Subjective vs Objective? Audible differences in hifi

Chat room for MikeC's Audio Craft, music & sound reproduction, streaming, audio PC, etc.
Post Reply
HammerSandwich
*Lifetime Patron*
Posts: 1288
Joined: Sat Oct 25, 2003 3:21 pm
Location: 15143, USA
Contact:

Subjective vs Objective? Audible differences in hifi

Post by HammerSandwich » Tue Dec 08, 2015 4:42 pm

Let me begin by agreeing with MikeC: good amps exhibit only minor audible differences. This is not a magic technology.
MikeC wrote:I would never use the ATI amps. Too big, too ugly, too inefficient.
ATI makes truly excellent amps IMO (the relay-free protection system is tops), but these are fair comments. A few years ago, ATI's factory outlet had significantly lower prices & was a good choice for people who value modular build & 7-year warranties from a USA factory.
MikeC wrote:I'm working on putting together an 8-ch am using 8 class-D Hypex and Connex amp modules with SMPS PSUs. They all have the same voltage gain, which is nice. Roughly 70W each for the tweets and high mids, 250W for the lower mids and 400W for the bass. So about 800W per side, 1600W total. The whole thing, with power supply modules, will cost under $1000, and fit in a dual side heatsink case less than 4" tall, 17" x 15". Efficiency is over 90%.
THAT's the project I'm interested in! Strikes me as more power than I'd ever need, but headroom for EQ makes sense with active crossovers, etc.

Speaking of Hypex, those new Kii speakers sure look interesting.

LostHighway
Posts: 38
Joined: Sat Oct 24, 2015 2:32 pm

Re: MikeC's Audio Craft: An Introduction

Post by LostHighway » Tue Dec 08, 2015 6:22 pm

HammerSandwich wrote:Let me begin by agreeing with MikeC: good amps exhibit only minor audible differences. This is not a magic technology.
Not to get too Bill Clinton but that all depends on what you mean by "good amps" and "minor". I am probably close to a decade out of date on amp technology but my own experience doesn't bear that out. Certainly the differences are not nearly as great as those among speaker systems (or rooms) or turntable/tonearm/cartridge systems but I wouldn't call them minor. If you use a good speaker system within it's comfort zone (relatively flat impedance above say 6 ohms) you can level match a pair of Atma-sphere MA-1s to compare to McIntosh amps (be they 275s or some of their SS amps) and I don't find the difference subtle. Alternately compare an early Bryston 3B or 4B (with the caps in good shape) to a modern 4B-SST and again the differences are not subtle. I really can't speak to the degree of difference among Class-D amps as the only ones I've spent significant time with are the circa 2000 - 2002 Bel Cantos (and to a lesser extent the CIAs from the same period).

MikeC
Site Admin
Posts: 12283
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: MikeC's Audio Craft: An Introduction

Post by MikeC » Tue Dec 08, 2015 9:32 pm

LostHighway wrote:
HammerSandwich wrote:Let me begin by agreeing with MikeC: good amps exhibit only minor audible differences. This is not a magic technology.
Not to get too Bill Clinton but that all depends on what you mean by "good amps" and "minor". I am probably close to a decade out of date on amp technology but my own experience doesn't bear that out. Certainly the differences are not nearly as great as those among speaker systems (or rooms) or turntable/tonearm/cartridge systems but I wouldn't call them minor. If you use a good speaker system within it's comfort zone (relatively flat impedance above say 6 ohms) you can level match a pair of Atma-sphere MA-1s to compare to McIntosh amps (be they 275s or some of their SS amps) and I don't find the difference subtle. Alternately compare an early Bryston 3B or 4B (with the caps in good shape) to a modern 4B-SST and again the differences are not subtle. I really can't speak to the degree of difference among Class-D amps as the only ones I've spent significant time with are the circa 2000 - 2002 Bel Cantos (and to a lesser extent the CIAs from the same period).
OK, it's true that I have had occasion to hear significant differences between "good amps", but in retrospect, I can't be certain what external factors were involved. It was, as you admit, never as big as the difference between source components or speakers. And while I might have had a preference for one or another, this does not confirm superiority, merely my preference under those conditions. But with regard to active xover speakers, I stand by assertion that amp difference are trivial -- again, assuming a certain minimal level of fidelity.
Mike Chin, SPCR Editor/Publisher
Support SPCR by buying your gear through this link: Amazon

