I find it hard to imagine that a reviewer can be 100% sure that a 16/44 on any of the streaming services is the same master that is on a cd that was used for comparison
In the post where I quoted from a review, I also put a link to the original so anybody interested can see the full context, which I took to be a generalisation across the industry.
You have identified one way in which digital sound quality may genuinely differ depending on the source.
There must be a surefire way of finding out - stream to a file and use a computer program to do a word-by-word comparison, maybe?
I very recently bought my partner a set of Barbirolli conducting Elgar on CDs, some of which are famous recordings which have been full price in the catalog for decades. The new set did not sound right and it turns out they have been remastered in 24/192, which possibly means remixed as well?
Another way for digital to differ is if what arrives at the dac, differs from the source. If any digital link in the chain does not implement full error detection and recovery, that’s a potential variance. Any digital that travels over an I2S cable connection is suspect. Same for Ethernet unless it is within an end-to-end error correction protocol.
If you can be sure your digital is arriving perfectly (which requires a memory buffer and an on-board clock), the next potential source of variation is analogue variability and interference.
After that, differences can only be psycho-acoustic in my book.![]()

