Yesterday, a friend of mine came with insane chain gear to try out my headphones. Gear I wouldn’t dare buying ever. A Gustard R26, Aune S7 pro, RME ADI-2 pro FS R. We have R2R dac, TOTL delta sigma with EQ, and class A amplification. This is TOTL for a lot of people and considered among the best in their price range for subjectivists. Compared to my Ifi micro idsd black label and stack of Topping E50 and S.M.S.L SP200, it should be night and day.
And when I first plugged my D8000pro, It was indeed night and day. Going from my stack to only the RME on Susvara and D8000pro, the scene was wider, more holographic, it was insane how my stack was soooooo bad. Where is that coming from ? I heard the DCS stack and Raal amp before and I thought there were no differences at all except volume control and features. Was I wrong.
Later that night, we decided to blind A/B the RME vs Ifi output (with switch box). Also, we compared only the DAC section of RME vs Ifi, DAC section of RME vs Gustard, both times connected to the Aune this time.
One of us would wear a headphones in the next room (I got long cables), the other person would switch the DAC input on the Aune randomly. Changings caused a 0.5sec cut in the sound. But we could also fake that change.
Conclusion : Both of us couldn’t differential anything, it was too hard. Him with his great gear, me with my skepticism but with doubt after hearing the RME (it was our biais). Sometimes there were differences, but most time it was just the volume not being 100% perfect. Giving the advantage to which gear was even the slightest louder. Nothing like wide soundstage, resolution, imaging… We will do more tests in the future as my Stack has not been compared to his stack yet.
And you, when was the last time you did an blinded A/B test ?
Yeah, a (long) while back I did an ABX of known hard to encode samples and my regular music, found the setting I could no longer tell a difference, and added one more quality level to the encoding. I forget exactly what the level was, but around 192kb/s the best encoders for Vorbis and AAC were pushing transparency for me, and by going with 256 kbps VBR I was satisfied they were better than required and I could never tell the difference in my tests.
10 years later I can’t really hear anything over 16Khz anymore, and pretty sure Vorbis/Opus/AAC at 256kb/s is fine for me despite the fact my equipment is a bit better than what I had at the time… I still have FLAC locally just because disk space is so cheap (have over 30TB of disk), but streaming spotify at Very high (320kbps Vorbis) gives me no pause, and high is probably fine for most listening environments/gear. Compression is a solved problem, and if you want better quality, you need to find better masters. Most high quality streaming services do in fact start from better masters when available, but then try to claim the difference is their compression/bitrate/sample rate/bitdepth, which is just bollocks.