LostHighway
Posts: 38
Joined: Sat Oct 24, 2015 2:32 pm

Re: MikeC's Audio Craft: An Introduction

Post by LostHighway » Wed Dec 09, 2015 2:56 am

MikeC wrote:OK, it's true that I have had occasion to hear significant differences between "good amps", but in retrospect, I can't be certain what external factors were involved. It was, as you admit, never as big as the difference between source components or speakers. And while I might have had a preference for one or another, this does not confirm superiority, merely my preference under those conditions. But with regard to active xover speakers, I stand by assertion that amp difference are trivial -- again, assuming a certain minimal level of fidelity.
I have vastly more experience to draw on in single amp per channel passive x-over contexts but I was fairly confident I was hearing differences in bi-amp set-ups using a good analogue x-over (probably most commonly, but not exclusively, the Bryston). I have never, to my knowledge, heard a system with digital x-overs. The problem in these contexts is that you've introduced yet another variable and it becomes very difficult to sort out what is going on, e.g. are you not hearing a difference because there is no difference or because the x-over is introducing a masking or homogenizing effect?

My hierarchy of differences:
Speakers (and rooms): I believe most people can almost always clearly hear the differences although they may not be able to articulate them very well.

Vinyl reproduction chains (turntable cartridge, tonearm and you could probably include phono stages too): Usually clear differences between chains but differences between individual parts of the chains are difficult to pin down because of the difficulties in setting up controlled comparisons. I believe with a fairly high confidence level that the sonic differences among cartridges are significant. I also believe that there are fairly audible differences among tonearms and among turntables but my personal confidence interval is lower. Phono stages are easier to compare and again I'd say that there are significant differences.

Digital sources: I have heard some comparisons between comparable (price and period) CD players where I thought the differences were dramatic but not in all cases, subtle to almost non-existent in some comparisons. The differences between early '80s "good" CD players and late '90s early/2000s "good" CD players are massive IME. DACs (above the dross level) IME the differences usually tend to be subtle but not non-existant. Of course if you go from say a $100 DAC to a $5k (or even a $1.5k) DAC the differences can be quite noticeable.
Amps and linestage preamps: Again, dramatic, or at least very audible, differences in some cases and subtle to non-existent in others.

If there are a takeaways from this IME (I'm not touting these as universal truths) they are:
1 Execution usually trumps theory
2 There is almost always an exception out there to try the rule

I'm a subjectivist who believes in science. There are quite a number of people out there purporting to represent the objectivisit or "scientific" perspective whose grasp of statistics and how science actually works seems tenuous to me at best.

None of this is directed at you Mike and I don't think we really disagree all that much.

MikeC
Site Admin
Posts: 12283
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: MikeC's Audio Craft: An Introduction

Post by MikeC » Wed Dec 09, 2015 8:19 am

LostHighway wrote:None of this is directed at you Mike and I don't think we really disagree all that much.
:lol: :lol:

I used to demonstrate the superiority of Naim amps by swapping in another amp in a Linn/Naim system, which usually lead to a drop in subjective sound quality most people could hear. But here's where the external conditions were unfairly biased for the Naim amp -- the entire system was a carefully matched ecology, so to speak, including the speaker cable, interconnects, etc. Even the midbass bloom of the LP12 (in evidence then, perhaps not now with changes in that product) was somehow ameliorated by this system. So most amps simply dropped into the system upset the ecology and didn't sound as good. I was vaguely aware of the bias of such demos, but there are always limits to how neutral you can get, how thoroughly you can explore differences between two similarly high quality products in a store. Oh, and I slept fine at night, too. :lol:

I had trepidation about the OpenDRC-DA8 taking over both the DAC as well as the xover functions, but I've lived with it for a month now and have no desire to go back to the analog xover or the Teac DAC. My wife thinks the system sounds as good as ever; I think it sounds better. The positive thing about this setup, imo, is that while all the xover filtering & eq & driver time delay are now done in the digital domain, before the signal is converted to analog, and there is still just one conversion from digital to analog.

Of course, when I play LPs, now, I run the analog preamp's output into an ADC (a Tascam UH7000), feed it via USB to the PC, which then processes it like any other digital source. That might add or detract something from the vinyl playback, but I haven't done enough listening yet to be sure of anything. Suffice it to say, the records still sound great... tho some would point to this as evidence of my tin ear. :lol:
Mike Chin, SPCR Editor/Publisher
Support SPCR by buying your gear through this link: Amazon

Quinnbeast
Posts: 94
Joined: Mon Apr 30, 2007 1:55 pm

Re: MikeC's Audio Craft: An Introduction

Post by Quinnbeast » Wed Dec 09, 2015 2:15 pm

When I see discussions of high end audio equipment, I can't help but be reminded of the commentary from Ethan Winer and Co with regards to subjective A/B test of related equipment, and the information bias of human perception when it comes to audio 'preferences'. The part at the beginning about the rigged switch that supposedly changes the amplification of a test setup between a tube and transistor amp really makes me smile.

https://www.youtube.com/watch?v=BYTlN6w ... e=youtu.be
Workhorse: i5 3570K, Asus Z77V-LX, Be Quiet Dark Power 550W, 8GB Corsair XMS, EVGA 970 SSC, Sammy 840 Pro 256GB, 1.5TB WD Green, Fractal R4, Noctua NHU12P. Dell 2312U, Dali Zensor 1 + Audioengine N22.

