That's from a listener perspective, If I had to take a PSWAG as to why they sound different, I would think that the power stored / available in the led sled has something to with it. With the bigger transformer and caps they are built with they pull a whole lot more current than the class d's, having that current on tap to drive the output would be the difference one would think.
Curiously...... the usual class-D can do MUCH better than most class A-B units on current, unless teh current limits are set poorly......
A class-AB amp pulls one amp from the power supply for every amp on the output.....
BUT a class-D amp pulls LESS than one amp per amp for almost all output levels..... the difference is the efficiency. Sounds like some sort of "free lunch" but it isn't.
At a particular time, the output voltage of each amp may be the exact same, and that may be, let's say, 1/3 of the power supply voltage.
The class AB is supplying 3 amps output, which means it is drawing 3 amps average current from the power supply.
The class-D amp is ALSO supplying 3 amps output, but it is only drawing 1 amp average current. The reason is that the Class-D is about 95% efficient..... the class-AB is lmuch less efficient.
That means that if the voltage out is 24 volts, at 3 A, both are putting out 72 watts, from a power supply of about 72 volts.
But, the draw from the power supply of the Class AB is 3A x 72volts, or 216W. Since only 72 watts is coming out as signal, 2/3 of the power is being lost as heat in those massive heatsinks. That's 33% efficiency at that moment.
meanwhile, at 95% efficiency, the Class-D is drawing 76 watts, which is a draw on the supply of 1.05 A at the same moment. A considerable difference in the demand for supply current, for EQUAL output currents.
With equal power supplies, the class-D can generally make far better use of the power supply energy storage..... OR for equal performance, the class-D power supply can have less storage capability.
The only time the Class-D actually draws 1 amp per output amp is at the very peak of the signal at a level just before clipping.
The argument about power supplies can only be true if the Class-D designer went overboard in assuming how much could be cut out of the power supply on account of efficiency.
