With so much high‑end analogue processing gear using valves these days, it's easy to assume that they invariably equate to a better sound — but that's not always the case. Paul White explains.
Though valve equipment has been favoured for years for guitar processing, there's recently been an upturn in its popularity for studio use. But why? There's no disputing that well‑designed valve equipment has a certain sonic 'something', but there's still no clear agreement over exactly what that something is. Some people would have us believe that valves distort in a subtle but interesting way that makes the sound 'warmer' or more detailed, while otherssuggest that the point is not that valves are magical, but rather that solid state devices actually damage the sound in some way that isn't shown up by simple measurements. It doesn't really matter which, if either, of these hypotheses is true — what's important is identifying where valve equipment can make your sound better, and where it might make it worse.
As I pointed out in a recent Leader column, high‑quality vintage valve gear wasn't designed to deliberately distort — the designers did their best to make the valves in their circuits behave as linearly as possible, so any remaining distortion should be quite low in level. Even so, we know that when valves are driven hard with peak signals they do tend to behave in a non‑linear fashion, and this behaviour is a lot kinder to the signal than the hard clipping associated with solid state circuits. However, a lot of older valve equipment also includes audio transformers, and these too can saturate in a non‑linear way if pushed hard, so it's not always easy to say exactly which part of the circuit is responsible for which aspect of the sound.
Some people suggest that the point is not that valves are magical, but rather that solid state devices actually damage the sound in some way that isn't shown up by simple measurements.
The increased distortion on high‑level transients exhibited by valve equipment does correspond, at least in part, to what we hear when using a harmonic enhancer or exciter. Adding small amounts of controlled high‑frequency distortion creates artificial harmonics that are musically related to the high end of the existing signal, and that has the effect of making the signal seem more detailed or transparent. This is quite counter‑intuitive, as you'd expect distortion to make things sound less clear.
Valve distortion also seems to produce this 'clarifying' effect, but only when the amount of distortion is relatively low and confined to transient sounds. If the drive level is increased so that even average level signals are distorted, pure musical sounds (such as flutes, or synth sounds with few harmonics) produce audible intermodulation distortion products, based on the sum and difference of the frequencies present. For example, if you were to play two sine‑wave pitches of 200Hz and 250Hz at the same time, the intermodulation products would be at 450Hz (sum), and 50Hz (difference). If you carry this over to musical intervals, you'll invariably find that the intermodulation products are unrelated to the musical pitches you're using, so the result sounds dissonant and unpleasant. The only exceptions are those musical instruments that we traditionally associate with very high levels of distortion, such as the electric guitar and the rock organ, where the intermodulation products have become an accepted part of the sound. Even then, a speaker that rolls off at high frequencies is required to prevent the result from sounding harsh and unmusical.
How Much Distortion?
Unfortunately, a number of modern tube products generate rather too much intentional distortion to be useful on all types of material. For example, a highly coloured mic preamp used on a pure voice can sound quite unpleasant, whereas a harmonically rich voice processed through the same device might sound fine. My own preference is for relatively clean‑sounding valve gear, but if you like the option of adding more obvious distortion when the material warrants it, a device with variable valve coloration will obviously offer the most flexibility. Drum sounds, for example, often benefit from relatively high levels of distortion, as their transient nature makes it difficult for the human ear to detect any obvious distortion. Instead the sound simply becomes punchier, more 'attacking', or crisper.
Because different audio material responds differently to valve coloration, this begs the question of how a complete mix should be processed, because, quite obviously, a typical mix contains many different sonic elements. I have to admit that some valve processors do make a complete mix sound sweeter and better integrated, but once again, the units with the clearest audio path seem to do the job best. If you need to be more heavy‑handed with drum and bass sounds, routing these sounds via a stereo subgroup, with a less subtle valve processor plugged into the insert points, is more likely to produce a satisfactory end result.
At this point you might wonder why harmonic enhancers and exciters can work well on a complete mix, even when the level of effect is quite noticeable, but the key to their success in this application lies in the fact that they confine their processing only to the higher reaches of the audio spectrum. Valve distortion, on the other hand, will affect everything. For this reason, you have to be quite careful about where and how you apply valve distortion.
