Electronics project notes/Audio notes - noise reduction
Dolby noise reduction
|This article/section is a stub — probably a pile of half-sorted notes, is not well-checked so may have incorrect bits. (Feel free to ignore, fix, or tell me)|
Dolby itself is a company / brand.
It is probably most associated with its noise reduction (aimed at reducing noise on tape), and Dolby surround.
For context, tape hiss refers (roughly) to unavoidable hiss due to the nature of the medium. That hiss is also also a few dB stronger above a few kHz.
Dolby Noise Reduction addressed this with some selective companding.
It does this to higher frequencies not only because the hiss is stronger there, but also because higher frequencies are typically much quieter, so it's unlikely this will cause any saturation / distortion.
For example, a Dolby B encoder (recording stage) boosts the frequencies about ~7kHz (with a well defined defined curve) by 10dB, and a decoder (player) lowers it the same way and by just as much. (So yes, it's much like a well defined EQ, and if you have a deck without Dolby, you can imitate it with EQ)
Playing a Dolby B recording on a dolby B player means the end-to-end frequency response is the same, but the higher frequency parts of the signal is ~10dB stronger on tape -- so the tape-introduced noise is effectively ~10dB lower end-to-end.
- playing a recording without NR on a player set to Dolby B will sound dull
- because it lowers the higher frequencies for no good reason
- playing a Dolby B recording on a player without NR or with it disabled will sound unnaturally bright
- because it's playing booster high high frequencies
Dolby B was popular in part because it's simple, and probably in part because it's subtle enough that playing B-recorded tapes in players without Dolby would sound bright (brighter than intended but not exactly a bad thing, particuarly if you had an EQ).
Dolby C is the same idea, but is ~10dB stronger and starts to come in two octaves lower.
This is more noise reduction, but also strong enough that playback without NR sounds wrong. Though, because it's roughly two B passes (strength-wise, anyway), playback on decks without C still sound decent when played back set to B.
Note that B and C do not particularly deal with transients.
Some Dolby variants introduced adaptive gaining and frequency masking (verify).
There are more than most of us have probably heard of:
- Dolby A - 1965, four-band, and embedding a warble tone that could be used to identify this type, and assist alignment. Used in some pro gear, cinema, and more
- Dolby B - 1968, consumer, basically one-band, mainly used in cassettes
- Dolby C - 1980, consumer, stronger than B
- Dolby HX - 1980, not reduction at all, just a way to void over-biasing while recording.
- Dolby SR - 1986, ten-band, sort of a refinement on A
- Dolby S - 1989, roughly something inbetween B or C and SR
The more professional variants take more care, to e.g. make the level production a little more truthful to the original.
That, and the warble tone stuff, was not so interesting to home recording so the fancier types, with more expensive hardware, rarely made it to consumer players.