When you listen to FM radio you may not notice immediately that there is a whole world behind it: the world of sound processing. With equalizers, compressors and numerous other steps, the sound of a radio station is optimized to sound consistent and have an own ‘sound’. With broadcasts through FM there are a lot of extra rules that your signal needs to comply with. Every radio station wants to sound as good as possible, and preferably as loud as possible, but within those rules. Globally there is a handful of companies that concern themselves with sound processing for FM radio. Thimeo is one of those. Our main product, Stereo Tool, is used globally by more than 1000 FM stations and many more thousands of internet radio stations. Because there are quite some rules for FM broadcasts, a lot can go wrong. Most sound processors have so many configurable options that you as a user can do a lot wrong. If you do not exactly know what is going on, or you are not paying attention for a moment, you do not comply with the rules (with a substantial fine as consequence) or your station does not sound as good as possible. Many (smaller) stations do not have an expert in this area and instead a hobbyist gets the noble task to configure the processor (consequence: misère) Many of the criterions where such an FM signal on a technical level needs to comply with are checkable simply by measuring. And that is good to automate. Other properties of FM broadcasts are more difficult to capture in a number (although for the human ear more easily perceptible). An example is “Radio 538 has heavier bass than SkyRadio”. But also for these kind of properties metrics can be invented with which you can more or less objectively compare two stations. The project aims first at checking the technical properties of such a signal, and then analysing a number of properties of the sound. To be able to do this a bit of R&D is involved, because there are no default measurements to analyze these properties. The ultimate end goal is to let a program analyze a FM signal and provide feedback on its characteristics afterward (for example “your station sounds a lot softer”).
Date: Jan 27, 2017
Author: Gerdriaan Mulder