The post below may be a duplicate, I thought I had made these comments but did not see them in the thread, so ...
There is some sort of continuum of responses to changes in streamers affecting SQ, from none, they all sound the same, to night and day differences.
I think that all responses may be correct, in that DACs appear to have greater or lesser sensitivity to the source of their data.
In the earlier days of digital, different CD transports and different digital interconnects had massive effects on the resulting audio from the attached DAC, and products such as the Digital Lens were created to alleviate the problems. I believe that in early DACs there was little (if any) buffering of the data and the DAC's clock was reconstructed, via a PLL, from the input stream and therefor the timing of the clocking of the data to the DAC processor was dependent on the quality of the input.
My DAC is the Esoteric K-01XDSE, with an external rubidium clock, and, to my ears, in my room, though my amplification and speakers, does not seem to be affected by changes to the source component. A change from the Aurender to the T8 has not affected SQ in any way. nor would I expect it to. I use an asynchronous USB connection and the DAC has engineered isolation of the circuitry from the source. It has a FIFO buffer and the clocking is totally independent of glitches on the input stream. Note that the chain outbound from the DAC (LA4, XA25, Atria II) is very transparent, though perhaps the Cardas Clear Reflection interconnects is slightly less so.
Other DACs may have prioritized their circuitry differently so that they are affected by the input timing, and that the sound is changed by different sources. It appears that the Fezz is one of those.