+ebanksms
Ear/brain systems differ and while technical measurements may be identical, experiences often are not. I’ll try to contribute a little to what millercarbon is saying. Imagine a system that has three removable or tweakable components that add similar noise and/or distortion to an underlying ‘true’ audio signal. Remove component #1 and listen. Hear no difference? Not surprising because the effect is masked by the additional garbage introduced by components #2 and #3. The overall signal/noise ratio hasn’t changed. Add #1 back in and remove #2 – same result. Now change all three components together and wow! All the garbage is gone because the overall signal/noise ratio has increased.
In my case it was going from no chassis damping, a good but inexpensive Sch**t DAC, Amazon RCA cables, and a garden variety Furman power conditioner. Added a steel and Sorbothane platform and cones – no change. Removed the cones and switched to a Berkeley Alpha DAC, definitely some change but for a high cost (plus it’s really a component, not a tweak). Changed back to the Sch**t and switched out the RCA for a balanced Synergistic Research active-shielding cable, no change. Went back to the Amazon RCA cable and switched the Furman out for a popular AC regenerator, no change. Very disappointing; nothing worked.
But then I decided to put in all of the new stuff and live with it for a while before selling it all as useless. I was puttering around a day or so later looking for a new novel to read and I had some background music on. Suddenly the music reached out and gripped me! It was [insert glowing audio jargon here] shockingly different and so much better, visceral. Every component mattered; no single change made the difference.
Aside: My example above is also an illustration of why I believe that as we get closer to the bleeding edge of system performance improvements, A/B testing becomes increasingly less reliable and essentially meaningless because of masking effects. The stimulus response function is just too weak. Double-blind testing with a small sample size of untrained subjects and a weak stimulus response function is the quickest way I know of to produce a false negative. But that discussion belongs in a different thread.
Ear/brain systems differ and while technical measurements may be identical, experiences often are not. I’ll try to contribute a little to what millercarbon is saying. Imagine a system that has three removable or tweakable components that add similar noise and/or distortion to an underlying ‘true’ audio signal. Remove component #1 and listen. Hear no difference? Not surprising because the effect is masked by the additional garbage introduced by components #2 and #3. The overall signal/noise ratio hasn’t changed. Add #1 back in and remove #2 – same result. Now change all three components together and wow! All the garbage is gone because the overall signal/noise ratio has increased.
In my case it was going from no chassis damping, a good but inexpensive Sch**t DAC, Amazon RCA cables, and a garden variety Furman power conditioner. Added a steel and Sorbothane platform and cones – no change. Removed the cones and switched to a Berkeley Alpha DAC, definitely some change but for a high cost (plus it’s really a component, not a tweak). Changed back to the Sch**t and switched out the RCA for a balanced Synergistic Research active-shielding cable, no change. Went back to the Amazon RCA cable and switched the Furman out for a popular AC regenerator, no change. Very disappointing; nothing worked.
But then I decided to put in all of the new stuff and live with it for a while before selling it all as useless. I was puttering around a day or so later looking for a new novel to read and I had some background music on. Suddenly the music reached out and gripped me! It was [insert glowing audio jargon here] shockingly different and so much better, visceral. Every component mattered; no single change made the difference.
Aside: My example above is also an illustration of why I believe that as we get closer to the bleeding edge of system performance improvements, A/B testing becomes increasingly less reliable and essentially meaningless because of masking effects. The stimulus response function is just too weak. Double-blind testing with a small sample size of untrained subjects and a weak stimulus response function is the quickest way I know of to produce a false negative. But that discussion belongs in a different thread.