The last 3%

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The "do all amplifiers sound the same" thread has no answer, but IMO all decent amplifiers certainly do sound similar. The same can be said for most other electronics, with the exception of speakers. Even those are converging more than they were 20 years ago. Regardless, I think everyone would agree that, even if all amplifiers do sound the same under certain conditions, when you lash it all together, different systems sound quite different. Giant big duh. But, it seems we spend most of our time tweaking and agonizing over the last 3% of the sound (ok, completely arbitrary number). The question- do you think our quality perceptions are highly non-linear, with some small as-yet-undescribed characteristic of the sound completely making or breaking the system, even after 97% (or more) of the technical requirements have been satisfied?
Interesting food for thought. I do think that our quality perceptions are non-linear, but I would contend that they are much like an asymptotic function. Once the basic technical requirements of full audible frequency response and low linear distortion are met, the curve starts tapering off towards the limit of our individual tastes. As people weight the importance of each technical requirement differently, the progression along the curve is going to happen differently for each person. We might find in the end that our limits somewhat converge, as you have mentioned.

For me, proper room setup might get me to the 50% range, low non-linear distortion might get me towards the 60% range, then uniform power response and controlled directivity might get me towards the 70% and 80% range respectively. Improvements to the noise floor and amplifier distortion might then get me into the 90%, but we are looking at a significant amount of effort and cost to make small improvements at this point.

I think one could sum it up quite well by quoting the law of diminishing returns.


  • perceived sound quality.gif
    perceived sound quality.gif
    5.7 KB · Views: 287
I like your graph!
Strongly reminds me an A-B test we did. "A" was an expensive (~2000EUR), so-called Hi-End equipment (factory made). "B" was a dirty cheap Lin-topology 2×~100W quasi-complementary (2N3773 outputs, BD139/140 drivers) amplifier with bootstrapped VAS (BD139) from the "audio stone-age" (late 70's) - DIY-ed by me.
"A" sounded better. I'd say... 75% for "A" and 25% for "B".
But hey! "B" costed only ~100EUR to build!

×3 better sound for ×20 price.
I tend to think that you may have minimized the estimate that it is only 3% we are after. I've heard quite a few systems and for the most part, you're right. They are only slightly different variations on sound/music, hovering around the same percentage of closeness to the real thing. But none of them approaches the sound of the best systems I have heard (which again doesn't approach the real thing by probably the same amouont). Notice I didn;'t say the most expensive systems, but the best sounding ones (to my ear).

I think there is a lot to be said for system synergy and pairing components together that compliment each other. Room acoustics play a large, large role in the overall sound. Silly things which shouldn't make a large difference (like teflon caps and such) on paper, seem to make a very large difference in real life.

There's a lot to be said, and a lot gained, from optomizing a circuit, isolating a ground plane, upgrading internal parts, etc. Tweaking, in a word. So yes, I think it is worthwhile time spent, because I think that the % to be gained is significantly more than you stated.

Yes, there are diminishing returns, but in many commercial designs I think there is certainly room for significant improvement, especially in cost effective designs.

I agree Bob.

A main problem is that many audiophile, or say HiFi listener thinks that a good amplifier or a good speaker is enough for good sound reproduction.
An audio system is a system. System created of amp, spks, room, pre, etc... etc..
Amp and spk are only a few percent (say 50%) of the system.
I mention this because I've found it easy to put together quality components, both commercial and diy, and get very acceptable sound that just doesn't speak to me. It seems like only after fooling with cartridge alignment, turntable mats, and some minor tweaks to the signal chain, do I get a system where the performers emotions seem to come through. The difference in overall sound is near zero, but there's some tiny edge, lack of edge, response difference, or something, that makes it seem like there's a real person at the other end of the wires. I don't question that it's just a matter of hitting the right response, phase, balance, or other technical parameter, but it really seems like it's the last couple percent that makes the difference between ho-hum and wow. (I listen mostly through headphones, so haven't even addressed the room yet.)

If you were to guess, what percentage of folks here, or even audiophiles in general, have a stereo system that "sounds as good" as the one Earl Geddes demoed at RMAF (in case you missed it, $220 worth of Costco signal prior to the soundwave generator into the room)?
What about if that same front end was used with SL's Orions++ (the Pioneer has MCH inputs and could be fed by the ASP, albeit from the source players analog output)? What would that percentage be reduced to? Zero?




diyAudio Member
Joined 2005
My take... and this takes the question in a slightly different direction.

Our ears hear what we train them to pay attention to. My old career in sound engineering (concert, recording, live theatre and film) trained me to pay attention to subtle qualities of multiple physical items simultaneously. It's almost a curse now.

I could tell you with my eyes shut whether the drummer uses nylon tipped sticks, what kind of mic is on the snare, what size of cymbals, whether the kick drum had 2 heads or 1, how old the guitar player's string are, etc, etc. And all at the same time. My job was to focus on 40 things at once, and after 15 years it's pretty much unavoidable.

So while -you- may focus on the last 3 percent of audio minutiae AS A WHOLE, and musical emotion as a whole, I cannot ignore the squeak in the kick pedal, that one bass note that was not sync'ed properly with kick drum, a snare sound that I would have 'done differently', or whatever.

Now that I don't do that work anymore, I'm trying to change mental gears to start listening to "previously mixed" music over which I have no control. I'm having a difficult time relaxing and 'listening to a song' instead of dissecting it into its myriad component pieces and mentally re-assembling it my own way.

If you want an exercise in critical listening, pick 5 of your favourite songs and focus on one instrument common to all the songs, and tune out the other instruments. Determine what the differences are between that instrument's sound in all 5 songs. Try a week of focusing on one instrument, then switch to another, and another.

Guitars usually offer the most sound variety between songs since they have so many control options. Sometimes they use the same drum setup for an entire album, sometimes it's different for every song (a budget issue since studio time is $$). Piano can have a lot of variation.

I guess my main point is that you will determine on your own what it is that draws your attention or satisfies or irritates your ears. It might be adjustable by altering the sound reproduction hardware, the listening room acoustics, or maybe it requires the band to sit back down after a fresh night's sleep to play the song again -- just differently this time.


This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.