Movie/FLAC/Misc Box: E8500, P5QL-CM, 4GB Corsair XMS, 4850 512mb Vapor-X, Corsair 450VX, 64GB SanDisk Ultra II + 1.5 GB WD Green, Zalman 7000AlCu w/mod.

HammerSandwich
*Lifetime Patron*
Posts: 1288
Joined: Sat Oct 25, 2003 3:21 pm
Location: 15143, USA
Contact:

Subjective vs Objective? Audible, significant differences

Post by HammerSandwich » Wed Dec 09, 2015 7:04 pm

Quinnbeast wrote:When I see discussions of high end audio equipment, I can't help but be reminded of the commentary from Ethan Winer and Co with regards to subjective A/B test of related equipment, and the information bias of human perception when it comes to audio 'preferences'. The part at the beginning about the rigged switch that supposedly changes the amplification of a test setup between a tube and transistor amp really makes me smile.

https://www.youtube.com/watch?v=BYTlN6w ... e=youtu.be
You can find a LOT of similar tests out there.

This kind of result has made me more objectivist - "pragmatic" might be better - over the last few years. IMO, any difference that's too subtle to detect blind or that requires weeks of analytical listening simply isn't worth much worry/effort/expense, even assuming it is real. Much better to put that time & money into more music.

LostHighway
Posts: 38
Joined: Sat Oct 24, 2015 2:32 pm

Re: MikeC's Audio Craft: An Introduction

Post by LostHighway » Thu Dec 10, 2015 2:55 am

HammerSandwich wrote:
Quinnbeast wrote:When I see discussions of high end audio equipment, I can't help but be reminded of the commentary from Ethan Winer and Co with regards to subjective A/B test of related equipment, and the information bias of human perception when it comes to audio 'preferences'. The part at the beginning about the rigged switch that supposedly changes the amplification of a test setup between a tube and transistor amp really makes me smile.

https://www.youtube.com/watch?v=BYTlN6w ... e=youtu.be
You can find a LOT of similar tests out there.

This kind of result has made me more objectivist - "pragmatic" might be better - over the last few years. IMO, any difference that's too subtle to detect blind or that requires weeks of analytical listening simply isn't worth much worry/effort/expense, even assuming it is real. Much better to put that time & money into more music.
I totally agree that people are far more suggestible than they would like to believe.

There are a number of documented examples of fairly well run double-blind ABX listening tests where the vast majority of listeners ("professionals" included) couldn't do better than chance, however, a number of those tests also turned up a few individuals who do far better than chance with repeatability. The problem with ABX tests is that good ones are hard to run, you need more time than most people are willing to commit (i.e. we're talking multiple days not just an hour or two) and a reasonably large number of participants plus the room(s) and equipment. Since there is close to zero funding for that sort of thing there are far more ridiculously flawed anecdotal "tests" out there, like the one cited in the video, than good science. A visit to Hydrogen Audio, home of the crank "objectivists", will turn up dozens of these meaningless "proofs" that everything sounds the same. And "yes" there are many people I consider crank subjectivists too.

The owner of one of the audio shops I worked for had gone totally blind in his early 20s. When things were slow we (the employees) would often go into one of the rooms and change the systems in some way, usually just subbing a different amp or preamp for the one typically in that system. We'd go get the owner who was in a different room, and most commonly working with a customer or on the phone during the switch, and bring him in and simply ask "what do think?" - sometimes he'd simply say "I think it sounds better (or worse) in this regard..." but with surprising frequency he'd come in and say "Oh, you took out amp X and put in amp Y" which was correct. He was not only recognizing a difference but identifying exactly what the difference was in a shop with a wide variety of gear to select among. Also, we occasionally tried to suggest a change when there in truth was none and we'd most commonly get either "I'm not sure what you did" or "I'm not hearing a difference". I realize that this too falls into the category of ridiculously flawed anecdotal evidence so make of it what you will.

I agree with much that Ethan Winer has to say but when he says, "We know how to measure everything that matters" he loses me. That is a fairly classic example of both hubris and how not to do science.
Last edited by LostHighway on Thu Dec 10, 2015 1:44 pm, edited 1 time in total.

