Hey Michael,
wondering the exact same as you do. I read everything I could find about it, there's even an interview with the Japanese designer of the ADI-2 Pro AD-DA's chips which delves deep into this same topic, but I still don't get it: first of all, and this will sound like a stupid question I know, but why is a filter needed in a converter? Could someone dare to explain it to me in plain words? Understanding this would help, I don't think the graphics on the ADI-2 Pro manual with all the different filter's curves convey this very basic information at all. I guess many people will wonder the same, otherwise just blame my ignorance..
I know from using linear or minimal phase equalizer plugins the differences in transients shaping from the 2, linear phase ones introducing some very subtle smoothing on some setttings (HP expecially). Is it the something similar with the ADI-2 Pro AD DA filters? I can hear that the sharp filters on the RME sound crisper and the slow ones lack a tiny bit of very high freqs.
I've been using the RME for listening to my own stuff, which is just tons of stereo unprocessed field recordings. I record with Sennheiser omni mics (they are more like baromethers than mics..) and know my sounds very well. As soon as I started listening to them with the ADI-2 Pro I discovered stuff I didn't know it was there. I'm enthusiast about how this little device sounds, I probably had heard a similar level of detail when doing the mastering of my stuff at a colleague's studio with Prim DAs.
BTW, even if I love it a lot, I don't consider the RME a neutral sounding device. I perfectly know the flaws and character in my field recording setup, Sennheiser MKH 8020's tend to sound a little boomy and dark, they pick up tons of low-mids and that's the reason I love them, they just enhance the rough organic side of a sound. I also know from years of experience that Sound Devices recorders, which I use, are tuned toward mid frequencies because they are mainly used and designed for dialog, the AD converters in their machines are no exception to this (when I tried to use my 633 for listening to music I was very disappointed to hear this subtle but clear focus on the mids).
Still, when listening to my recordings with the RME everything sounds more crisp and detailed, which is perfect for me and I wouldn't go back.. but I'd dare saying that it has its own sound character which is oriented toward ultra high definition. I feel like there's a slight enhancement to the signal, like applying the Kush clariphonic but of course much much more subtle and natural sounding with the RME. Long ago I had the same kind of feeling when demoing some so called hi fi player apps for the mac, 'better' substitutes to Itunes, like Amarra etc.. They say (their developers) that their sound file reading gets rid of all the interferences from other processings going on in the machine and stuff like that but in the end they don't fool anybody with good hears.. it's clearly at play some enhancing algorithm which explains why all of these apps sound very different. I'm not saying that the RME is fooling me with enhancers or so, and the kind of improvement I hear from this device is worth every last bit of data.. but I wouldn't say it's a neutral sounding processor. That's why I'm very interested in understanding the filter options, given its NON neutrality..I don't think any AD DA will ever be transparent, even just because there's some analog components in them.. leaving apart the chips mess. I'd like to better be able to 'play' with the filter options and to do so I need to understand them better.
The very last nerdy check I did was this, I brought the RME into the field, I used an iphone with the camera connection kit and a portable 12 v battery. I plugged my mics into an SQN mixer I just got, it's an all transformers rather heavy indeed portable preamp from the 90s, one of the very top sounding coloured portable mixer around along with Nagra's and Cooper Sound's. I then captured the balanced out from the SQN into the RME, recording the digital sound with the Apogee Metarecorder app on the iphone. Again, the results are excellent, I'm not complaining.. but well, the sound coming out from the headphone amps in the SQN was one thing, and the one coming out from the RME was a completely different thing.. I know the headphone amp is different.. but come on, I'm working with sound since years, mostly on some very subtly textured low level sounds.. I sometimes use a Sound Devices USB Pre 2, which allows to monitor both the analog input and the signal after the conversion and the difference between the 2 is rather negligible. The 'converted' SQN mixer's sound sounds like a HD version of it, which is just great but I wouldn't call the RME a neutral device..
