|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
70.168.149.115
This link was posted in a comment by zobel on the Innerfidelity site and I found it very interesting, apologies if it's been posted in the Asylum before. Several designers give their takes on how much measurements matter. I like Nelson Pass's response the best.
Regards
13DoW
Follow Ups:
Definately
Quote from a John Curl interview.
Also, we couldn’t use mylar capacitors, which are fairly efficient coupling capacitors. While mylars are fairly efficient from a size and costpoint of view, we realized they have problems with dielectric absorption. I didn’t believe it at first. I was working with Noel Lee and a company called Symmetry. We designed this crossover and I specified these one microfarad Mylar caps. Noel kept saying he could 'hear the caps' and I thought he was crazy. Its performance was better than aluminum or tantalum electrolytics, and I couldn’t measure anything wrong with my Sound Technology distortion analyzer. So what was I to complain about? Finally I stopped measuring and started listening, and I realized that the capacitor did have a fundamental flaw. This is where the ear has it all over test equipment. The test equipment is almost always brought on line to actually measure problems the ear hears. So we’re always working in reverse. If we do hear something and we can’t measure it then we try to find ways to measure what we hear. In the end we invariably find a measurement that matches what the ear hears and it becomes very obvious to everybody.
From pages 15/18 and 16/18
Edits: 01/26/15
Case in point - the Japanese amplifiers of the 70s and 80s that measured vanishingly low Total Harmonic Distortion, you know, on the order of 0.0001%. These amps sounded quite bad relative to many other amps with rather higher THD, notably tube amps with THD circa 0.05%. If you can't go by THD what, pray tell! can you go by? Frequency response? Slew rate? If we all went by measurements we'd all have Kenwoods and Pioneers.
" If we all went by measurements we'd all have Kenwoods and Pioneers. "
Those manufacturers were just such jerks for marketing to irrational customers and giving them exactly what they were looking for.
You can also get air cleaners to put on your 2 liter eco commuter which flow more air than a 747 engine needs during takeoff....
I pretty much stopped looking at specs in the 1980s. I buy audio equipment by how it sounds to my ears, not by how some manufacture says it measures..
.
Of course as always, YMMV.
Edits: 01/27/15
Hi
personally i have come to the conclusion that accuracy is the key word.
And tests can tell a lot of things, even on the all chain (see link)
In the picture for instance a 400Hz sawtooth wave at 44Khz sample rate after passing through speakers and microphone.
Kind regards,
bg
Never confuse what is on the scope with what humans hear...it is not a good correlation for the most part.
Hi as i said accuracy in reproducing a signal is a basic concept in high fidelity
The principles are principles ...
If you send in a wave form and you get a very distorted one at the output is not good.
On principle.
If you send in a monotone signal and you get at the output many other signals not present in the original one again this is not good.
I would start with a test to screen out the not accurate systems, to move on maybe with listening tests.
I cannot say that good sounding systems cannot be also accurate of course. That would be the best.
But a system that distorts badly a signal or even more it generates signals not presented in the original input ... it lets me perplexed. Very.
And i also think that in the long term distortion is fatiguing.
It is like looking at something with wrong glasses. It stresses the brain.
Kind regards,
bg
Edits: 01/26/15 01/26/15
Beppe61 is right on the money. High Fidelity.
You are thinking about this in the same way an engineer would typically look at this: distortion bad so I use my bag of tricks to reduce that distortion so that wave in = wave out. However, you have to realize that ALL amplification devices are inherently NON-linear and this means that they will all make extra harmonics in addition to the main signal and to one degree or another distort the waveform. That is the imperfect world we live in...no such thing as a truly linear amplification device.
Then in the 1930s a guy discovered that if you use negative feedback, suddenly the whole mess seems to disappear and it became standard engineering practice...along with push/pull circuits that partially cancel out even order distortion.
However, it was noticed by some that sound quality did not take a leap forward as someone who is obssessed with HOW MUCH distortion was present would have assumed. In fact, some felt that there was a big problem with the now orthodox view of applying feedback to "linearize" a circuit. No forethought was done to address if what was being done in the name of lower distortion was actually lower sounding to the listener.
