why expensive streamers

Everyone keeps talking about noise.  I guess this is natural.  but that isn't the main problem.  The problem is dropped bits.  If a bit is missing, a streamer has to improvise.  guess.  fill it in, often with an interpolation.  So there is no noise generated by that, just an inaccuracy.  I think of it as rounded edges and thinned out middles.

The error correction people keep mentioning is generally not the correct term.  For file transfer, there is data correction available.  It will go back to the source and get the correct data.  For streaming, there is not data correction, there is just "data error handling", perhaps "data error mitigation".

Jerry

@jaytor 

respectfully disagree. maybe i’m not following, but not sure your objection is any different for measuring dacs.  regardless, others have done it; for example, you can measure at several different dacs with a control source (e.g., a computer) showing jitter and signal to noise ratios - at the dacs - for both the computer and the streamer, and then do same for other streamers against same control (computer).  that’s just one way, which is how paul miller (hifi news) has done it.  if you look  at his measurements, it shows that dac performance on jitter and noise dominates. that outcome is consistent with what the folks at Benchmark have said - that their dac’s performance is agnostic as to the streamer.  

@invalid technically true, but it is like AM-FM. Both transmit analog but FM is more advanced that AM. Digital is ultimately beams of light in a fiber or modulated currents in a twisted pair or coax cable, or radio waves of a satellite link. but it is also designed to be error-free, otherwise that Word doc you are opening from a server half a globe away would be garbled.

Since when has man ever created something error free or perfect?

Even the gods man has made are full of errors and faults. 
(that's snark, folks)