A side-view of the RND Shelford 5052 Preamplifier
RND Shelford 5052 Mic Pre Rupert Neve Designs

Let’s break down each part of this: “class A” is a type of amplifier circuit that reproduces the whole audio signal. This is different than a “class B” amplifier which only produces the top or bottom of the sound wave. class B amplifiers are used two-at-a-time so one reproduces the top while the other reproduces the bottom. Class A has the highest fidelity reproducing both together because there is no “crossover distortion” or, more simply put, imperfections when joining the top and the bottom signal from the class B style circuit. The negative about a class A circuit is that it uses much more power and thus creates more heat.

“Discrete” that means that the primary circuit for the mic pre uses individual components that are selected for the best combination from the point of view of the designer (the components are seperate, or “discrete”). For example a tube, op-amp, or transistor are examples of a discrete components found in a mic preamplifier. Alternatively an integrated circuit (“IC”), or “microchip”, is a small package that contains a very dense layout of semiconductor components.

There is a deeper meaning to whether a mic pre, compressor, or converter is class A and what that means to you, the purchaser: A class-A, discrete piece of gear was most likely designed to be higher-end with less compromise than the average integrated circuit design. Though, a poor design can still be marketed as “class A, discrete” because a popular thought is that to be of a certain quality a microphone preamp must be both class-A and discrete. That is a myth.

There are preamplifiers that are considered among the best that break one or both of these rules.

Let’s take Rupert Neve for example. His design for the Neve 1081 is not class-A, it is class-AB yet it is considered one of the very best preamps of all time.

My Grace Designs Model 101 (ribbon version) is a distinctly clear-sounding preamp that is wonderful for its fast transient response, meaning it catches a more accurate representation of the sound wave, particularly the leading edges of a sound. But it is not a discrete design but rather an integrated circuit.

It all comes down to the designer and the design they they make. With that said one can often generalize that truly discrete and class-A piece of audio gear will sound good because of the effort it takes to make one.

This is because while it is more costly and difficult to build a preamp with discrete components it is also often an indication of quality as it reflects on the designer’s intention to achieve a performance goal regardless of cost. If trying to create a really nice-sounding preamp it is easier for many electrical engineers to select and optimize components than having to build a dedicated integrated circuit chip since creating a custom chip requires a lot of resources and work in comparison. Additionally most integrated circuits (read: the pre-made ones) cannot handle the voltages that discrete components can. A high voltage can be very beneficial to headroom, dynamics and frequency range. A clever designer like Grace Design can do amazing things with an purpose-built integrated circuit (with a higher voltage and bias than the average integrated circuit) but a microphone preamp is not a complicated circuit and usually those seeking this certain level of quality will go with discrete.

Let’s be mostly discrete

My own opinion is if the integrated circuit in the Grace Model 101 is specifically designed that kind of makes it a discrete component of sorts in the sense of the non-technical word, right? To me the important part of ‘discrete’ from the prospective of one who’s job is an intersection between physics and black magic is that everything that was picked to be specifically a part of a mic preamp. None of the components were designed to also do other things beyond their job in the chain of the analog audio amplification part of the preamp. If a chip can amplify but also check your email it’s not as focused. There will be less fidelity because it is not purpose-built, other parts of the circuit introduce noise, distortion, etc. WIth discrete (individual) components specifically selected all the components in the chain can do is is their part in amplifying the gain from mic-level to line-level along with whatever features were specifically designed into the preamp e.g. a low pass filter. Everything is there with a purpose.

One trend that stays very consistent is that discrete circuits are usually more expensive than integrated circuits. Another easy generalization is that integrated circuits are more easily (and commonly) smaller than discrete circuits.

For all of these reasons prosumer audio interfaces will usually have integrated circuits and high-end pro gear will usually be discrete/class-A which brings us to:

You’ve got A lot of Class

“Class-A” gear uses a single device (tube, FET, or transistor) to do the full swing through of the signal through both positive and negative. Class-AB and B use a separate amplifier device for the positive and for the negative.

In an audio waveform there are moments where the wave is above silence, below silence, and at silence. A class-B amplifier pushes the top of the wave up with one device on the circuit and pulls the trough of the wave down with a second device. A class-A amplifier uses just one device for the entire input signal (the conduction angle, or “Θ” equals 360°). Because it is both amplifying the positive and negative sides of the wave at the same time but the wave can only be either in the positive or the negative a class-A amplifier device must constantly dissipate ½ or more of its available output power as heat — meaning it is wasted.

Class-B and class-AB dual-devices (Θ = 180°) don’t have this issue, making them more efficient because it splits the sound wave into separate positive/negative (or top and bottom) parts and and uses a separate device for the positive and negative parts of the sound wave but the drawback here is there is a distortion created at the point of separation because one might have switched off before the other has switched on. Class AB circuits are biased so that both are on at this crossover point. The benefit here is that a class B or class AB circuit will be more efficient because with the two output devices reproducing the upper and lower waveforms are switched off when not in use and already use less power on top of that.

A class-D amplifier (found in many loudspeakers) uses a pulse-width modulation instead of directly using the input signal at all and is much more efficient yet retains less fidelity

A class-A design can be simpler than the rest because the concept itself is simpler, has generally better high-frequency performance (after all the higher frequency the wave crosses ‘0’ the more often there is distortion which runs into issues kinda related to the nyquist theory), reduced high-order and odd-order harmonics, no crossover distortion (and low distortion in general).

However class-A amplifiers are inefficient and rarely used outside of audio. The old studio consoles were filled with them and required a hefty power bill (but also doubled as heater). While glossing over a lot of technical points the power-inefficiency is essentially the reason a class-A amplifier has more headroom as well. But this means a more expensive power supply and better heat sink is required.

Class-A designs are more demanding on components and harder to design and build. But this also means that the components are of higher quality

You’ll find transformers in many audio devices as well and these can provide a passive (meaning no power-amplification) gain, can help match signals/impedence, and provide galvanic isolation between things. But a circuit might be advertised as ‘transformerless’ because transformers can potentially introduce phase shifts, can have a ‘sound’ if used in some ways, and can be large and heavy.

Tubes provide amplification and can give a tonal character but they can also be used in different parts of circuits and different ways to create very clean audio as well.

Here is more detail and some citations: gearslutz gearslutz gearslutz cakewalk forums power amplifier classes.

What makes a good preamp? That’s for another article but the short answer is a good designer.