This got some people thinking about HOW distortion is reduced and what side effects there are from those approaches and what it might be doing sonically that has a negative impact on perceived quality. It turns out that people are quite tolerant of even a few % 2nd order harmonic but may well be intolerant of < 0.1% ninth harmonic or even less tolerant of even higher orders.
Obviously a truly distortion free signal is desireable but it is not really possible at least not with current technology. What masquerades as almost "zero" distortion has been shown by Crowhurst and others to be nothing of the sort, which along with signal modulated noise floors, dynamic distortion, noise burst etc. seems that the distortion is shifted around but not really eliminated.
Your confusion arises from the thining that what is on the scope is what you hear...it is not really the case because also the scope doesn't often have the resolution necessary to show you all the really tiny distortions that seem innocent but may be very damaging to the sound.
Some distortions are fatiguing and some are simply invisible to your ear brain as they are masked by the ear/brains own mechanisms and distortions. High order harmonics are almost always fatiguing because it takes very high SPLs to generate them in your own ear and they are associated therefore with loudness. I think it is these high order harmonics that are the most damaging to soundstage and full 3d stereo imaging. They bring the perception of highs forward and thus shorten the depth and flatten the images.
Many no feedback amps, whatever their other flaws, have good long term listenability because they are creating a distortion pattern that is closer to nature and often have exceptionally 3d imagining and depth of soundstage because of a relative lack of high order harmonics that damage such perception.
"ALL amplification devices are inherently NON-linear"
Many class D amps measure very nearly perfectly linearly.
Hi and thanks for the very helpful explanation
But i have an opinion on why solid state usually sounds less musical, less "dynamic"
It is not for the use of negative feedback but more for the lack of enough capacitance in the power supply
This can be verified easily in those rare cases when an upgrade of capacitance is provided even from the factory for some SS power amps
Usually the upgraded version sounds more full, less noisy, more dynamic
If you increase the capacitance enough you can make a solid state amp sound like a tube one.Another point.
SE tube amps are usually tested with very efficient and easy loads
I would try a decent solid state amp with the same speakers and listen
I am sure the result would be equally satisfying
Try instead the SE tube amps with a difficult speaker ... and listen
I can agree that efficient speakers are much more amp friendly and maybe a smart choice in the end
Why to complicate life with difficul loads ?
I said about capacitance because i made some experiments that confirm this. The capacitance used in ss amps power supplies is usually too low
Because big and very good caps cost a lot indeed.
A good starting value could be good quality 20.000uF per voltage rail per channel
for a total of 80.000 uF for a stereo amp
Instead i have seen values of even 15.000 uF ... i can understand that then the resulting sound then is flat ...
Caps are like shock absorbers in a car ... if they are undersized and the road is bumpy the travel is not very pleasant.
Kind regards,
bg
Edits: 01/26/15 01/26/15
If you increase the capacitance enough you can make a solid state amp sound like a tube one.
I would not agree with this as I have heard amps with over 1 Farad (i.e. 1,000,000 uF) that still sounded nothing like a tube amp. It might sound more dynamic...or not. Clearly an amp will sound less dynamic with an underdeveloped power supply but it is not that all there is to it.
"SE tube amps are usually tested with very efficient and easy loads
I would try a decent solid state amp with the same speakers and listen
I am sure the result would be equally satisfying"
Have you done this experiment? I assure you I have and you would likely not believe the results if I tell you. I used to run my KR Audio VA350i on very difficult electrostatic speakers and not one SS amp ever sounded more dynamic. I had a similar results on a number of other speakers.
I have only had SS amps with very large power supplies so I don't know what amps you are referring to but I had ones with 100,000 to 250,000uF.
Hi and ok .. i trust you
i do not have a significant experience that tells me the opposite
And tube amps are usually more musical ... i agree.But what about a let's say good Naim integrated (even a small one) with your speakers ? have you tried it ? just out of curiosity ...
