Warming Trends Tube Warmth: Cure? Or Crock?

When I did blind listening tests for my comparison CDs in 2000, one of the objectives was to pick out which preamps were tube and which were solid-state. Easy, right? We were shocked to find that even with a room jammed to the rafters with experienced engineers, none of us could reliably pick out the tube preamps — a task we all thought would be relatively simple because of the “warm sound” of tubes. It was a real ear-opener. Since then I’ve pursued the tube versus solid-state argument, seeking to use my ears as the only criteria, instead of looking for the little glowing bottles to assure my conscious mind of a unit’s “tube-iness.” I inquired of tube experts whether my experience was singular or universal. I asked them to comment on some common statements concerning tubes. This marks Part I of the answers they gave. Watch for Part II in a future issue.Several of the designers simplified their replies in order to conform to our space limitations and broad audience. For

MYTH #1: Everyone knows that tubes are “warmer” sounding. This is important because digital recording is so sterile and we need tube warmth to balance out that sterility and harshness.

Oliver Archut: The warmer tube sound is quite a marketing tale; when pentodes and beam-power tetrodes were first introduced during the 1930s, the hardcore triode users at the time (and still today) brought up the same warmer tone argument. Yes, there are tube designs that do sound horribly sterile and just putting a tube into the signal path won’t change the tone response too much.

But what some people refer to as a “warmer” sound is a sum of components and design philosophy. Some classic audio designs require a certain tube or tube maker and cannot be replaced without the sound being compromised.

Hutch Hutchison: There are a ton of misconceptions about tubes including the belief that they sound “warm.” Tube circuits have about as many sounds as there are circuits. It is more about the topologies used and aims of the designer as to how a circuit will sound. In fact, it is usually the transformers that are typically used in tube-based products that create the illusion of “warmth.” There is a lot of vintage solid-state gear that used transformers and is commonly described as “warm.” A typical transformer adds some odd-order distortion at the lowest frequencies and tends to restrict the ultrasonic frequencies, which we subtly perceive because of the phase shift. Part of this effect also seems to smooth transients. Tubes probably tend to offset some of that effect, because a little THD tends to exaggerate transients.

Perhaps the only aspect that is generally true is that tubes rarely sound sterile. Digital converters, and specifically their filters, have completely different problems sound-wise, but it might be over-generous and a holdover from the days when we were told that “digital is perfect” to just use the word “sterile.” It’s a bit more complex than that.

Aspen Pittman: There is a common myth that tubes are “warmer” sounding. It certainly can be said that cranking up a tube amp will make an electric guitar sound “warm, fat, or distorted.” That scenario, however, is one in which distortion is desirable. On the other hand, distortion is the enemy of the engineer who is attempting to record a sound source faithfully and realistically. Here you want accuracy and transparency rather than any coloration that might be described subjectively with a word like “warm.” Fortunately there are many types of tubes and related circuitry that result in a comparatively transparent sound.

As far as tubes “warming up” digital recordings, there seems to be a lingering implication that there’s something inherently deficient in digital recording. While some purists will always make a case for analog over digital, the fact remains that the vast number of pro recordings today are made with digital recorders. Rather than saying that tubes “warm up” digital, it would be more accurate to say that tube mics deliver a truer, more pleasing sound when auditioned against the comparative dynamic improprieties of a solid-state mic.

Doug Fearn: Let’s face it: no recording has ever sounded exactly like the live event. No audio professional is likely to be fooled into thinking any recording, no matter how good, is equivalent to listening to a live performance. But that’s okay, because what we do is provide an alternative experience that isn’t precisely the same, but is potentially equivalent in emotional content.

In the analog age, equipment evolved that was less precise than digital, but provided a reasonable musical experience. All the old technologies (vinyl disc, analog tape, vacuum tube amplification) emphasized the second harmonic as the primary distortion component. Our notion of a favorable recorded sound was based on that experience. Digital eliminates that, leaving us with a more accurate representation that’s often lacking in the “musicality” that the second-harmonic provided. The second harmonic, to some extent, compensated for the fact that the recording never sounds as good as the live sound.


