Is Tidal reducing the bit depth of their "HiFi" streaming to reduce their need for bandwidth?
I'm a little embarrassed to admit that I've had Tidal for almost three months now but only just this morning decided to A/B the following two chains:
Tidal HiFi > iPad 3 > Pure i-20 Coaxial Out > Metrum Octave MkII > Metrum Aurix > HD800
44/16 WAV > FiiO X5 Coaxial Out > Metrum Octave MkII > Metrum Aurix > HD800
Playing a well-recorded track I'm very familiar with, Michael Jackson's Billie Jean, swapping between Tidal HiFi and the Fii0 X5 as SPDIF sources to the Coaxial input of the Octave MkII, I was surprised when I immediately noticed a big difference in volume at the HD800. I started pulling up some other tracks for comparison and, with every track, I am hearing about a 1.5 dB lower SPL at the headphones using Tidal HiFi > iPad 3 > Pure i-20 as the Coaxial source.
This could be a problem with the Pure i-20 docking station's Coaxial output, but other than a TeraDak Teralink X2 USB-to-SPDIF converter that's currently packed away, I don't have any other Coaxial output sources to play with, as a tie-breaker.
I may be wrong, but to my thinking, there's only one way to reduce the volume of an SPDIF source - by reducing its bet depth. So, although I have no evidence, I am considering the possibility that Tidal is reducing its bandwidth requirements by reducing the bit depth (from 16-bit to something less than 16).
Is there anyone else here who can compare the SPL heard at the headphones when playing Tidal HiFi (44/16) vs. a 44/16 WAV or FLAC file of the exact same artist/album/track? If it really is a due to a reduction in bit depth, then it wouldn't matter what chains you use for the comparison, as long as your comparing a true Redbook file to the Tidal HiFi equivalent.
Oh, I just talked myself into another experiment I can perform myself - with these two chains:
44/16 WAV > Windows 7 > Foobar 2000 > USB > CEntrance DACmini CX > LCD-2 rev.1
Tidal HiFi > Windows 7 > USB > CEntrance DACmini CX > LCD-2 rev.1
OK, I'm convinced - even when comparing these two chains, the Tidal HiFi source gives me a lower volume at the headphones. I just now compared Jennier Warnes' First We Take Manhattan, which has some dynamic drum hits, etc. Playing from a 44/16 WAV file (vs. Tidal HiFi), it's not only louder at the headphones by at least 1 dB, it seems to have more punch. Tidal HiFi sounds fine until I make a direct comparison with a WAV file - then it sounds compressed. (Note that the "HiFi" indicator is lit in Tidal.)
UPDATE: Using Eva Cassidy's Blues in the Night, from her Eva By Heart album, it's really obvious, right at the beginning of the song, that her voice has a lot more "snap" with the 44/16 WAV file than with Tidal HiFi's streaming of the same track.
I'd really like to hear anyone else's observations when making similar comparisons between Tidal HiFi and a Redbook file - using the same DAC, amp and headphones.
Mike