Gentlemen,
If I may provide a designers perspective. I don't want to wade into a he said, she said. I am just providing my design experience.
Class D is going through a maturation stage and developing fast. The prior lack of experience in higher frequencies, above 20kHz, of Audio designers has led to some bad class D designs, while non-audiophiles have entered the market with good designs but bad Audio sound.
Audio sound, aside from being subjective (tubes vs SS), requires real time application and the human ear is very sensitive to sounds not just in the Audio range but outside it as well. Combined they make up the sound. The staging, "air" etc is also made up for many aspects that are not generally measure by Audio designers and manufacturers.
So with that Preamble.
Class D has suffered from lack of high frequency knowledge - designing traces that interacted with each other, designing power supplies that don't have kickback or noise, designing layouts that don't propagate RF, Capacitance in the MOSFETS that don't allow fast switching (that is turning off from on, and on from off - this is not the switching frequency but speed on which it can respond), Inductance of the MOSFETS cases, timing of the deadtime vs switch time and so on. While it is not very complex, Class D requires a lot more knowledge AND design then a Class A or Class B or a tube amp, only because there are a lot more complexities to be considered.
A good Class D will keep the dead time to a minimum, reduce interactions between traces and power supply ( overshoot, ring) and have a clean power supply so you don't hear it. The amp boards and Power supply boards (Switching power supplies) are very sensitive to placement, wiring, components and so on.
Hence just making a amp board that sounds good is not enough. The whole amplifier has to be engineered. Switching frequency, theoretically is nice to have higher, does not work well with MOSFETS, as the capacitance in MOSFETS don't allow for a fast(er) turn off/on.
The filters can be as clean or as horrible as the designer makes it to be. Having a sharp filter on paper with lossy components makes for bad sound as lossy components have a lot of parasitic impedance and will change the whole curve and filter.
Finally the Class D amps are of age where they can be superior to any other class available today - it just takes good design. Class D has many benefits that you cannot get from the other types of amps, including better sound.
BUT as all things, there are no absolutes. It must be better design, better built, better implemented, better engineered to be better. Just saying Class D is better or worse does not make any sense as saying Tubes or SS is better then the other.
Hope that helps a bit with the understanding of Class D
If I may provide a designers perspective. I don't want to wade into a he said, she said. I am just providing my design experience.
Class D is going through a maturation stage and developing fast. The prior lack of experience in higher frequencies, above 20kHz, of Audio designers has led to some bad class D designs, while non-audiophiles have entered the market with good designs but bad Audio sound.
Audio sound, aside from being subjective (tubes vs SS), requires real time application and the human ear is very sensitive to sounds not just in the Audio range but outside it as well. Combined they make up the sound. The staging, "air" etc is also made up for many aspects that are not generally measure by Audio designers and manufacturers.
So with that Preamble.
Class D has suffered from lack of high frequency knowledge - designing traces that interacted with each other, designing power supplies that don't have kickback or noise, designing layouts that don't propagate RF, Capacitance in the MOSFETS that don't allow fast switching (that is turning off from on, and on from off - this is not the switching frequency but speed on which it can respond), Inductance of the MOSFETS cases, timing of the deadtime vs switch time and so on. While it is not very complex, Class D requires a lot more knowledge AND design then a Class A or Class B or a tube amp, only because there are a lot more complexities to be considered.
A good Class D will keep the dead time to a minimum, reduce interactions between traces and power supply ( overshoot, ring) and have a clean power supply so you don't hear it. The amp boards and Power supply boards (Switching power supplies) are very sensitive to placement, wiring, components and so on.
Hence just making a amp board that sounds good is not enough. The whole amplifier has to be engineered. Switching frequency, theoretically is nice to have higher, does not work well with MOSFETS, as the capacitance in MOSFETS don't allow for a fast(er) turn off/on.
The filters can be as clean or as horrible as the designer makes it to be. Having a sharp filter on paper with lossy components makes for bad sound as lossy components have a lot of parasitic impedance and will change the whole curve and filter.
Finally the Class D amps are of age where they can be superior to any other class available today - it just takes good design. Class D has many benefits that you cannot get from the other types of amps, including better sound.
BUT as all things, there are no absolutes. It must be better design, better built, better implemented, better engineered to be better. Just saying Class D is better or worse does not make any sense as saying Tubes or SS is better then the other.
Hope that helps a bit with the understanding of Class D