Measuring measuring gear

January 21, 2022
 by Paul McGowan

One of my readers, Jake from Boston, sent me the following note that might someday make it to an Ask Paul video, but for now, it’s certainly worthy of a post.

Hi Paul, I have been doing some reading about measurement tools such as those produced by audio precision, and I was wondering how accurate these tools can be considering they produce their own noise, etc. It doesn’t make sense to measure it because the tool you would use has the same issue. Are these tools good enough to the point where their own inherent problems are truly negligible?

The simple answer to Jake’s question is yes, the measuring equipment we rely upon for our designs is certainly quiet and low distortion enough to measure beyond the point of audibility. And that’s what’s important.

Our mainstay of measuring equipment is, as Jake suggests, made by an Oregon-based company called Audio Precision. (For those who are observers of URLs you might notice that Audio Precision has one of the rarest URLs known, a two letter URL AP.Com).

From their website on the equipment we use: “With a typical residual THD+N of -120 dB and over 1 MHz bandwidth, the analyzer surpasses the analog performance of all other audio analyzers, including a 5 dB improvement compared to our 2700 Series analyzer. Add to this FFTs of 1.2 million points and full 24-bit resolution, and you have performance unmatched by any other instrument.”

So two comments about this. First, from even years ago, the measurement equipment we rely upon to design our products have specs that outperform the human hearing mechanism by a major magnitude. Second, measurements are critical to the design process. From a time standpoint, 80% of all our design efforts are spent at the measurement bench getting the product to do and perform as we wish.

Lastly, as folks who read this blog know, the last 20% is spent with our equally important measurement tool, the listening room.

To make a world class product it takes the best measurement equipment for the job—both on the test bench and in the listening room.

Subscribe to Paul's Posts

40 comments on “Measuring measuring gear”

        1. Aging ears for sure LOL! I was referring to Pareto’s vital few vs trivial many. My data sets have been falling 90/10 in payoff distribution. Apply it as your creativity desires.

  1. Here you are, grown adults, arguing over measuring how much measuring is done.

    The interesting word is “audible”. Years ago, before digital and then before solid state, proper reviewers for magazines like Gramophone had good test equipment to measure things that mattered, like power output, frequency response, distortion etc. Gramophone called them “Technical Reports” before they were called reviews. One measurement for control units (pre-amps) was maximum cable length based on 20pF/ft screened cable. These guys were professional engineers.

    When John Gilbert of Gramophone measured the Quad 303 he had to buy new test equipment, finding that his existing machine had more harmonic distortion than the 303 amplifier he was testing. So he went out and bought a Dynamco Precision Oscillator, providing the specifications in his review.

    There are measurements and measurements, those that manufacturers use as tolerances in the design process and measurements that are actually relevant to the end user. As suggested several times yesterday, it seems the relevance of these is often confused and the design measurements are as much quality control, whereas ultimately the ears have it.

    And as a bald bloke from the Midlands once said:
    If we shadows have offended,
    Think but this, and all is mended—
    That you have but slumbered here
    While these visions did appear.

    1. “We instinctively admire their suave fakery, their artful dodging, their expansive self-congratulatory phraseology , their mellifluous padding. We have been through it, too, and we know good trickery when we see it.”

      From Jean Sheppard’s book “In God We Trust All Others Pay Cash”

      This is an excellent commentary that can be easily applied to specs/reviews found in the popular audio mag, product manuals and the “FAKE” help from call in centers at the manufacturers.

      I get a big “kick” out of the info from the TALK centers I’ve called. I had a problem with the amp and the pre amp from a company. On the advice of the pre and amp hifi manufacturer, I was told to call the manufacturer of the streamer, my internet service, and the streaming service for resolve which was a total waste of time. It took 6 months of calling to the company for them to accept responsibility for a pre-amp and an amp with inherent known, unresolvable, problems. I was then told that the engineers at the company was working on the software to correct the problem which was a smoke screen. However, I was not offered a reimbursement, or credit. What I was offered was a 33% discount on more of their equipment. I would NEVER buy another one of their products.

      1. My interest in reviews is more akin to fiction than non-fiction. I go for the informative writing style, even if dull and boring. When the likes of Mr Fremer try to jazz it up I treat it more like fiction, but if I want fiction I’d rather read a book than an audio rag. Anyone who believes manufacturers’ marketing claims gets what they deserve.

  2. I’m a bit confused… could you spread the design effort percentages between circuit design, measurements of the same and listening test of the same please?

  3. Hopefully someday a measurement device and measurements will finally replicate the human ear / brain interface and then be able to be 100% correlated to each individual’s on the planet hearing acuity. In fact maybe the technology needs to be developed to correct an individuals hearing acuity so they can hear ‘correctly’ according to some pre determined standard.

    Until then, accurate measurements are a necessary part for those in the audio design or modification phase. Some of the consumers only want to read specs, drool over graphs and form their opinion that way. Other consumers want to use just their ears to judge. In the 3rd group of consumers there’s middle ground who utilize both methods. (The 3rd group is probably the majority, but can be held in disdain by the other groups)

    There was an article a little while back in one of the review rags about the way Dan D’Agostino developed gear at Krell versus the way he does it now. A very interesting read…. For all the postulating groups…. If I find it again I’ll post a link to it. Found it. 😀

    Check out the 6th paragraph… assuming I counted correct..

