Measuring the wrong thing

February 28, 2016
 by Paul McGowan

There’s nothing wrong with using measurements to prove something – as long as you’re measuring the right thing – which is where the subjectivists and the measurementists often get off on the wrong foot.

Take for example, heart attacks. If you suffer cardiac arrest, you’re dead within moments. Right? Stopped heart = stopped life. Turns out that’s not necessarily true. Scandinavian scientists have discovered packing a person’s head in ice after cardiac arrest buys them time, and lots of it. In fact, they may not even perform CPR on them. We die not directly from the stopped heart, but from brain failure. In other words, when the heart stops the brain fails: one is the cause, the other the result.

When your heart stops pumping, blood to the brain stops flowing, calcium floods in, and your brain cells die. Pack it in ice, calcium does not flow, and you can survive without a heartbeat for quite a long time without damage. It’s why people falling into frozen lakes can sometimes be revived after even a few hours. We don’t die from cardiac arrest, we die from the brain not getting blood (because of the heart failure). So measuring a person’s pulse tells us not the whole story, but a clue as to the outcome -which can be changed.

The point of all this is not to carry a bucket of ice if you’re prone to heart attacks. But rather to point out the fallacy of believing a limited set of measurements. Before our recent understanding of brain cell death, we could say with certainty that heart failure = death. Today that same measurement is not accurate. Wrong conclusion based on accurate measurement.

We have many means of measurement: our senses and our machines. Let’s be cautious in our proclamations based on our measurements.

Subscribe to Paul's Posts

11 comments on “Measuring the wrong thing”

  1. ‘The honorary Award for best musician portraying an actor in a lead role’ …

    and the Oscar goes to …. John Denver depicting mild-mannered Jerry Landers in the ground breaking film “Oh God” with a compelling performance by God himself, George Burns.

    Cute, stupid little movie http://www.imdb.com/title/tt0076489/

    And speaking of artists, Lady Gaga owning the National Anthem @ Super Bowl 50 definitely grabbed me, can’t remember the last time experiencing goose bumps. Expecting yet another poignant performance from her this evening.

    Frankly, it was a somewhat limp year for film, mostly computer animated, digitally enhanced bullshit : -(

    1. Yeah, more than a few hours would be pushing it. We are never really sure when the people in that situation succumbed to the cold, after all. It is possible to keep people in a state of ” suspended animation” under controlled conditions in circulatory arrest for a few hours, though. “You’re not dead until you are warm and dead.”

  2. One reason this is all so much fun for me is because it is all so hopelessly mixed up. It’s like watching a large number of cats chasing each other’s tails in one endless fight that can get pretty nasty at times. At least that doesn’t happen too often around here.

    Leaf through a bunch of JA’s amplifier measurement reports. Notice anything? For the most part they all look pretty much the same. The differences are usually negligible. Yet read the reviews of how they sound and you will hear this one has better bass, that one has better imaging, the other one has airier highs, and yet another has more presence. Look at the speaker reports and on some of them you will see FR curves of two speakers superimposed on each other showing that they are very similar. Yet you will also read how different their sound is. What’s wrong? How is it explained? Between the so called subjectivists who go by what they hear and so called objectivists who go by what they measure….both are wrong. But IMO the real burden of proof is on the objectivists. They’ve failed leaving the arguments to the subjectivists.

    The first thing an objectivist has to do is prove or disprove that two components really sound different. That’s not as easy as it seems. Fair tests are hard to devise, set up, conduct, and evaluate. And they only give conclusions for the specific conditions of the test. Change anything and the results may no longer be valid. Not only that but were completely descriptive tests performed, the data would fill a book and only engineers would be able to understand them. The search for one or a few simple numbers we’d like to have to explain everything to those not expert may not be possible.

    We rely on scientists to discover and explain phenomena and devise ways to measure them. But methods useful in the 1930s are of little value today. By those standards modern equipment no longer reveals critical differences between competing products as they once did. What are the sciences that need further study? Sound, electronics measurement, acoustics, acoustic measurement, and psychoacoustics. There are important shortcomings in each of them, any one of which alone would invalidate any conclusions drawn from measurements of the others. For example, amplifiers are tested on a bench with a 4 or 8 ohm resistor for a load. Amplifiers that perform nearly identically under those conditions, among the easiest challenges they can be put to, can perform very differently when connected to loudspeakers which is how they are used. So the measurements are of no value because they can’t predict what will happen in real world conditions of use.

    Another mistake is that we don’t listen to amplifiers, speakers, CD players alone, we listen to entire systems of recording and playback from the point where the sound reaches the microphone to the point where it reaches your ears. An amplifier with a rolled off high end and poor damping characteristic of many tube amplifiers may mitigate the characteristics of a loudspeaker that is otherwise shrill and lacks sufficient bass in ways amplifiers that measure better on a test bench can’t. So for that speaker, which is the better amplifier. We also don’t have a specific agreed to goal or set of goals for home hi fi audio systems to perform to. For example, Dr. Mark Waldrep is not interested in accurate reproduction of what is heard live, his goal is to produce the most pleasing “manufactured” music. Others want accuracy recreating what happened live. For many manufactured music recordings, there is no such thing, parts being added at different times.

    Bottom line, the state of the art of measurement based on scientific understanding in this field remains inadequate, in some ways even primitive at a time when other disciplines have made remarkable strides. Rather disappointing but this is why this remains more art than science. Is there any hope things will get better? Maybe one day but I don’t see anything on the horizon.

