Yes, but it's pretty sophisticated. They use pattern matching broadband noise reduction. The way it works is they take a section of silence... the lead in groove in a record, or the space between songs in a tape... and analyze the noise in it using a computer plugin. It determines the frequencies where the noise exists and the proportion from one frequency to another. Then it gives you a slider control where you can dial in more or less noise reduction. This can also be applied dynamically, more noise reduction in quiet passages where hiss would be intrusive, and less in loud passages where the hiss is buried under the signal. If the settings are carefully judged, it works very well. If you want to try it, there is a VST plugin called Sound Soap that does this.
If the recording is band limited and doesn't have a lot of information above 6 or 7kHz, the engineer usually leaves a small bed of even hiss in the track. Having some higher frequencies, even if they're just noise, tricks the ear into thinking it has treble that doesn't really exist in the music. It's a psychoacoustic trick.
Edited by bigshot - 4/21/14 at 10:51pm