The question of volume control is a complex one to answer. Somewhere between the DAC and the amp, we have to adjust the volume. It can be done a) in the digital process, b) in the DAC - analog domain, c) in the passive preamp, d) in the active preamp and e) in the amplifier.
Each way can be done well or badly, each one is equally good if done well. Many people would scream that digital volume looses bits, but I would defend this but this is not the subject of this post.
In our DACs we use the highest existing level of volume control quality. if level 1 is the pot, level 2 is the stepped rotary switch with discrete resistors, level 3 is microprocessor controlled relay network with one resistor pair in the chain. We are level 3. (well, there is also level 4 with non-binary ladder, but it is as rare as white raven. (We are working on releasing one passive preamp like that). We install this volume AFTER tube stage and just before the output and we find this scenario optimal. Theoretically, if our DAC works into the next component with volume control at the input, the two controllers work simply in parallel. It does not matter technically, if A is maxed out and B adjustable, or vice versa or both somewhat adjustable. For listening - each scenario can bring a different flavor and each one is "allowed". Do we need a preamplifier in the chain - with associated cost, shelf space, pair of interconnects and power cable - we don't, but if the user finds it useful or more attractive sounding - it is fine with me. Personally in such scenario I use DAC fully open. The series resistance is shorted then, and parallel resistance is a straight 10k to the ground.