    1. The basic misunderstanding of Hifi is that the designers of HIFI equipment see themselves as artists – as the performer of music – producing a good sound (!) and not as engineers committed to highest fidelity. The comfort zone for artists is much more enjoyable than the science and standards based zone of an engineer. Once you have reached the status of an artist you get much better payed than an engineer. 🙂
      And if artist-reviewers would be interested in objective evaluations they would test every piece of equipment under identical (!) conditions in a reference listening room.

      1. I don’t know if there are more starving artists than there are starving engineers but I think the artists have it only because there are more of them.

        Engineers are taught to think in terms of systems. They not only look at the trees, they look at the forest. They also look at how each tree works as part of the forest. Focusing on just one tree or a few doesn’t show the big picture. For that you have to zoom out. The first thing you have to do is figure out what the system has to do in specific details, not in vague generalities. Then you have to invent a way to measure whether or not the system does what it is intended to do and if it doesn’t, where and how does it fail and to what degree. That is what this industry fails to do, it has no actual goal let alone ways to measure if it has achieved that goal. It’s paradigm was inherited from prior generations going back many decades. The problem is that the hardware has reached the limits of performance of every variant within that paradigm yet it falls short of expectations. This is why no matter what they do to it, they’ll never be happy. It isn’t the hardware that is flawed, it’s the paradigm itself. It does not successfully address the goal because it doesn’t have a goal. Therefore there is no way to measure it.

        There was a time when another paradigm was tried, quadraphonic sound. It was a technical failure because it was based on a lack of understanding of the items I listed above. So the industry reverted back to its previous paradigm and has ceaselessly tried to perfect every last element of it by trying every possible variant of each element unwilling to accept that there’s no way to go beyond what it already has with this paradigm. Its possibilities have been exhausted and each new variant claiming improvement costs staggering prices.

        There are a few intrepid people who are experimenting with other paradigms that are entirely different. They look at the forest and see something different than just a bunch of individual trees. Some actually have specific goals based on their analysis and understanding of what they want to achieve. Acuvox is one of them. I’m another. Those experimenting with WFS and phase cancellation methods like Ralph Glasgal and Edgar Choueiri are others. Whether these efforts succeed of fail they will each produce different and interesting results that probably can’t be obtained in other ways. Frankly its a lot more fun than trying to build yet another amplifier or speaker.

        1. Having a collection of some 70 SQ Quadraphonic LPs, and both a Sony and Fosgate decoder, I got the idea that by using Ambiophonic reproduction methods, these recordings might now be able to produce a stable 3D sound field. This has turned out to be the case. The separation of the four signals is marginal so you must reproduce the front and rear pairs without losing or distorting any level or time differences. I believe, among other problems, it was the complete loss of the encoded interchannel time differences, when these discs were listened to over speakers arranged in a square, that made Quad sound less than a success.

          The logic circuits used in the decoders only worked in the level domain so again time localization cues were not involved. In my experiments, I can have a full circle of sound without logic. The amplitude cues are also delivered intact by using RACE and a better pinna functionality is provided by having speakers close together in front and rear.

          Unless you are a real vinyl fanatic, it makes no sense to do this, since any SACD multichannel recording or the equivalent in DTS encoded pop, opera, or movie BD is a lot easier to work with. But the exercise does prove that eliminating localization cue distortion does wonders for both surround and ordinary LPs.

        2. [Soundmind]
          I got the chance to go to the last rehearsal of Jacques Offenbach’s “The Tales of Hoffmann” at our local theater.
          There is no doubt, that it will again take some weeks until I will hear an opera on my home-system again. To take your words: In the theater i’ve been aware of the forest, at home there is only a bunch of trees, but I’ve to complement that on other systems a lot of trees were missing 😉
          I’d like to say that most of todays amps will be able to reproduce a soundfield like in a live performance but I doubt that todays loudspeakers – no matter what kind they are – and the recording technique (or must I say the recording engineers) meet the required standards.
          Greetings

    2. What has changed from the 1930s is human hearing. By then the urban and suburban soundscapes were taken over by post-industrial noise pollution – the sounds of pavement, metal, glass and china; engines in transportation and lawn care; motors in industry and household appliances, all of which changed hearing for the worse. Also phonographs and radios became ubiquitous in industrialized societies within a decade of the Kellogg & Rice patent for the loudspeaker (1925).

      Our genes do not contain a blueprint for a completed human neural network, only the feedback mechanisms for growth in response to stimulus. How your brain grows to perceive the world depends on the stimuli while the brain is growing. My research indicates that individuals raised in a pre-industrial environment beyond roads have vastly superior hearing in terms of spatial cognition of the soundfield. This increase in aural information processing is also present in those who practice an acoustic instrument for some hours a day from childhood and avoid listening to bad recordings and audio gear – which is pretty much all of it.

      This is confirmed by hearing measurements and psycho-acoustic observations from pre-loudspeaker era.

  3. “…we could say with certainty that heart failure = death. Today that same measurement is not accurate. Wrong conclusion based on accurate measurement”
    Strange sentence, in fact a kind of “contradictio in terminis”.
    First you say the measurement is not accurate, then you say the measurement is accurate, but the conclusion is wrong…
    I agree with the latter (measurement accurate, conclusion wrong).
    IMHO the content of this post is a bit inconsistent, a bit like comparing apples and oranges.

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301
1-800-PSAUDIO

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram