Quite a long time ago, a friend of mine (who holds good qualifications in electronics and related fields) claimed that a certain brand of very cheap hi-fi systems had a ‘noise reduction’ feature that worked by injecting noise into the signal when it was turned ‘off’ - i.e. when you turn ‘on’ the noise reduction button, the machine stops adding noise to the signal and it sounds better, so you are left with the impression that it works.
I’m skeptical about this claim on several levels:
-It sounds like it would be illegal.
-Adding circuitry to generate noise is not going to be much cheaper than adding genuine noise reduction.
-It sounds like a conspiracy theory.
So, has anyone ever heard about this actually happening?
I don’t know the answer to points 1 and 3, but I don’t think 2 is right (i.e. that the costs of adding noise would be the same as taking it away). Assuming we’re not talking simply about tape decks (most of which would have dolby b/c noise reduction - I don’t believe this is an expensive add-on), I guess we’re talking about single-ended noise reduction. The unit I have to accomplish SENR works in two ways; it has a downward expander (sometimes called a noise-gate) which monitors the signal level and essentially turns the signal off during periods of near-silence, i.e. at the noise-floor level, plus some quasi-intelligent high-frequency filter which acts to reduce hf noise when it detects that hf are above a given % of the total signal - on the basis that hf hiss is more easily heard.
Ramble ramble. The point is that this unit cost me about 200 quid, and I’m going back nearly 10 years. Admittedly it’s designed for a semi-pro project studio, but even so…
BTW; did he mention the brand? I’d be interested… (knowing my luck, it’ll be what’s sitting in my living room… )
I don’t know about your question about the ‘cheap’ consumer hi-fi problem, but as to the general question, yes sometimes ‘noise’ is injected into an audio signal.
In digital audio production it is called dither. Its a low level digital noise that is used to mask errors caused by the conversion of an analog signal to a digital signal.
It’s not really used for noise reduction, but to make digital audio sound less ‘harsh’ or ‘shrill’ and more pleasing to the ear.
If you google for ‘dither’, you can eventually find some technical explanations that explain the process in detail.
Noise reduction is definitely NOT as cheap or easy as noise creation. Injecting a little noise would be an absolutely trivial thing. Removing noise takes lots of hard work. I’ve done some work on implementing single ended noise reduction on a DSP. Not easy, not trivial, and fairly expensive even at today’s low prices for components.
To generate noise, all you would need is an amplifier (which your radio has out the wazoo) set to a high gain with one input tied to audio ground directly and the other to audio ground via a resistor. Noise for the price of one of the op-amps in one of the quad op-amp ICs that you had already there.
I don’t know the specifics of Dolby noise reduction, but it is much more complicated than a single amp and a couple of passive components.
As for somebody actually doing it, I don’t know. It sure sounds like the kind of crap some marketing doofus could come up with, but I don’t know if any one has actually gone through and carried it out.
There are ways around licenseing and in any event Dolby type B noise reduction isn’t all that complex, just a band limited form of compression. In the '80s I bought an add-on noise reduction unit for a walkman. No, I’m not kidding, I don’t know what I was thining. It had two modes, one for dbx type II compression and one for “type B,” apparently it’s clone of Dolby B.
