Source or speaker

Think about what the various parts of a system have to do. Your speakers are the link that moves, or tansduces, the information from an electrical information to sound mechanical information (driver movement) to sound information. This transduction between ways of 'displaying' the information brings in all sorts of subjectivity and problems with comparative measurements. The speakers (an of course the room acoustics) perform a function that can only be tested as to accuracy by going back to the original performance from which the information was obtained. All other parts of the system are in one environment where each stage (whether an internal amplification stage, input and output of a switch, or from one box (such as a preamp) to another (such as an amp) can be tested for accuracy with methods and measurement resolution that swamps any noticable differences. As to working with a digital source - the only movement you have there, still in an electrical enviroment, is from the digital word to the analog equivalent. Again, what makes digital so advantageous is noise immunity and that the resolution as far as audio swamps noticable differences in the conversion from digital to analog. As to cables and interconnects - this is of so little consequence (assuming you aren't connecting everything with corroded #28 wire) as to not warrant attention - as someone as stated - just buy large wire. So, in short - the speaker and the room are pretty much the place to put efforts and money. One thing I have always wondered - we often start with the idea that a flat response across the audio freq range in a given room and true representation of the signal give us the the desired outcome. However, we do not take into account our own transducers - the physiology. Not really an issue if one has "normal" hearing. But for persons over 50 years old or who have been exposed to acoustic trauma (including live music at 120 db for hours). Seems like the ideal to get the information into our heads would be to start with audiometric testing, thus obtaining where the individual had db loss (i.e. perhaps 10 db down at 4000 hz, 20 db down at 8000 Hz. normal below 4000 Hz. etc) and then using appropriate attenuators to create what would be for that individual a flat or accurate response. Just a thought. But until then - money is best spent on speakers not digial sources. If you want to try it yourself do blind A/B tests at a showroom between the cheap model digital source and the one that costs 5 times as much- this likely will be discouraged by most salesman - I have heard " Oh our switching gear is not very good and you won't be able to hear the difference here in the showroom" Interesting.