Quinnbeast
Posts: 94
Joined: Mon Apr 30, 2007 1:55 pm

Re: MikeC's Audio Craft: An Introduction

Post by Quinnbeast » Thu Dec 10, 2015 3:45 pm

LostHighway wrote: I agree with much that Ethan Winer has to say but when he says, "We know how to measure everything that matters" he loses me. That is a fairly classic example of both hubris and how not to do science.
LostHighway wrote:I'm a subjectivist who believes in science. There are quite a number of people out there purporting to represent the objectivisit or "scientific" perspective whose grasp of statistics and how science actually works seems tenuous to me at best.
Apologies if this is taking things off-topic, but I am a little curious (and enjoying the discussion). You make a similar reference a couple of times with regards to what you consider to be an incorrect use (or understanding) of science in certain circles, but I don't get your meaning I'm afraid. While I can appreciate that - for example - speaker / equipment design is very much a overlap between scientific principles and a personal or artistic choice of how to construct gear correctly, I'm not sure what this immeasurable element is. If something cannot be measured, how do we then perceive it? How does the immeasurable come in to play?
Workhorse: i5 3570K, Asus Z77V-LX, Be Quiet Dark Power 550W, 8GB Corsair XMS, EVGA 970 SSC, Sammy 840 Pro 256GB, 1.5TB WD Green, Fractal R4, Noctua NHU12P. Dell 2312U, Dali Zensor 1 + Audioengine N22.

Movie/FLAC/Misc Box: E8500, P5QL-CM, 4GB Corsair XMS, 4850 512mb Vapor-X, Corsair 450VX, 64GB SanDisk Ultra II + 1.5 GB WD Green, Zalman 7000AlCu w/mod.

LostHighway
Posts: 38
Joined: Sat Oct 24, 2015 2:32 pm

Re: MikeC's Audio Craft: An Introduction

Post by LostHighway » Thu Dec 10, 2015 4:51 pm

Quinnbeast wrote:Apologies if this is taking things off-topic, but I am a little curious (and enjoying the discussion). You make a similar reference a couple of times with regards to what you consider to be an incorrect use (or understanding) of science in certain circles, but I don't get your meaning I'm afraid. While I can appreciate that - for example - speaker / equipment design is very much a overlap between scientific principles and a personal or artistic choice of how to construct gear correctly, I'm not sure what this immeasurable element is. If something cannot be measured, how do we then perceive it? How does the immeasurable come in to play?
It doesn't bother me if it doesn't bother the mods, although we have already gone a fair way afield from discussion of Mike's excellent DIY audio projects.

I'm probably confusing the issue by using references to bad science in two different ways. In the case of blind testing, which I'm not opposed to at all (it is educational to find how not-so-immediately-obvious differences can be in blind comparisons as opposed to sighted), my take on a proper ABX test methodolgy involves a switch mechanism that is for all practical purposes a (short) straight wire both measurably and in listening tests, double blind, a system and room familiar to the listeners, utilizing both rapid and long interval switching (more of the latter or listener control), listeners get to choose at least half the source music selections, breaks when anyone reports fatigue and a large listener pool or many trials to get you somewhere in the direction of statistical significance.

The second reference to good science versus bad science is one of approach. If you start by thinking, "We've got this dialed, we know exactly how it works and we can measure everything that matters" you've either picked an incredibly simple and boring subject or you've begun by shutting the door to exploration. If you assume you already know the answer you generally get the answer you assume. Richard Feynman (and I'm sure many others) wrote about this. To me the interesting questions are more along the lines of, "What if what we think we know is wrong? Is there an alternate hypothesis or methodology to get at this?" I'm probably not articulating this very well. To cite another example there were late 19th C scientists, doubtless smart men, who thought that post-Maxwell we very close to a complete understanding of all physical laws. The guys who believed that were not the ones that advanced science.

It isn't about immeasurable, although in some cases it may be about immeasurable using present technology, it is about the questions being asked and openness to being wrong and/or to challenging accepted "truth".

