...
I can't agree with you. The purpose of a bridge is to support a load. The purpose of a traffic signal is to regulate traffic. These things can be measured objectively. Ah, but consider the following. Traffic jams are stressful. Stressed people become agitated and distracted and this leads to traffic accidents. Perhaps one measure of the utility of a traffic signal network is the extent to which drivers subjectively perceive traffic to be smooth-flowing, and whether or not the delays they encounter are fair. How do you measure that with a meter? This reminds me of the social engineering theme park designers do to make long waits in line tolerable to their guests.
The purpose of an amplifier is not to produce a chart on an audio analyzer. It is to render enjoyment to a listener....
Two things you've said here. On the first one, I have professional expertise, so I will address that (maybe not so) briefly. First, the purpose of a bridge beam is not just to support a load, but to support that load within a range of parameters, including deflection, resonance (vibration), size, weight, and cost. If a bridge beam supports the load but sags too much, or not enough, the traffic on it will bounce along. Very embarrassing. Yes, it happens more often than you think. Lawsuits often result. Even worse is when the structure looks great until the vibration causes the joints to fail or be damaged, or until the vibration scares people on the bridge, or until the vibration sets up an uncontrolled resonance that tears the bridge apart (the poster-child example for this, of course, is the Tacoma Narrows bridge disaster).
And, bravo (no sarcasm--I mean it!), you have identified two main objectives for traffic signal timing: In a network, smooth flow, and at an intersection, fair distribution of green. That puts you ahead of a lot of people responsible for signal operation. (Those two objectives don't work in the congested regime, but that's been one of my specialties.) How do we measure smooth flow and fairness? We listen to citizens, of course, and do so very carefully. But we are doctors and they are telling us where it hurts. We certainly do not assume their diagnosis is correct, nor do we accept their suggested treatment at face value. Then we measure. It is true that measurement used to be a highly trained traffic engineer making observations on the street, because that's the best sensing analysis we had, and this led to all sorts of disagreements at all sorts of levels. These are not subjective measurements even if they are qualitative (there is a difference) and anecdotal (there is a difference there, too). The engineer is looking for specific features based on understanding cause and effect. But in the last decade or two we have learned how to use data far more effectively for performance measurement. For example, I can measure smooth flow using a Purdue Coordination Diagram (Google it). I can measure fairness using degree of saturation on each approach, or lack of fairness by looking at the reason the signal controller terminated the greens for the movements at the intersection. There is a vast body of knowledge there, and the science of it is absolutely based on careful measurement and data. That science may be theoretical--I studied under Robert Herman, who developed a basic theory of town traffic called the Two Fluid Model--or it may be empirical. Let's just say that traffic guys at the more advanced levels paid attention during statistics classes.
And, believe me, I have heard about where it hurts. For the time I was in local government, I personally answered something like 12,000 citizen complaint calls about signals in the large city where I was in charge of their operation. (I believed the guy in charge should explain the decisions, not some poor call-taker who has no idea.) Traffic engineers have to understand "preference" as much as "operation", and how they are different. They also should understand the difference between objectives and measurements.
On your second paragraph that I quoted, the purpose of an amplifier may not be to transparently enlarge the signal. But lots of amplifier designers claim that this is their goal, but that achieving their goal requires stuff the plain measures of transparency don't capture. So, they are claiming something mystical, and cry crocodile tears when detailed and careful measurements refute their claim. Is Nelson Pass claiming this? I don't think so, and Amir's tests indicate that his measurements are
completely consistent with the specifications Pass claimed. No, it's not a refutation of Nelson--he designed a fun project that teaches some things, including that a home-made distortion box might still sound pretty good within the very limited context of its operation. But his followers have expanded that claim to include that the design actually sounds
better. For that to be true, high levels of distortion must be favorable to music reproduction, yet the artist and the producers have already added all the distortion they wanted to add. Is it the amp's job to add more? If Nelson presented this as, "This little project will demonstrate high distortion--see if you can hear it" then we wouldn't be having this conversation. If he said "This little project will demonstrate that distortion makes some recordings sound better to me" then we might disagree but that really is in the domain of preference. If he says everyone prefers it, then we would want to see preference testing. But even then it would be advice to producers, and the home equipment should transparently reproduce the distortions added by the producer and artist as a result of that preference testing. But that's not what I see--I see a bunch of people saying this little amp represents something
better, and defending that point of view when measurements challenge it.
But it's like most specialized fields: Only the specialists in the field really understand how much engineering and measurement there really is. Untrained users often run into things they don't understand, assume the wrong causal relationship subjectively, and then draw the incorrect conclusion.
Rick "guess what I do for a living" Denney