Valves are inherently noisier than solid state devices, not least because they run at high temperatures, and electronically generated noise is proportional to temperature. If valve equipment is designed properly, its thermal noise level can be kept low enough for serious studio work, providing not too much gain is required. In equipment where a lot of gain is required, such as microphone preamps, it's normal to use an audio transformer before the valve, to increase the signal voltage. As valves are inherently high‑impedance devices, they team up particularly well with step‑up input transformers. Unfortunately, really good audio transformers are expensive, so you'll find that most of today's less expensive valve mic preamps use a solid state input stage instead of a transformer. This hybrid approach can work well, but purists will tell you that they don't sound the same as an all‑valve circuit with a transformer at the front end.
Valve microphones use a valve preamp inside the mic body, and, for whatever reason, the best vintage models sound very nice indeed. Their noise performance tends to be a little worse than an equivalent solid state design, but when you're miking instruments or voices at fairly close range, this isn't really a serious issue. Some companies, such as Sony, have built valve microphones with elaborate cooling systems to try to improve noise figures, but in most routine music applications this isn't necessary.
Unfortunately, a number of modern tube products generate rather too much intentional distortion to be useful on all types of material.
If you choose to use a solid state mic and then process this via a valve preamp, it's possible to gain some of the tonal benefits of a valve mic, but unless the preamp has a transformer input stage it could be argued that you'll do just as well using your regular console mic input, with a line‑level tube preamp in the channel insert point. I've done this using a Ridge Farm Gas Cooker preamp (which has variable valve drive), and it seems to work very well. However, no two audio chains will ever give you exactly the same result, so being too perfectionist in your pursuit of a specific sound you once heard on a vintage recording often leads to disappointment.
While valves can be made to run reasonably quietly if proper attention is paid to circuit design, there are other sources of noise specific to the valve. One is so‑called 'microphony', where the internal metalwork of the valve vibrates in response to an external sound and ends up modulating the signal being amplified. Guitar players are familiar with this, as their valves often share the same cabinet as their loudspeakers, but pro‑audio equipment can suffer from it too, especially if the monitoring level is high. The usual cure is to select non‑microphonic valves when choosing replacements, but you can generally tell if you have problems by tapping the valves with your fingers and listening for corresponding noises through the speakers.
Another common source of noise is oxidisation on the valve pins, caused by the heat of the valve. This usually results in intermittent, 'spluttery' or 'fizzy' noise, caused by varying contact resistance. In these cases, cleaning the pins and valve socket using a proprietary electronic cleaner, such as Deoxit, is a good idea. In older equipment, heat from valves can cause accelerated ageing in nearby capacitors and resistors, again resulting in noise (or even complete failure), but unless you have a good grounding in electronics, tracking down and replacing these is a job best left to the professionals. For these reasons, owning and maintaining vintage valve gear can be expensive.
Good valve circuitry can have a positive effect on sound, but you can't assume that a unit will sound good simply because it has a valve or two inside. I'm a fan of good valve microphones, as they do seem to present a better integrated sound than most solid state models, but if you haven't had a lot of experience with tube products, you might find the differences are smaller than you expected. Indeed, valve processors that make the biggest tonal differences are also likely to have the most detrimental side‑effects, unless there's a proper facility for adjusting the amount of valve drive.
Processing a full mix via a quality valve equaliser or valve compressor can bring about a subjective improvement in sound quality, especially if the source is all‑digital and a little sterile‑sounding, but beware of anything that seems to change the sound excessively when any EQ or compression element is switched out. Try to listen to the music without concentrating on it too much, and see if the effect is smooth and involving or harsh and dissonant. A mix with a brash edge may seem impressive when you first hear it, but if it has the effect of making you want to keep turning the level down it's probably not the right effect to use. A good processor will leave the music sounding transparent and cohesive, without making it aggressive. Similarly, the much‑vaunted warmth of valve processors should manifest itself as a more solid bottom end, not as an obvious bass boost or bass muddiness. You'll also find that some solid state devices have a sound that is more 'valve‑like' than some genuine valve processors, so don't be fooled by the glow inside the box. Let your ears decide for you, and always use whatever does the job best.