Cistron
Posts: 617
Joined: Fri Mar 14, 2008 5:18 am
Location: London, UK

Re: MikeC's Audio Craft: An Introduction

Post by Cistron » Fri Dec 11, 2015 12:27 am

This is awesome. Also loved the presentation on audio myths.
The second reference to good science versus bad science is one of approach. If you start by thinking, "We've got this dialed, we know exactly how it works and we can measure everything that matters" you've either picked an incredibly simple and boring subject or you've begun by shutting the door to exploration.
I don't think I can follow LostHighway's argument. What is way more frequent in science, is measurable effects with immeasurable causes. I'm, roughly speaking, a molecular biologist and there are countless such problems. We understand it, but elucidation of the conundrum with current technology is bloody damn hard. Think protein folding with its massive computational space and difficult to abstract the physics or gene regulation, which is just a mess of thousands of factors working in concert, feedback loops, temporal, concentration changes, seasoned with DNA structure and epigenetics. (Now this is truly OT :mrgreen: )

When I think about audio and acoustics, the physical realm is pretty well explored (isn't it?). We even have a great understanding of how the cochlea works; after all we can explain Tinnitus and even fabricate prosthetics to make the deaf hear again. However, perception, as Ethan Winer explained at the beginning of his presentation, is a totally different issue. We haven't the faintest how all our brain function amount to a consciousness. Neuroscientists can tell you where in the brain certain signals are processed. That's about it.

Many psychologists will tell you that our brain produces models of how it understands the world while we grow up. At the centre of this sits the "I", and we spin a hero-story as we venture through life. If the brain encounters something that doesn't fit the model, figuratively speaking, all hell breaks lose. We confabulate: tint, make-up, remember incorrectly, ... and possibly (if we are lucky) slowly and painstakingly adjust our models.

By now I've lost trail on where I was going to go with my argument. So take it as a bit of smart-arsery ;)

LostHighway
Posts: 38
Joined: Sat Oct 24, 2015 2:32 pm

Re: MikeC's Audio Craft: An Introduction

Post by LostHighway » Fri Dec 11, 2015 3:28 am

Cistron wrote:This is awesome. Also loved the presentation on audio myths.
The second reference to good science versus bad science is one of approach. If you start by thinking, "We've got this dialed, we know exactly how it works and we can measure everything that matters" you've either picked an incredibly simple and boring subject or you've begun by shutting the door to exploration.
I don't think I can follow LostHighway's argument. What is way more frequent in science, is measurable effects with immeasurable causes. I'm, roughly speaking, a molecular biologist and there are countless such problems. We understand it, but elucidation of the conundrum with current technology is bloody damn hard. Think protein folding with its massive computational space and difficult to abstract the physics or gene regulation, which is just a mess of thousands of factors working in concert, feedback loops, temporal, concentration changes, seasoned with DNA structure and epigenetics. (Now this is truly OT :mrgreen: )

When I think about audio and acoustics, the physical realm is pretty well explored (isn't it?). We even have a great understanding of how the cochlea works; after all we can explain Tinnitus and even fabricate prosthetics to make the deaf hear again. However, perception, as Ethan Winer explained at the beginning of his presentation, is a totally different issue. We haven't the faintest how all our brain function amount to a consciousness. Neuroscientists can tell you where in the brain certain signals are processed. That's about it.

Many psychologists will tell you that our brain produces models of how it understands the world while we grow up. At the centre of this sits the "I", and we spin a hero-story as we venture through life. If the brain encounters something that doesn't fit the model, figuratively speaking, all hell breaks lose. We confabulate: tint, make-up, remember incorrectly, ... and possibly (if we are lucky) slowly and painstakingly adjust our models.

By now I've lost trail on where I was going to go with my argument. So take it as a bit of smart-arsery ;)
Crap! I just lost a long reply by being timed out on my log in. I'm out of time for now but I'll try to get back to it. Short version: I really liked your post and replied referencing Luca Turin's attempts to revive the vibrational theory of scent (possibly put to bed by Eric Block, et al this year) and Alfred Wegener's proposal of plate tectonics for a couple of contrasting examples. I admire Turin even though he may be both wrong and somewhat prickly.

MikeC
Site Admin
Posts: 12283
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Re: Subjective vs Objective? Audible, significant difference

Post by MikeC » Sat Dec 12, 2015 3:48 pm

Quinnbeast wrote:When I see discussions of high end audio equipment, I can't help but be reminded of the commentary from Ethan Winer and Co with regards to subjective A/B test of related equipment, and the information bias of human perception when it comes to audio 'preferences'. The part at the beginning about the rigged switch that supposedly changes the amplification of a test setup between a tube and transistor amp really makes me smile.

https://www.youtube.com/watch?v=BYTlN6w ... e=youtu.be
Thanks very much for this link. I just viewed/listened to the whole piece & have to say it pretty much reflects my own thinking and perceptions in hifi.

The NULL demos are killer! Ditto the audibility of distortion/noise tests. For me, on my main PC, which doesn't have very good speakers, with sound level at moderate, once the noise/distortion got down to -70 dB, it became inaudible. That's pretty telling about system noise in general.
Mike Chin, SPCR Editor/Publisher
Support SPCR by buying your gear through this link: Amazon

Post Reply