[geoff]

On 11/02/2019 9:06 AM, Scott Dorsey wrote:
James Price wrote:
I was under the impression that the primary difference is that overloading
the front end of the ADC would potentially produce digital hard clipping at
the output, that is assuming the amp sim's volume isn't set particularly low
and the input level is cranked high enough. For example, a 50 dB boost applied
to the input of an ADC vs. an analog amp. Both will clip, but my presumption
was that the ADC would produce digital hard clipping in the form of a square
waveform, whereas hard clipping in the analog amp would produce a waveform
that wasn't chopped.

Clipping is clipping. Some analogue circuits will "clip softly" which is
to say they have a little gain reduction right before they come to the
clipping point. Some analogue circuits will overshoot when they clip.
Some will get stuck for a cycle or two when they clip.

All of these will result in a different spectrum. You'll still get the
odd harmonics that you get from perfect clipping, but the ratio between
them will be different and you'll get other stuff too.

People that build guitar amps think a lot about behaviour of circuits
when they clip, but most designers don't care about it because they don't
expect people will ever clip anything. Hi-fi tube amps often use splitter
circuits like the paraphase inverter or the see-saw, which you will never
see in guitar amps because they behave very badly when clipped.

When you go into clipping, you have left the normal operating envelope of
the equipment, and you have become a test pilot. Don't make any assumptions
about how anything will behave when it's clipped. If you intend on using
clipping as an effect, try it and listen.
--scott



And then there is the solid-state v. vacuum tube predominant even/odd
distortion harmonic thang, as you obliquely alluded to in your second
paragraph.

geoff

geoff