![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Rereleasing songs
BBC Four shows a good variety of documentaries. Earlier this month they showed
One discovery was that in terms of the extracted features the more successful songs tend to be those closest to the current mainstream. Today it occurred to me that the algorithmic research could be monetized without being able to adjust a given song to increase its likelihood of success. Instead, select songs that were not much successful and were some considerable distance from the mainstream at the time that they were released, but that are close to the current mainstream, then rerelease some version of them. After all, it is hardly unprecedented for an artist to have a hit with a song that had been released years earlier to considerably less fanfare.
The Secret Science of Popfrom which my impression now is that from the past few decades of popular music in the UK they were somewhat successful in doing some kind of feature extraction and weighting then using that data to derive a statistically predictive measure of success in the sales charts. They however failed to extend this to be able to adjust a given song to make it more likely to succeed commercially.
One discovery was that in terms of the extracted features the more successful songs tend to be those closest to the current mainstream. Today it occurred to me that the algorithmic research could be monetized without being able to adjust a given song to increase its likelihood of success. Instead, select songs that were not much successful and were some considerable distance from the mainstream at the time that they were released, but that are close to the current mainstream, then rerelease some version of them. After all, it is hardly unprecedented for an artist to have a hit with a song that had been released years earlier to considerably less fanfare.