I too would love to see this, a simple 0-1 value global blend or strength would be great. The OpenImage implementation as it stands now is almost always far too heavy-handed to be used for final output frames. It would be a huge quality of life improvement if we could use it for times when de-noising in post isn’t in scope