In hifi, anything not the signal which somehow arrives at the ear is distortion. In this context, the signal is what's measured on an oscilloscope. What kind of dead-obvious distortion can we list? Let's start with noises. There's the pop and crackle of a dirtier LP. There's transformer and ground-loop hum plus power-supply hiss close to the drivers. There's scratching noise from dirt in a voice coil. There's also traffic noise and refrigerator hum and all the other contributors to our ambient noise floor as that which we call silence when no music plays. Measure it and it'll show anywhere from 25dB to above 40dB. That too is distortion; actual anti signal which creates a S/N barrier. Incidentally it can be reduced or outright disappeared with sealed 'phones, particularly well-fitting in-ears and their ultimate expression, the active noise-cancelling kind. For more obvious distortion we have tube rush, crackle or microphonic motorboating; or structural resonances like rattling aircon ducts. What about boom notes where specific bass tones trigger ringing, making those tones noticeably louder and more resonant? On the recording they all measure the same amplitude but played back in our room, they're uneven. By definition, that's distortion too.
How about playing things at higher SPL when the treble keeps scaling up fine but the bass lags behind to make the tonal balance brighter? How about noticing that after two hours of play, dynamics feel really congested because voice coils have heated up enough to raise their resistance? Dynamic peaks that should crest to twice or four times as loud only manage a 10% increase. Dynamic compression is another form of distortion but not nearly as obvious as a screech or pop, overdrive/clipping or pitch shifts from a cassette's or LP's rotational instability.
How about insufficient bandwidth and digital? What should look like square waves with clearly defined transitions between high/low voltages instead shows rounded edges, overshoot and wiggles. In the time domain this blurs the exact point at which each voltage state switches. It's a form of jitter. During playback it's not at all obvious unlike ticks and stutters from a bad Internet connection and dropped USB packets. If we can eliminate or reduce digital jitter, we might notice that before, things sounded a bit softer or slightly more diffuse. What about a speaker's impulse response which shows that sounds don't stop properly; or that various drivers don't rise in unison but staggered in time? Then there's harmonic distortion as every tube hater's favorite red flag because most valve gear measures higher THD than transistors. Curiously, adjustable negative feedback on valve amps to reduce such distortion often has experienced listeners prefer less or no feedback. By definition that means higher distortion. Once again this type of measured deviation isn't 'external' like obvious non-musical noises. It's 'enfolded' inside the sound. It can make it softer, warmer and denser. If that has your vote, being told that you're a distortion addict could sit wrong.
It also overlooks that our ear/brain runs its own distortion analyzer. Higher measured distortion of the right kind may actually cancel out to register lower distortion. Brands like Audiopax, FirstWatt, Hiraga, Riviera and Tenor deliberately tune their THD behavior to this assumption. Now higher measured distortion fed to our ear/brain's processor results in lower perceived distortion. Voilà, more natural sound. While THD and amplitude measurements often dominate distortion discussions, what about port ringing or our room's coarse very unpredictable EQ on the frequency and time response? What about our speakers' built-in phase shift?
This type distortion rarely makes Hifi Police news. Yet its severity crushes any 0.005% or 0.05% THD concerns. Speaker dispersion interacting with our room has a say in how reverb times differ across the audible bandwidth. Some tones linger longer, others clip off sooner. That textural nonlinearity shifting back and forth between dry/damped and loose/ringy is yet another deviation from the recorded signal. By definition, that's still more distortion. Because we live with our room on a daily basis, we've long since tuned out its acoustics. That makes the milder forms of this distortion very hard to notice yet could be crystal to a visitor unfamiliar with our room.
How about vertical baffles versus adjustable time alignment between speaker drivers as championed by Wilson Audio? How about high digital latency in DSP subwoofers which are additionally placed in room corners to sit well behind the loudspeakers? How about energy storage and delayed release in coupling capacitors and dielectrics, phase shift in output transformers, signal reflections from mismatched impedances? Timing errors are another form of distortion that's rarely talked about. Regardless, it should be clear that just as life is messy, living without any playback distortion is a Quixotic dream. It's also a very distorted view on the matter as though reality was at fault. Playing back various sound sources of varying size and height over two speakers locks us into a very limited reality from the start. A lot of the distortion mentioned above is sonically embedded. That contributes to the sound's overall feel or personality. As long as that coincides with what we enjoy, it's all good. Zero distortion is actually impossible. The real goal must be to manage it so we end up with distortion that's no bother; or better yet, distortion which enhances our experience. At least to me that seems a more mature and correct view on the subject than declaring all distortion the enemy to promote a campaign of total annihilation in pursuit of any absolute sound. Such war is best left to those who 'listen' exclusively over oscilloscopes. They hide in an ivory tower of idealism. They never get down to deal with reality. Reality requires a rather more practical hands-on approach; far less theory, lots more practice. It also needs sufficient stones to declare "I like this" then to not give a bleeding toss what anyone else thinks. Listeners differ. What they listen to and for, how loud and what triggers them positively or negatively differs. There's no consensus reality, just individual encounters with an illusion. Artificing this illusion so we find it persuasive—a common expression is suspension of disbelief— is our job. It takes learning what's important to us; learning what's available in different solutions and results; then getting co-creative to curate it all such that for us and our budget it works the best. It's very basic and quite complicated all at once.
And then our tastes or circumstances change and we start all over again.
And we must count ourselves lucky for the opportunity and wherewithal to engage in this form of personal entertainment. Obsessing over better sound or a more musical hifi is a real luxury. Personally I'm allergic to noise so want my hifi quiet like a church mouse; to stereo imbalance to want my center image centered; and to boomy bass. All of those are external forms of distortion easily recognized. I've also recognized that I prefer direct-coupled amplifiers of wide bandwidth. Over the years I must have developed another allergy to the effects of most coupling capacitors and output transformers. But all the rest of the embedded distortions I can't identify per se. A/B comparisons net differences. Sometimes they lay bare preferences. Now I conjecture about what created my preference. Could it be lower jitter? Less IMD? A more benign form of THD? Less or more negative feedback? A simpler crossover? Lower noise on the AC? The list of potential suspects is long. Determining clear cause is impossible shy of doing circuit or filter surgery to isolate a specific part. If we listen and compare enough over the years, we'll simply recognize general trends. Perhaps we prefer passive power distributors to most active filters. Perhaps we prefer active preamps to none or autoformer volume controls. The possibilities are endless; as are the chances for exceptions. To assemble a pleasing stereo, we in fact needn't know what causes our choices. We simply recognize what we like better and act accordingly. It's when others ask advice that it's all too easy to go beyond simply sharing what we've learnt about our own sonic tastes and how to best accommodate them. We might propose certain rules to the correct doing of hifi which correlate with our preferences and associated beliefs. Alas, it's far more likely that our 'rules' are in fact based on inaccurate assumptions and technical misunderstandings; or at best are just personal so far from universally applicable. Which would be yet one more form of distortion to add to this steaming pile. That wouldn't be helpful at all now, would it?
Of course life is messy. On that score, all is as it should be…