    For now I’m very happy there are manufacturers out there that model, measure and listen. In the end the equipment choices made for myself will be made by me based on the parameters I set.

  4. Hi Paul, a filosophical reflection. The soundstage built up behind the loudspeakers after a perfect set up is a illusion (although very real). The same can be said about aging (classic concerts trained) ears. The brain registered the perfect sound (when younger), but can restore the perfect “music illusion”even through aging ears ( no longer capable of hearing the full scale frequency). So we don’t only hear with our ears (certain bass frequencies can be felt). Greetings from Belgium, Marc

  5. I looked up Audio Precision’s scope of accreditation and it requires three pieces of gear to calibrate their products:

    A Keysight 3458A, an 8.5 digit multimeter, is used to measure the DCV, ACV outputs from the AP unit and to check its input and output resistances.

    A Keysight 53131A, a frequency counter, is used to check the frequency of the outputs from the AP.

    A Fluke 5522A, an electrical calibrator, is used to generate known signals that the AP then measures.

    These are all very high quality reference assets, which is definitely a good sign. And, the fact that AP is accredited is also reassuring.

    I know the guy at Fluke Cal that manages the AP account and he spoke highly of them. If memory serves, I think I heard that some of the guys at AP got their start at Tektronics, the oscilloscope manufacturer also out of Beaverton, OR. This would explain why they know what they are doing when it comes to calibration.

      1. Ha, yes. That was actually a saying there: “If it measures right, it must be a Fluke”. Another one I heard (from the more cynical employees): “You can buy a better product, but you won’t find one more expensive”.

  6. A sort of related tangent: Guitar tuners (measuring instruments) vary from manufacture to manufacturer and even unit to unit within the same manufacturer.
    To really dial in the band everyone should use the same tuner and then adjust by ear. For example the guitar B string needs to be flattened just a bit or you will get dissonance from a perfectly tuned B string in relation to the high e string. Intonation goes out on a lot of stringed instruments (temp/ humidity, etc) so tuning to pitch needs to be taken at several places along the neck and making an overall summation of the desired pitch. The electronic measuring instrument is not the end all to end all. It is just tool in the hands of the craftsman…
    I use a BOSS TU-3 tuner on my pedal board and unlike convention (or maybe herd mentality?) it is the first in line.

  7. At the tender age of 19 after finishing my freshman year I was allowed to become a lab rat in a professors NMR research lab as a summer job. I became exposed to measurement devices like multi-meters, oscilloscopes and spectrum analyzes made by companies named Fluke, Techtronix and some other names that I can no longer remember. For a 19 year old kid it was like being told I was part of the Starship Enterprise crew ( even though Star Trek would not air for another 3 months ). I learned that high quality measurement devices are essential to any science or engineering research lab.

    Something like 50 years later the Head-Fi organization had its national Can Jam in NYC which I went to. To my pleasant surprise AP had a room at the show where they were demonstrating their measurement gear. It is very impressive gear, well made and the AP folks seemed like really smart people.

  8. Every time we go to a doctor and have our vital signs and blood tested we are relying on measurements to indicate our overall health. If the measuring equipment is not in order and properly calibrated, they can give false readings, but usually they are in good working order. We may feel perfectly well, but measurements can tell us we are not normal and may be in danger of failure under certain stresses and conditions. Measurements often do not tell the whole story and do not take the place of the physician’s observations and experience, since every patient is different.

    I would think measurements do the same thing in audio: give an indication of a piece of gear’s overall balance and health, how close it is to optimal and normal, and its risk of failure under stress. Measurements can confirm what is heard with the human ear but cannot take the place of listening.

    1. And even a properly done medical examine is only indicative of that instant in time-space. You really need a large number of data points to get an idea of what is going on for dynamic systems. When I first diagnosed with type 2 diabetes from an A1C test at my annual physical, they had me taking a blood glucose tests twice a day for a couple weeks. As the results were mostly in an ‘acceptable’ range, it was reduced to daily testing for a few more weeks, then once every 3 days and now weekly. It has stayed more or less stable. Due to my background in environmental geology and keeping track of long term ground water monitoring networks, I run some basic statistics on the data for my own purposes and the Doc seems to appreciate it also.

      1. You’re right. Take BP, for example, it varies throughout the day. Just sitting, relaxing and thinking about the waves of an ocean can drop it. When I go to my doctor’s office my BP is always higher than when I am home–the white coat effect. Also, when the person taking my BP is asking me questions while taking the measurement, I tell her to stop talking so I can concentrate on relaxing. When she puts the nitric oxide measuring device on my finger, I take deep breaths to raise the oxygen level in my blood for a better read. Anything to improve my stats 🙂

  9. Wait, what? Test equipment is not perfect? Shocked! Shocked I am.

    Jake from Boston (if you’re reading any of this), if you expect perfection from any human activity, you’ve got the wrong universe. Test equipment is not perfect. Audio equipment is not perfect. The human central nervous system is not perfect. However, it’s what we’ve got. Get over it and move on. It’s no excuse for slacking off and doing things half-a**ed for fun and profit, but sometimes good enough really is good enough.
    Happy listening.

Leave a Reply

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

4865 Sterling Dr.
Boulder, CO 80301

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