By the way i will try some hybrids one day.
I do not like the bass of the average tube amp and i cannot afford big money amps.
But hybrids can indeed get the best of the two worlds for reasonable money.
Kind regards,
bg
Edits: 01/27/15
Things are much much worse than you have described it. Actually, perhaps you've noticed, that almost ALL systems, even ones where the owner might swear up and down that the distortion has ben reduced to near ZERO, distort when the volume is pushed up past a certain point, that point in fact NOT that loud as to be confused with clipping or speaker distortion or any such thing. The distortion actually occurs at a fairly modest volume level. That should not happen. The reason for this distortion at rather modest loudness levels actually has NOTHING WHATSOEVER TO DO WITH THE AUDIO SIGNAL! It has nothing to do with the signal in the electronics, with the AC in the house power or power cords, in cabling or anywhere else in the system. This distortion is produced by information fields that interfere with the listener's hearing ability. The good sound, the real sound his system is producing is in the room, the acoustic waves are very accurate and complete. But because the INFORMATION WAVES are interfering with the listener's brain, confusing and affecting his PERCEPTION OF THE SOUND, what he actually hears is a more compressed, more distorted and noisier version of what's in the room. INFORMATION FIELDS are not electromagnetic in nature, and they do not attenuate over distance, even great distance. They are somewhat like the Higgs field in that respect.Cheers, GK
Edits: 01/26/15
"Things are much much worse than you have described it. Actually, perhaps you've noticed, that almost ALL systems, even ones where the owner might swear up and down that the distortion has ben reduced to near ZERO, distort when the volume is pushed up past a certain point, that point in fact NOT that loud as to be confused with clipping or speaker distortion or any such thing. The distortion actually occurs at a fairly modest volume level. That should not happen. The reason for this distortion at rather modest loudness levels actually has NOTHING WHATSOEVER TO DO WITH THE AUDIO SIGNAL! It has nothing to do with the signal in the electronics, with the AC in the house power or power cords, in cabling or anywhere else in the system. This distortion is produced by information fields that interfere with the listener's hearing ability. The good sound, the real sound his system is producing is in the room, the acoustic waves are very accurate and complete. But because the INFORMATION WAVES are interfering with the listener's brain, confusing and affecting his PERCEPTION OF THE SOUND, what he actually hears is a more compressed, more distorted and noisier version of what's in the room. INFORMATION FIELDS are not electromagnetic in nature, and they do not attenuate over distance, even great distance. They are somewhat like the Higgs field in that respect.
Cheers, GK "
Please define "information fields".
What are Information Fields? They are non electromagnetic fields that result when there are. Great many things of the same type. For example there is an Information Field for atoms and another one for electrons, the fields of which would very strong since the intensity of an Information Field is proportional to the number of things or objects that create the field. And the fields is a natural biproduct of the things or objects. It's when certain things or objects in the immediate surrounding represent a danger in the mind of the listener that his hearing is affected/degraded. This is especially true for such man made things as squares and rectangles but also phrases and even words. Thus simply throwing out old newspapers and magazines in and of itself will usually improve the sound. You would hear the sound as being less distorted and more transparent. Unfortunately there are a great many things in the local area that are linked to information fields of different types, all of which are bad for the sound. Term *Compact Disc* printed on all CDs, CD trays in cases and CDs cases in the world produces a very strong Information Field.
This is especially true for such man made things as squares and rectangles but also phrases and even words. Thus simply throwing out old newspapers and magazines in and of itself will usually improve the sound.About 25 to 30 years ago I was in a Medical records storage room where there was several long row of shelving about 7' or 8' high filled with paper folders of patient’s medical records. Just going from memory each row was about 40' long. If a person stood at one end and another person stood at the other end of a same row even when speaking very loudly neither person could hear the person on the other end of the row. LOL, even hollering at the top of the voice could hardly be heard. All the cardboard and paper was absorbing up the sound waves.
When I see photos of members systems with a whole wall full of vinyl records I wonder how all the record jackets are affecting the sound from the speakers in the room.
