB. This plot also hints that the linearity is still pretty good at -120dB (despite some other claims.)
Here’s a -40dB 1k Sine
The noise floor is higher on this one because I was using my lower resolution (but wider bandwidth) scope. Once again an older plot, the ultrasonic noise now starts at a higher frequency and doesn’t go as high in amplitude.
Here’s an older -60dB 1k Sine
The blips at roughly 16k, 27k, etc. were also there with the scope probe directly to ground in that setup.
I though I had some of louder 1k signals, I’ll look around some more.