Law of Accelerated Returns

So we have a proposed performance curve which is very pretty and supports diminishing returns towards the higher end of gear, but what about the PERCEPTION curve? Asymptotes can work in either direction! Say that we have a threshold for "suspension of disbelief". If I was sitting at 99% of this threshold, and I upgrade for a 1% absolute improvement, now I’m suddenly enjoying music past the threshold where it "feels" real - i.e. the 1% difference is perceived as a VAST improvement. If breaching this threshold was the primary goal, then any amount short of it (no matter how small) represents failure, and not good value! Of course the threshold varies by record, day, and mood, etc. And this is all a vast over-simplification (as is our limitation for discussions here). Human perception makes everything messy lol.

Well, that performance / response curve is representation of average human behavior, like all theories.  Like I said, there are many human factors affecting the pattern of the curve.  In the other topics I posted that discussed the DR, I touch on the possibility of "measuring" (or "quantifying") the SQ and, if we could do so, it will help ruling out significant part of human perception disparity and provide us a more objective way for gear evaluation.  Part of the discussion pertains to creating metrics and standard procedure to measure SQ via the measurement of quantifiable psychoacoustic "entropy."  One aspect of theory you might be familiar with relates to analyzing certain types of musical harmonies, such as the even order harmonics give smooth, rich more pleasing sound and odd order harmonics give edgier more exciting sound.  Well, just my 2 cents.

@lanx0003 

Excellent… I like it. The question is, can you keep your own situation out of it. Or, are you unable to not imprint: “I can’t afford it, therefore it is not worth it.” I am able to appreciate, let’s say a million dollar system (using $ as a proxy for sound quality). At some point, I am sure I would say we were into the diminishing returns. But that level is probably quite high.

IMHE this is the hardest characteristic to reproduce. Energy created by the room and reflections blur out the third dimension. Channels that have different amplitude and impulse patterns also blur out the third dimension. These three problems compromise at least 90% of the systems out there. Some speaker/rooms will never be able to perform at this level. Others can but are not adjusted correctly. Only by luck can you get this out of the box and only the very misinformed are going to get there by placing little discs next the their interconnects. Do you have to spend a lot of money? Depends what you think a lot of money is. I think you can get there at a lower volume for maybe $50K. The full Monty takes at least $100K for a system with a turntable. Many are spending $250K just for speakers.

A lot of megabuck systems are crystal clear, cover the full frequency range, have cavernous soundstages with pinpoint imaging, but fail to do what some more modest systems can achieve in terms of PRAT. and musical expression.

My workshop system uses a 46DHT direct coupled parafeed amp using custom Magnequest iron - only two tubes in the signal path, into high efficiency full range drivers. I listen nearfield and do not get a huge soundstage, nor is the imaging what you would call pinpoint, neither does the frequency response approach full range. But the speed and articulation is mercurial, and all the musical inflections are laid out with great clarity. This enables me to get a great deal of satisfaction listening to music, rather than using music to listen to a system..

I can’t think of any valid examples of comparisons where marginal returns in audio gears is increasing.

The reason is that the fundamental technology must remain constant. This is mostly an engineering issue, and usually there are more features as you pay more and this is a similar issue. So, I should stop now. But just to illustrate -

A trite example, about as generic as I can think of - balanced connections added to a DAC. Topping, Gustard and others have this kind of range, and they cost more.

The balanced is supposed to be better than just RCA.

Let’s pretend that the percentage "betterness" is more than the percentage more than you paid, and Harley announces that he has discovered increasing marginal returns.

This isn’t correct. He’d be wrong. There are two different technologies on the table.

There are plenty of other examples.

Edit - there maybe obvious engineering thresholds. Below the threshold, blah!! Just add one more tiny component, and it goes from blah to actually OK, a hundred times better. This is hardly relevant to what Harley is talking about - high-end, or at least pretty good stuff to start with.

.