The RME, with Auto-Ref (or +13 dBu output level) at 0 dB attenuation (so, showing 0 dBr on the screen), has a MUCH hotter output level than the CYAN unit. You shouldn't need a passive attenuator after it and should be able to run it at MAX (or fixed) output level without any problems. But, to answer your specific question, the RME's volume control is the best software/digital implementation I've come across absent things like DAVE or Hugo 2. Even so, I run the RME at 0 dB myself as doing otherwise, while improving some measurements, has a negative effect on my perception of micro-dynamics. The CYAN's implementation is as close to indistinguishable to the RME's performance as I can tell by listening, but it still sounds its best at max level. In other words it is not a concern at all with the CYAN when used as an DAC/amp (the CYAN is a more enjoyable unit in that role), but if just being used as a DAC I'd bypass it. Hard to say ... mostly because I haven't really spent any time with a Gungnir MB since October last year, and that unit was an April 2017 purchase. They have very different presentations, though. That Gungnir MB had more weight on the low-end, but otherwise delivers a drier, more neutral, performance than the CYAN does in PCM NOS mode. The blacker background and sweeter tone of the CYAN do stand out. Then vs. the CYAN in DSD OS mode ... well, there the CYAN probably winds up being the more neutral performer, and if running HQPlayer to feed it maybe more resolving as well. I would say that, if someone was able to stretch for a Gungnir MB of that era, and wasn't looking to use the CYAN as a DAC/amp (just as a DAC), then a better comparison would be between the Spring DAC L1 and the Gungnir MB.