docholliday
100+ Head-Fier
- Joined
- Mar 14, 2017
- Posts
- 445
- Likes
- 354
I just listened and compared these 3 tracks (not FLAC, but 320kbps mp3) between N3 and X1ii, using UERR and MDR7506, went back'n'forth multiple times, and don't hear any issues with drums/percussions as you have described above. I assume you're referring to some ADSR envelope anomaly you hear, perhaps release time? What I would recommend, in Music Settings take a look at Digital Filter (AKM4490 has different filters) and play around with it. I'm using Sharp filter. Also, my N3 already has over 35hrs of burn in.
Just a quick follow up, the issue was related to the EQ bug when associated with High gain. Once the EQ would stop working, it would make all the cymbal rings cut off waay early and the drum resonance dry. A reset of the player (and subsequent reconfiguration of settings BEFORE playing a track) resolved it. Also, in high gain, the cymbal resonance was much less, when working correctly, then in low or medium. The player was originally set to high gain to keep the volume 35-45 for the UERR (which is what the X1-II was set at for the same output level). In Low gain, it requires 62-65 on the N3 for the same sound output level.
Once I figured out how to keep it from going wonky, I did look at the design documentation of the AK4490EN chip and the 32-bit, 8x sampling SCDF (Sound Color Digital Filter) cutoff filter. To begin with, I had a session drummer lay down a nice track for me (since I was in the studio anyways) to use for testing, with full cymbal solo. I monitored and recorded it via the board with the UERR as well as the Genelec near fields. I also listened from the other side of the glass to get the recording as accurate as possible.
I played it back first on the Fiio X1-II to see how it sounded and it was decently close to the studio recording. Then, on to the N3, set to L gain with no EQ...SHARP was the worst filter of all, with the clean brassy resonance of the ride cymbal losing all harmonic resonance and became a vibrating sheet of aluminum foil. The full pre and post ringing of the harmonic cutoff made the mid drums sound solid with little harmonic variation.
The most accurate was the SUPER SLOW, which actually drops the SCDF's sampling to half, and resulted in the closest sounding cymbals as well as drum "thumps" that had harmonic resonance. The next closest was the Short Delay SLOW.
I also had a go with the keyboard and noticed that the recorded track did work best with SHARP, but wasn't much worse with any other filters. I also tried both tests with a light EQ set at 63/+1db, 125/+1db, and 16K/+1db. Same results, with a bit cleaner ring to the cymbals and drums were a touch more "accurate" - the UERR is known to be very flat, so I gave it a push for the final test.
As long as I don't touch the EQ and/or set the EQ with nothing playing, power down and restart, the issue hasn't returned and all test tracks sound very close to how they play on other devices.