Sorry if wandered too much from your post.
Edits: 01/27/15
While I understand what you said what I'm talking about here is much more, uh, mysterious, what I'm referring to has nothing to do with things resonating or absorbing, nothing so mundane. What I am referring to is the effect of shapes and geometry on the listener, not on the audio signal or the acoustic waves. However, having got that off my chest, the albums are bad for the sound, though most audiophiles probably believe the albums are absorbing bad acoustic waves. The same problem with albums and CDs in terms of geometry is also the case for the room itself, you know, being square or rectangular, and I won't even get into the shape of the house itself and all the right angles. You can see how these shapes can produce information fields, no? Since there are SO MANY things in the WORLD that have that particular shape. Another way to look at it, all things being equal the same audio system would sound much better if it were inside a geodesic dome rather than an ordinary room of a house.
"all things being equal the same audio system would sound much better if it were inside a geodesic dome rather than an ordinary room of a house. "
How so?
Uh, because there aren't any 90 degree angles? Spherical things are found in Nature, but not square or rectangular things. They're so unnatural. Furthermore, even if you were blindfolded you would be affected by the change in geometry since it's an example of mind matter interaction, kind of like extra sensory perception.
Edits: 01/27/15
Wow! If tossing newspapers helps imagine how my sound will bloom when I delete your post!
Oh wait... Since the 'information field' doesn't decrease with distance and it still exists in your sent file folder and on Rod's server deleting it on mine won't help one iota! Rats.
What fun... Even audiophiles don't have to be serious ALL the time.
Rick
Fun, right? Next stop Teleportation Tweak and the Photos in the Freezer Tweak. Don't think with the Right Side of your brain so much. Oh, wait, I forgot, you only have a Right Side. Fun, fun, fun. :-}
Edits: 01/27/15
Like the Higgs Field? LOL. No, the Higgs field is something real that real scientists can write down math to define and explain, and it can be detected in expensive machines.
Information Fields as you have just described them are some BS that you made up, and have no other objective existence, let alone are the cause of audio distortion. A better name might be a Disinformation Field.
I know it's obvious but someone should say it.
Ah, now I see where you got the tox in beau tox. You little fukkers down under bite, eh? As for reality I may be slow but I'm ahead of you.
Edits: 01/26/15
Ahead of you? Mate, it's already the next day here.
'Fraid that ain't gonna do you much good, Mate. I'm many years in the future.Remember, it only has to be sufficiently advanced to be indistinguishable from magic.
Cheers
Edits: 01/27/15
"What are Information Fields? They are non electromagnetic fields that result when there are. Great many things of the same type. For example there is an Information Field for atoms and another one for electrons, the fields of which would very strong since the intensity of an Information Field is proportional to the number of things or objects that create the field. And the fields is a natural biproduct of the things or objects. It's when certain things or objects in the immediate surrounding represent a danger in the mind of the listener that his hearing is affected/degraded. This is especially true for such man made things as squares and rectangles but also phrases and even words. Thus simply throwing out old newspapers and magazines in and of itself will usually improve the sound. You would hear the sound as being less distorted and more transparent. Unfortunately there are a great many things in the local area that are linked to information fields of different types, all of which are bad for the sound. Term *Compact Disc* printed on all CDs, CD trays in cases and CDs cases in the world produces a very strong Information Field. "
Ok, whatever.
The young kids these days say, What Ev. And sign W and E. Just a heads up.
lol!
Nt
The ear/brain of listening humans is the only thing that can define whether or not a piece of gear is good sounding or not. Measurements are just numbers that have to be correlated with human perception.
That is why an amp can product 0.0001% THD and still sound sterile and lifeless...it is not the AMOUNT of distortion that is important it is the quality of that distortion and how it impacts the listener. It could also suggest that other distortions that are unpleasant are lurking somewhere that are not covered by this THD measurement.
Speakers have lots of materials related distortion that sits below a perfect looking frequency response curve but for sure has a serious impact on the overall sound quality...driver materials, cabinet materials, bass loading designs etc. all contribute. Crossover parts as well generate different levels and qualities of distortion.
