Monday, October 20, 2014

Failing Audiophile 103

Let's talk metrics. The audiophile community loves talking metrics. Since I can't match the technobabble of a true audiophile, I will retaliate with rambling.

Most metrics are meaningless. They mean nothing. And assembled into lists, metrics have even less meaning. It's not because the metrics are wrong. The metrics are all correct. The reason that metrics are meaningless is because you and I, the average consumer, don't really understand what the metrics mean or how they go together. If a consumer cannot make a meaningful decision with a given set of metrics, those metrics are meaningless.

What do we use instead? $$$. We look at the price of a product as a metric for all those metrics. That ought to work pretty well, but you can see where the mischief starts, can't you? If a manufacturer wants a better performing product, it just give the product new packaging and price it higher. "Now with 30% more statistics!" Now worse products are selling next to better products at the same price range.

How can we tell the difference between the two? Metrics! They're both useless and useful at the same time. Talk about a Schroedenger's cat problem.

The thing about metrics is that metrics are all that we really have to distinguish quality from garbage. Metrics are the only way to guarantee that we are getting the quality that we've paid for. Metrics are also the way that we can determine the fair price of a product in the marketplace. A sales force can whip up lies about "experience" and "impression," because those features are subjective, but they can't lie about measurable features. Lying about your product is false advertising.

An arena where this phenomena plays out is in cables. One group claims that their cables "sound better" (which is subjective), and charge significant bucks for the cables. Another group finds that the cables carry a signal no better than an ordinary cable. Who do you believe? Metrics are our only arbiter.

Claims are easy to make. Measuring is hard. It's no wonder that claims hold such power in the marketplace.

Some claims fall apart more easily than others. I've often heard how amps can sound better by removing their cheap Chinese power supplies sound significantly better with replacement power supplies. That sounds good at first, but once you start thinking about it, the argument falls apart. In the brutally competitive audio marketplace, I'm expected to think that sound engineers somehow don't know that the power supply influences sound. I find that that claim dubious. If an engineer can easily get better audiophile sound with a better power supply, then they would build their equipment using a better power supply. They would talk to manufacturers to improve their cheap power supplies. Nobody's going to screw over 95% of their audio design work by using cheap parts (unless it's the lowest end consumer products, which the audiophile won't use anyway).

How do the engineers know when they've done a good job? Metrics. They know and understand the metrics. They have the tools and know-how to measure those metrics well. Using the design process, they use those metrics to get the best sound in their price range. If there's an easy and cheap way to improve sound, they do so. To think otherwise is to think the engineers amazingly stupid.

Another example of poor thinking operates around speaker materials. "This speaker is made with paper, so of course it's sounds bad. All the cheap speakers are made of papers." Compare this with, "The car is made of steel, so of course it drives bad. All cheap cars are made with steel." The materials used in a product do not determine the final quality of a product. Bad speakers are bad because they are designed that way, not because they are made from paper. Even "paper" isn't just paper. Different formulations of paper at different price points will produce different speakers. A well designed and constructed paper speaker can sound good because engineers, you remember them, record the output of those speakers and measure the distortion. If the measured distortion is too high, they redesign the speakers. Paper may be one of the indicators of quality, but by no means is paper the only indicator of quality.

Products are not made and manufactured in a vacuum. They are designed to markets with certain price points in mind. Over time, improvements are made to every level of product design. True audiophile equipment is made by mindful engineers who put their best attention into making a good products. Any easy improvement to the products have already been made. These products meet their design goals for that price point, for that market, and for that intended use. This means that, generally speaking, money becomes a good metric for quality.

Some things can't be reduced to metrics. Given a list of promises on a box, does the product deliver on those promises with your setup? How does the product compare against the competition? How do the headphones fit? How aesthetic is the unit in question? I don't know a way to measure those things.

Note that knowing this doesn't mean a thing. It won't keep you from get ripped off. We've all spent badly because metrics don't mean anything, and there's always someone willing to lie and take you lunch money at every price point. Metrics work great until somebody starts lying.