Reverse engineering

Prev Next

Reverse engineering
I remember my first trip to Hong Kong back in the 1990s. Breakfast that morning was to be fruit, Miso soup, rice, and salted plum. When the server brought my plate out I was pleasantly surprised to see my fruit, a lovely orange/yellow mango, opened for the eating like the top half of a ball with individual-sized mango squares ready to be devoured. It was the cleverest thing I had ever seen. How did they do it? I pulled at the edges to "deflate" the half ball back into its original form of a simple half of a mango. Clever. By carefully scoring the mango meat into the small squares and then folding open the shape, a lovely presentation was created. I had reverse engineered the technique. This method of working backward from a finished product is nothing new. All engineers do the same thing if they need to unravel an existing product or technology. Moving from mangos to audio, if we hear two pieces of identically measuring equipment sound very different, how do we determine what's going on? Do we propose that because they measure the same they must be the same and therefore dismiss that which experienced listeners report? Or, can we take another route? Perhaps we would consider reverse engineering the problem. Imagine for a moment measurement equipment designed to measure only THD and frequency response (as once was done as the only two meaningful measurements). We take the two different sounding units and measure them both. Let's for sake of argument imagine they measure the same. Yet still, they don’t sound the same. One crowd dismisses the other by proclaiming them to be the same despite the imperial evidence to the contrary. Now, imagine we add another aspect to the test. IM distortion (which we previously ignored). One is high, the other low. Now we see why one sounds poorly while the other sounds good. The point I am attempting to make is that if one were wanting to really identify the mechanisms at play between two different sounding units, the process of working backwards until a measured difference was identified might just be a good way to go. If I had more time and desire for such exercises I might be the first to jump in. For me, it's a lot easier just to listen.
Back to blog
Paul McGowan

Founder & CEO

Never miss a post


Related Posts