So Much "Harshness"

While many good points have been made above, my feeling is that you hit upon the most significant reason in your original post. Recordings that are poorly engineered, or intentionally brightened, or both.

Most of my listening (80% vinyl, 20% cd) is to classical music on labels that are either audiophile-oriented, or are otherwise high quality. I hear very little that I would describe as harsh. When I have occasion to put on a popular recording, or many big-label classical releases, my reaction is often (although certainly not always) simply "ouch."

Regards,
-- Al

Tim,
This is the url for the 4 Gauge Cable;
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=100-196

Commcat,
Well I'll be damned. But why scrimp? Might as well go for the 1 gauge.
I'm going to stop in at one of the local car audio places this week to see how thick that stuff is.

In my opinion, in the majority of cases these days, the cause of harshness is the loudspeaker drivers and crossovers. Especially, when used with tube amplifiers.

Drivers exhibit rising impedance as frequency increases as a function of the voice coil inductance. This impedance rise will obviously be at its maximum right below the frequency where a driver is crossed over. At any rate, tube (and, solid state, according to Ralph Karsten) amplifiers prefer higher impedances. What this means is that they put more power into those higher impedances, resulting in increased (and, out of balance with the rest of the musical spectrum) volumes in those regions.

Given that the preponderance of loudspeakers these days are using a crossover between the midwoofer/midrange driver and tweeter in the presence range (upper midrange/lower treble), usually falling somewhere between 1500 Hz and 3000 Hz, which is precisely where the lion's share of complaints of brightness and harshness are centered, it all seems pretty elementary to me.

The classic way around this is a Zobel (resitor - capacitor in parallel with the respective driver) network in the loudspeaker crossover, which flattens the impedance rise due to voice coil inductance. However, in my opinion, the Zobel causes at least as many problems than it solves ala robbing music of immediacy and drama, in addition to making the loudspeaker that much more difficult for the partnering tube amplifier to drive. At any rate, since most current loudspeakers do not implement the Zobel (or, at least, in this manner), here we are with this issue being exhibited in the majority of high-end audio systems of the current times.

I do think today's higher resolution equipment plays a large role in the epidemic of harshness. I don't recall hearing this kind of harshness in systems years ago. Higher resolution is quite often obtained by boosting upper mid and high frequencies, this acerbates harshness in recordings.

I've also found rather mundane parts quality in a lot of equipment. Upgrading capacitors and resistors can result in even higher resolution and much greater refinement.

I find too many recordings are excessively bright, again a problem with boosting upper mid and high frequencies.

All of this resolution does require careful system matching, getting the highest resolution and refinement is a delicate balancing act, easily the most difficult thing I've encountered in audio.