Quote:
Originally posted by akdrama:
Great discussion.


If we look at the two graphs Altec provided...
the tube signal after 10khz, it flattens out. As compared to the SS after 10khz, there are a lot more peaks (distortion peaks?). In regards to the tube signal, would this account for the "liquid highs" characteristics of tubes? Tubes are often characterized as having sparkle? How would that show when analyzing the graph?

And with the distortion of the solid state after 10khz, would it be safe to say this is the "brightness" aspects of SS?

AK
Distortion components ideally should not extend beyond the 5th harmonic, and they should decrease linearly in level as the harmonic number increases.

The 2nd harmonic component is essentially harmless because it is fundamentally musically related, being an octave above the original frequency, and in fact has the beneficial effect of masking the presence of the other, higher order distortion components. This effect is one reason tube buffers smooth out the sound of an otherwise solid state system.

The 3rd harmonic is a musical 5th, and is still not overly objectionable, but it should never be higher in level than the 2nd harmonic.

The 4th harmonic is just a double-frequency replica of the 2nd, and is likewise musically related and relatively harmless. It should be at a lower level than the 3rd.

The 5th harmonic is the first harmonic which can start to sound harsh if it is not lower in level than the 2nd, 3rd and 4th.

The 6th and above start to have really negative impact on sound quality, no matter how low their absolute measured level is. The structure of these higher harmonics is one of the main reasons different pieces of audio gear sound different from one another, albiet in subtle ways.

The distortion components I showed of the solid state amplifier photo go all the way up to the 20th harmonic and above! This will color the sound, causing sound quality differences between it and other amplifiers, especially tube ones, which have a virtual absence of distortion components beyond the 5th.

In subjective terms, tube amplifiers, because of their benign distortion signature, "get out of the way" of the reproduction of music. That goes a long way toward explaining the "liquidity" and "sparkle" in the high end.

The presence in solid state amplifiers of high-order and non-musically related distortion components will contribute to their sounding "hard" or "sterile".

From a sonic importance standpoint, the earliest components in the full audio chain have the most influence. In my recording work, this means a tube microphone preamplifier. In home audio, the same thing applies, with the preamp having more influence than the power amp.

However, all components in the chain will together influence the ultimate sound quality, so that is why as an ideal I always recommend a totally vacuum tube setup for music reproduction if at all possible. Short of that, a tube preamp (or tube buffer) at the low level audio stages can be greatly beneficial.