1. Using tubes in a circuit does not necessarily result in a “warmer” sound.

2. What many listeners describe as a “warm” sound often results more from transformers than tubes.

3. Older technologies “trained” us to hear second-harmonic distortion as more “musical.”

MYTH #2: Tube designs are generally slower and mush out the transients and that makes them seem warmer, like tape. Solid-state amps are faster, which makes them less warm.

Oliver Archut: The speed of an electron tube depends on the plate voltage. If designed correctly, a tube is as fast — or faster — than a transistor-based circuit. The slowest part in a tube design is generally the transformer or a combination of coupling capacitor and transformer.

Doug Fearn: Vacuum tube amplifiers usually have far fewer active devices than their solid-state equivalents. In reality, tubes are “faster.” However, poor design and lousy components (especially poor transformers) can cause that loss of transient detail.

To many of us, tube sound is “right” and solid-state is harsh. It’s not that tubes fail to reproduce transients, but that solid-state clipping, especially in mic preamps, adds a burst of odd-order harmonic content to the transients, making them “harder” sounding. This could be desirable in some situations, but with long-term listening, many people find tubes less fatiguing.

Hutch Hutchison: Tubes are quick enough for radio frequency applications and are commonly used in broadcast transmitters today. Once again, in pro audio it’s usually the transformers that limit the frequency response and are to blame (or credit). It’s the transformer that tends to sound vaguely like tape — and should — because both rely on magnetism and have hysteresis effects. With solid-state one of the common flaws is the use of Class AB output stages, which can cause a form of distortion that might be described as cold or harsh. Better solid-state gear suffers less from that.

Mitch Margolis: There’s no doubt that many circuit designs employing vacuum tubes, past and present, are indeed slower then current solid-state amps. By “slower” it is meant that the circuitry’s slew-rate and ultimate bandwidth are lower than that found in amplifier systems of strictly modern, conventional cost-sensitive design.

It should be pointed out, however, that the tubes themselves are capable of furnishing enormously high-speed response while maintaining inherently high-linearity and can do so with very little feedback compared to conventional solid-state amp designs. Bringing these desirable characteristics to the end-user depends largely on the circuit designer’s skills, and the budget limitations on the particular system being built. Circuit topologies, layout, and component choices have a very strong influence on the ultimate “speed” of the finished system.

For example, it’s an expensive undertaking to have input and output transformers fabricated whose bandwidth can keep up with a decently performing tube audio amplifier. It’s these pivotal components that tend to dominate the measured and perceived speed of an otherwise well designed tube amplifier stage. In other words, if the tube can easily handle 30MHz but it is connected to a line-matching transformer capable of 100kHz maximum, the transformer will win the high-frequency response argument.

But there are examples of “warm” sounding high-bandwidth tube production gear on the market, so bandwidth by itself is not the sole characteristic responsible for a given amplifier’s relative warmth or lack of warmth. In addition, the relative “slowness” of some of the tube gear helps in lowering the amount of high-frequency signal energy hitting the anti-aliasing filters in the A-D input chain. This can be very helpful in keeping record-channel sibilance under control.


1. A correctly designed tube can be “faster” than a solid-state component.

2. Well-designed tube circuits have fewer “active” components than solid-state designs.

3. Poor transformers can result in loss of transient detail.

4. The sound of a tube circuit depends on the skill of the designer and the components used.


Doug Fearn: Although the notion of even- versus odd-harmonics has been well researched over the years (see Russell Hamm’s article, “Tubes versus Transistors: Is there an audible difference?” in the Journal of the Audio Engineering Society, May 1973), might there be other factors that explain why tubes sound different? For one thing, tube circuits have far fewer active devices than their solid-state equivalent. Do all these silicon junctions each add some as-yet undefined character to the sound? Most engineers have discovered that the simpler the audio path, the better the sound in many circumstances. Tubes provide a very simple audio path.

I often wonder if there are other factors that haven’t been fully described that contribute to the difference. Some of my experience as a designer has revealed things that sound different, although measurements show no difference.

It’s interesting to speculate what we would have now if tube research hadn’t stopped by the 1970s.