One Way Analyzers Can Lie

One Way Analyzers Can Lie

Today I was helping a buddy of mine bench-test some active monitor wedges. My audio interface decided to kick the bucket, but that’s another issue. These are 12″ coaxial boxes that have a very small form factor. I really like them. He had six of them and it was clear that at least some had an issue.

The method is a comparative one, described here. It’s very difficult to take a full-range measurement free of boundary effects, etc, but there’s also no need. Measure each box under the same conditions and the ones that deviate from the norm are the ones with problems.

First up is two normal boxes on top, vs one with a dead LF driver (or amp channel) below.

I split the panes to show the “good” and “bad” coherence, but when overlaid the HF from the bad box matches perfectly. That one’s easy to spot and (obviously) clearly audible as well. Sounds like a telephone.

But the analyzer doesn’t always tell use the whole story.

The maroon trace in the image below matches the previous “good” boxes perfectly.

The small phase deviation is just a result of the boxes not being in exactly the same spot for the measurement. About 90° at 10 kHz, or 25 microseconds which is placement off by a third of an inch. That’s not a problem. So we can’t just look at whether the lines overlay or not. Sometimes they don’t overlay, and there’s no issue. Conversely, the coaxial driver mount in this wedge has snapped, and the HF driver is just sitting in place due to gravity. It tests fine with relatively steady-state test signals, but music playback makes the thing jump and rattle crazily. Also tilting the wedge from side to side while it’s playing results in immense distortion. If we were only looking at the measurement, this would appear to be perfectly normal, and we’d screw up a gig in a big way.

The analyzer can’t tell you everything you need to know. It can only answer the questions you ask it, so we always need to keep in mind the context of what we’re measuring at a given moment and how that measurement is obtained. Eyes, ears, and brain will always be the best analyzers.

Leave a Reply

Your email address will not be published. Required fields are marked *

Secured By miniOrange