Every theory is grey so I tested practically.
Setup:
ADI-2 Pro USB mode, source Tidal on iPhone.
ADI-2 Pro Outs set to Auto Ref Level, Hi-Power.
ADI-2 Pro Analog In (Monitoring) set to +19dB
All processing in ADI-2 Pro set to bypass.
AD and DA Filters to Sharp.
Recording:
ADI-2 Line out 1/2 to:
Nakamichi 680ZX, Rec Level 7/10, Master max.
Cassette Sony MA-X 60min, best Metal Tape available.
(BTW: CrO2 tape you would need to substract 6-8dB from all levels below!)
Azimuth, Bias and Level newly and precise calibrated, frequency response checked.
Monitoring:
680ZX Line Out to ADI-2 Pro Analog In, routed to headphone out 3/4.
Listening on STAX SR009 / SRM1-Mk2-pro (known for not distorting by itself).
Comparison done by switching Tape/Source on the 680ZX during record.
Results were as expected, typically for cassette recording:
Nothing unexpected happened.
The tape recorder behaved with RME ADI-2 Pro like it does with any other source.
Conclusion:
Source signal always was clean, even at levels way above those used.
From tape, with Dolby switched off, pulses and attacks were almost unchanged up to a recording level of -5 to 0dB (on 680ZX peak meter).
Above progressively the percussive attacks were softened, bass becomes audibly distorted above ca. +7dB at max, depending on the kind of program used.
With Dolby B attacks and brilliance were slightly reduced, so Dolby B was not used in this test.
The transition from clean to distorted is very soft, so it largely depends on personal preference at which point one would could consider the signal as distorted.
As a studio professional having started my career working with analog, but for a really long time with digital now, I am not very tolerant to distortions.
If the signal is changed, attacks washed out and punch in the bass range is compromised I don't like that.
As usual, your mileage may vary, especially with analog tape the transition between clean and squashed is quite soft, starting ca. at -10dB ranging up to +10dB peak input level.
At that point the output has lost about 3-5dB peaks by saturation, even more in the upper treble range.
With Sinewaves the tape behaved normal too.
Max rec. Levels rel. to Dolby Level 200nWb/m:
0dB @ 33Hz / 3% K3
+10dB @ 333Hz / 3% K3
0dB @ 10kHz Saturation Level
-5dB @ 12.5kHz Saturation Level
-15dB @ 16kHz Saturation Level
Here it's obvious what I explained in my former posting above:
As usual with small track, slow speed analog tape, highest level can be achieved in the mid range.
Maximum clean Bass level is 10 dB lower.
Treble is saturated much earlier too, and saturation means it's already completely distorted.
This figures fit to the way music was produced at the time Cassette recording was invented, strong midrange and recessed bass and treble.
This kind of fits to vinyl discs too, which have comparable limitations.
Some details and a selection of musical examples:
Following music taken from current Tidal steaming version, that is partly different to the original CD release.
Levels:
The Nakamichi 680ZX has a frequency independent Peak Meter with an attack time of 10ms, shorter peaks are not fully shown.
Real peaks can be estimated as ca. 6dB higher than visible on 680ZX's Peak Meter.
0dB is calibrated to 200nWb/n, = Dolby Level, not very hot!
Dire Straits - Six Blade Knife:
Clean up to ADI-2 Pro Out -33dB, 0dB on 680ZX Peak Meter
Higher levels give progressively more distortion in the bass, and the percussions become softened.
Donald Fagan - Morph the Cat:
Clean up to ADI-2 Pro Out -36dB, -5dB on 680ZX Peak Meter.
Above the very peaky percussion attacks are progressively squashed, sound becomes dull.
Michael Jackson - Bad (2012 Remix):
Clean up to ADI-2 Pro Out -37dB, -2dB on 680ZX Peak Meter
With this one you can go much higher without obvious distortions, it just sounds softer overall.