So I've done my research and I understand on a technical level (as much as possible) the theoretical differences between amplifier classifications.
My question is a more subjective one. What are the differences in sound between amp classes? Is there a typical A/B sound? A typical class G sound? Or does it depend more on the amp design and components?
The "class" of operation provides certain benefits that one listener might prefer for a specific system. Class of operation is mostly concerned with bias setting for the output transistors (therefore the power supply) in classes A through C and more broadly so through classes D, G, H and T.
Class A has the outputs switched on at all times and therefore has no switching noise or notch distortion. Class AB1, AB2 and C have higher amounts of this specific distortion as the outputs are switched off at more "efficient" schedules in their cycle. This efficiency reduces heat but also reduces fidelity to the original input signal. (One of the many reasons given for the presence of "tube sound" is the time it takes for a "valve type" vacuum tube to actually turn completely off compared to that of a transistor that reacts as an "on/off switch" device.)
Class C is typically relegated to non-critical use such as intercom systems due to the relatively high levels of notch distortion. Class B is found in many lower cost or older sound reinforcement amplifiers. Class A and AB1/AB2 are consumer audio systems with more of class AB2 now being found in pro/studio amplifiers of lower wattage.
Virtually all consumer audio pre amplifiers operate in class A. Any single ended amplifier must by its nature be a class A amplifier.
Class D is "digital" though the "D" isn't meant to designate digital rather just another step in the evolutionary chain of events. Class T is a variation on the class D operation designating a different type of power supply usage. Both class D and T are typically using switching type power supplies (which results in the ouputs switching on/off at exceptionally high speeds) which ups the efficiency of the system but also has the potential for injecting "switching noise" into the signal path.
In home consumer audio class D is typically used in high powered subwoofer plate amplifiers where high frequency switching noise is filtered out by the crossover of the processor/speaker system.
There have been class G and class H amplifiers that never stayed in the market due to various problems. These were more a marketing gimmic used by specific manufacturers and not true classes of operation.
To answer your direct question I would suggest you consider the amount of variance between amplifiers broadly grouped into class AB. A $99 receiver will be a class AB amplifier as will a $20k McIntosh. Most tubes amps not designed as single ended - in other words any push-pull amplifier - will be class AB as will most transistor amplifiers. A forty year old Marantz receiver and a new Marantz power amplifier will both run in class AB.
So, through the last sixty years of amplifier design thousands of amplifiers for home use have been designed around class AB operation.
It's tough to answer your question David. As JV notes, within certain classes (like AB) there are a multitude of different "sounds". Some of the others, like class D, are still evolving. I haven't heard all of the class D amps, but I'm sure that a Rotel would still sound quite different than a Bel Canto or a CI Audio. Perhaps the house sound lies more in the brand (and design philosophy) than the amplification class.
Some really horrible generalizations here. For example there's no doubt plenty of class A amplifiers that "never stayed in the market due to various problems", as there are also plenty that have stayed in the market that, probably still suffered other various problems. None are necessarily relevant to how they sounded. For typical consumer audio the quality is so bad already that it's hardly worth a bother what's used and class H offers respectable quality with increased efficiency for home theatre or whatever, also used for pro audio.
The D in class d isn't "digital", it is simply the next available designator, A, B, C.. Associating the D with digital is erroneous in almost every case, but even for those where it may be questionable based on the method of modulation, they're still completely analog machines.
Forget everything mentioned about class T here. It is just another class d, and not at all a good one. "Class T" is just a marketing trick, or logo, for "T"ripath, the now bankrupted-into-non-existance company that produced the flawed technology.
The class of an amplifier has no bearing at all as per the exact type of supply it uses. class D just as any other can use a battery supply, common in portable applications where battery life is important, due to class d's more effective utilization of the supply, and may be coupled with a switching regulator or a linear regulator.
It can use a SMPS for lightweight in portable high power PA applications where weight is of prime concern.
It can use a standard unregulated raw DC supply, as does the new Mark Levinson class D amp that weighs in at 135lbs and sells for 30 grand, and probably isn't worth it.
Full range class d is becomming common place in consummer audio, which is often of limited quality anyway. Class D has also gotten a firm hold of pro audio, and even high end audio, producing some of the lowest distortion amplifiers known, Halcro for example.
Most commercial class d amps are just modules licensed from companies specializing in class d due to their complex nature there is a serious lack of expertise and as such quite often quality but the possibilities in the hands of the skilled practitioner are limitless.
What makes any amp sound different than another is the quality of the design, ie accuracy of the modulation process, feedback and control, right down to the final degree of implementation, ie. layout and component selection.
It's generally regarded that the very best amps of any class will have more in common than not, as you might think they should. In a really good and flawless as possible implementation of any particular topology the sound will be transparent to it.
You're right, I should have stated class D amplifiers are analog systems. The sentence above should have read, "Class D is thought to be "digital" though the "D" isn't meant to designate digital rather just another step in the evolutionary chain of events."
It is also true that plenty of amplifiers of all classes of operation have been removed from the market due to "problems". Problems aren't restricted to class G and H amplifiers.