There have been some attempts to correlate what we hear with what we measure and one of the first was by British BBC engineer D.E.L. Shorter when it was noticed that sound quality and the measurements of tube amps were not going hand in hand. At this time the push/pull tube Class AB tube amp with negative feedback had become the standard and sound quality no longer correlated with THD as it is likely to do with no feedback, Class A SETs...although there are lots of reasons those won't correlate so well with a simple measurement either.
Cheever did some nice historical summary in his 2001 master's thesis and also the introduction of a new metric based on how the ear/brain actually perceives (to the best of current understanding) distortion. His metric is actually also SPL related...so the sound level matters too.
Geddes published a couple of good papers on the subject and came up with a metric that had a good correlation (r = 0.9) between listener preferences and different types of distortions that gave a low to high Gedlee number (some of the worst were actually some of the lowest THD).
He even states that they observed NO or even a slightly NEGATIVE correlation between THD and sound quality!! Think about it.
Norman Crowhurst wrote many papers on the subject of negative feedback but the most interesting was one where he pointed out that increasing the negative feedback ends up generating what amounts to a false noise floor that is actually modulated by the signal. This will never show up in a normal measurement but it has the effect of masking low level information that is vital for sonic realism. Why is that different from real noise? Real noise is not correlated with the signal and it has been shown that you can effectively hear music signal correlated sounds BELOW the threshold of the background noise. Now if that "noise" is correlated with the signal then the brain can no longer distinguish between noise and signal because both are correlated and that information gets lost to us.
Finally, Nelson Pass's white paper on what is happening with negative feedback application is interesting because it shows that feedback tends to concentrate distortion rather than really getting rid of it.
So, measurement do matter and SHOULD be used in ways mentioned in the articles above to use what we know about human perception to make better sounding gear. However, this is not being done at all by most designers who have stumbled upon one or maybe two of these ideas but nearly none who are applying these ideas systematically.
Great summary!
Of course a set of links would REALLY make it easy for us lazy folks. But I have read at least many of them at some time or another and likely can find em' again.
I know there are quite a few of us on this forum that used to design electronics for a living or otherwise have an involvement beyond just being a user and we still find home audio fascinating in retirement. For me the discernment of our hearing (especially the WSP) is a big part of the fascination. That's Wet Signal Processing.
I am thus not troubled by our current "Technical Difficulties", indeed they add a great deal of interest. The saving grace is that the system I have seems accurate and satisfying (to me) so there is no stress involved in just playing around since further improvements are likely but not essential. That to me is about as good as it gets, is there any audiophile anywhere that has resolutely arrived at the journey's end?
I've worked on other products that to a large part relied upon an aural interface and you know, we suffered the same ilk of issues. I also once shared a Lab. with two other groups of folks (I was doing Avionics), some were working on video faders and others on TV exciters. Guess what? They had the same problems. The common element: Humans! Wretched humans.
When they wanted an ignorant, unbiased observer they would sometimes turn to me.
"Can you tell a difference between these pictures?"
"Yes".
"Which is better?"
"That one"
"Shit!"
And like our stuff, the two things had "identical" group delay, gain, bandwidth...
Audiophiles are not alone. Oh, another example: "birders" and their binoculars and scopes, "Optics" they call them. Audiophiles are really OK, we just have normal discernment coupled with an especially intense interest.
We CARE about the sound!
Rick
Interesting stuff to be sure. For me it is THE interesting part of the hobby beyond listening to music...that and designing some of my own gear.
However, sound is even more complicated to understand than vision because it is not as centrally processed and hearing relies on phase relationships as well as amplitudes.
They say size doesn't matter and I'm sure they've taken plenty of measurements by now.
YAWN.
Someone recently wrote a letter to Stereophile taking them to task for including measurements in reviews. I didn't agree with that individual at all.
