Bartokfan - I believe you and I would probably come to the same conclusion, but this is only one data point. The tube outputs on the Shanling sound better than the SS outputs as well, that is until I mod it. Then the SS outputs leave the tubes in the dust...very dependent on the design. Again, I feel that generalizations cannot be made.
I have heard good SS and good tube designs. Many more SS designs though. If you are looking for rolled-off and undynamic, then even poor tube designs can deliver this. This is why there are so many tube designers out there. It is easier to come-up with a tube design that sounds decent. It is just simpler design. It is also very difficult to design one that outperforms the best SS, particularly when it comes to HF extension, dynamics and bass tightness. The technical reason why this is the case is as follows:
Tubes require high-voltage, therefore the currents are quite small to the plate. Delivering these currents is therefore easier than the equivalent power delivery to bipolar transistors or mosfets. As a result, even "broken" power supply and power delivery designs for tubes can sound decent. I know because I have fixed a number of them. But, making them world-class requires the same care as design for SS. I have found that most tube designers do not have the experience required to achieve this. The Wavac is an exception, but I suspect that even the Wavac could be improved a bit. The other issue with high-voltage is coupling. There must be isolation, either by transformer or by capacitors. These ultimately add coloration. I have tried and measured all types of coupling caps and only within the last year or so have I found capacitors so transparent that they sound close to DC-coupled. Most designers select capacitors that I reject as too colored due to high dielectric absorption, losses or other factors. I have a DAC product myself that has tubed output, so I'm not against tubes. It is my best DAC.
Steve N.
Empirical Audio
Manufacturer