The field is called audio engineering for a reason. Its practitioners are engineers who are people who try to measure and quantify the electrical, mechanical and acoustical properties of their designs. To paraphrase Paul Klipsch, "you can't make what you can't measure because you won't know when you've got it made."
I have read reviews in The Absolute Sound where the writer included frequency response "measurements" of a component under review. I thought that was the publication that didn't believe in measurements!
Specifications can be misleading and marketing and advertising types can get very "creative" when including specifications in copy.
Speakers are especially difficult to nail down--there are so many variables--room size and surface treatments, speaker and listener location and so on. I think John Bowers had as good an approach as possible using good engineering principles then subjecting the speakers to listeners who were musicians who had performed the works they were listening to.
The Aristotelean mean seems to me best--not no measurements, not no auditions, but a judicious balance of each, both contributing to a well designed speaker.
For me, the audition in my listening room ultimately determines whether or not speakers stay there, but I still want to read the specs and measurements.
...no one "doesn't believe in measurements".
I suspect TAS believes observational listening is more important than measurements.
I agree with that from a consumer's perspective.
And they don't want to take the time or invest the money to do them.
John Dunlavy insisted that his speakers have a good step response in support of his design phiosophy of time aligned-in phase speakers. After that was properly extablished he went to cabinet size, shape, bracing, stuffing and alll the rest.
I have heard equipment of two different designs have almost the same specs which sounded very different from each other.
I am a pro violinist. I have only one violin, but no matter where I play it it sounds different. I have played in Carnegie Hall, and in Alice Tulley Hall....not far from each other, yet my instrument sounds miles different from the sound at each venue. ..as a matter of fact, on that same instrument if I choose to play with my alternate bow, the sound is different...or strings, or even rosin used on each of the bows. All of this sound is right....just different. I would encourage designers of equipment to concern themselves with the minutia of measurements, but for the listener....just enjoy.
Measurements matter. But many of today's measurements and/or how we use them often don't correlate with what we hear(of course that assumes a golden ear). Does that mean throw out measurements? No! It means learn to use the ones we have now better and also to find new ones. The sooner we do that the less reliance of the prophet's Golden Ear.
By the way I have seen graphs and measurements work quite well. I've seen a friend pick a cartridge from a frequency curve and a square wave(not knowing what the pickup was when he said it was good). I've seen the same friend design speakers on a computer in half a day and they were finished except for fine tuning that was just fine tuning the final colors of the speaker(since none is uncolored) to fit his own tastes.
Some measurements work for some people who actually know what they're doing. And if we keep working on the problems they will continue to be even better tools. Besides without even our current 'poor' measurements we couldn't even design the products that are then judged by Golden Ears.
...at least two of the designers responding are no longer with us.
Interesting article.
Agree with you about Pass' response.
This from the article I posted earlier, and was hopping more people would be interested:
"The complications and controversy stem from the fact that music is played to be heard by human beings, whose nonlinear ear-brain hearing systems are far from fully understood. Since no one knows exactly how to model the human auditory system, no one knows exactly what engineering measurements are appropriate to evaluating the performance of audio equipment. A smidgen of some kinds of distortion may sound worse to the ear than larger amounts of other kinds. So ultimately, the only way to judge audio equipment is by listening to it. Hence the controversy: subjective human perception—especially when flanked by questions of artistic merit—is made to order for arguments and disputation."
“Somebody was always controlling who got a chance and who didn’t. - Charles Bukowski
... that doesn't mean we can disregard measurements. I am a circuit designer by profession (not audio) and I know when I've done my job because I hit the specs. I suspect the professional audio world is similar but, obviously, not the audiophile world. I find extreme subjectivist and extreme objectivist arguments equally annoying, which is why I liked Nelson Pass's well-balanced approach.
I remember reading that tube article when it came out, nice to see it again. One passage may me smile "Another general tendency has been for tubed amplifiers (of both guitar and high-end types) to be designed and built by technically unqualified people"
Regards
13DoW
You are right...Nelson Pass response was very good.
“Somebody was always controlling who got a chance and who didn’t. - Charles Bukowski
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: