[metadodo]

Ok so to make it short, what does positive gain mean?

 

[mikeaj]

edit: see below

 

[brconner]

Quote:

A positive gain means that whatever the input was, the output is larger (more).  i.e. you gained something
 
A negative gain means that whatever the input was, the output is smaller (less).  i.e. losing is the opposite of gaining, so you lost something

Usually output/input is described in ratios, which lends itself to expressing those numerical relationships in decibels.  http://en.wikipedia.org/wiki/Decibels.  Positive dB means positive gain (output > input), 0 dB means no gain (output = input), and negative dB means negative gain (output < input).

 

Actually a negative gain is a 180 degree shift in phase. Doesn't necessarily mean that its less volume. Also, gain can be expressed as a ratio (unitless) or in decibels (dB)
 
If the gain ratio is equal to 1 then there is no change in volume.
 
Greater than 1 and the output is louder than the input.
 
Less than 1 and the output is smaller that the input.
 
Also to change a ratio gain to dB...
 
dB = 20 * log( ratio gain )
 
Or to change dB to ratio gain...
 
ratio gain = 10 ^ (dB/20)

 

[mikeaj]

I hope discussion of negative gain is taken as relevant and on topic, though the OP did just ask about positive gain...
 
At first I thought the definition was one thing.  Then I saw your definition and thought you were right.  Then I looked around, and I'm not sure.  Either this may be an ambiguous case, or I'm not understanding something.
 
I see several places online where people use the term "negative gain" as I did to describe attenuation as I did (i.e. ratio < 1, or negative dB gain).  This is especially the case where people are referring to power quantities.  Of course with power quantities, negative power may or may not make sense depending on the application.  Maybe they're incorrect with this usage of the term.
 
From a field quantity perspective, what's of interest is signal amplitudes.  In this setting, negative gain implies multiplying the input by -0.1 or -5?  So-called inverting amplifiers are usually denoted with gains like that: -0.1 or -5 or negative-{some number}.  This eliminates confusion as to what is "negative."
 
In audio in general, negative gain refers to signal inversion?  Inversion of audio signals translates to 180 degree phase shift, as you mentioned.
 
I'm not sure if saying "gain ratio" and giving the 20 dB log rule is the most precise way of explaining things.  Naturally, the question becomes "ratio of what?"  dB = 10 * log (ratio gain) if the quantities being compared are powers, but it's 20 * log (ratio gain) as you have if it's signal amplitudes.  The meaning of "dB" needs to be inferred based on context clues.

 

[brconner]

The only reason i mentioned it was because of your previous post before mine. Didn't want to get the OP confused. A negative gain always means a 180 degree phase shift. Negative is a term thats used outside of the field but isn't really correct, unless the amplifier is inverting the signal.
 
The reasons decibels are used is because it makes the math really easy in certain situations, specifically when you are combing different systems together. If you want I can send you a pm with a very specific example, but I'd have to go very in depth. If people are interested I can post an example in this thread though. I'd just need to work on it so that its easy to understand haha.
 
Also what you said about my conversion is correct. The formula I gave was for signal amplitudes and not power. For amplitudes use the 20, for power use 10 instead of the 20. The reason for the different numbers has to do with how power is calculated.
 
Like you said decibels need to be based on a reference. You may have noticed that headphones have a sensitivity rating thats in "dB SPL" per milliwatt. That spec tells you how loud the driver will be given the power put into it. The decibel is used as a tool to show differences between numbers logarithmically.
 
So you might wonder why we don't just describe everything for audio in ratios. Well its actually not very intuitive that way. Sound is logarithmically based and a ratio is linear. Using dB's converts the linear ratio to a logarithmic system. If you have an amp with a gain of 2 that does not mean the volume will be twice as loud. 
 
I realize I haven't really explained what a dB actually is but you give you an idea...
 
If you're listening to music at 60dB and you want the volume twice as loud, it needs to be 70dB. If you convert that to a ratio it's a little over 3. Most people wouldn't think that makes any sense.
 
Here's a great link that explains decibels pretty well.
 
http://trace.wisc.edu/docs/2004-About-dB/
 
Hopefully I didn't confuse anyone further. This is definitely a very technical topic and can be a bit difficult to explain sometimes and I'm a little tired so thats not helping haha.
 

 

[mikeaj]

Replies in bold.  I cut out some material to make it easier to read.
 
Quote:

The only reason i mentioned it was because of your previous post before mine. Didn't want to get the OP confused. A negative gain always means a 180 degree phase shift. Negative is a term thats used outside of the field but isn't really correct, unless the amplifier is inverting the signal.
 
Yes, I mentioned whether or not it might be appropriate because I'm the one dragging this discussion along.  The fault lies with me, no doubt.
 
If you're listening to music at 60dB and you want the volume twice as loud, it needs to be 70dB. If you convert that to a ratio it's a little over 3. Most people wouldn't think that makes any sense.  
Just to reiterate what you're saying, what's perceived as twice as loud is roughly 10 dB greater SPL (sound pressure level), where SPL is on a 20 log scale.  The SPL at twice the perceived loudness is therefore about 3.162 times greater than the other SPL.
 
Here's a great link that explains decibels pretty well.
http://trace.wisc.edu/docs/2004-About-dB/
 
Thanks for the link.
 
Hopefully I didn't confuse anyone further. This is definitely a very technical topic and can be a bit difficult to explain sometimes and I'm a little tired so thats not helping haha.
 

And I spent about half an hour looking around academic publications.  I saw papers published in IEEE Transactions on Vehicular Technology, IEEE Transactions on Antennas and Propagation, etc. about antennas use the term "negative gain" to mean less power (referenced to the power output of the 1/2 wavelength dipole antenna) and not inversion.  One author was careful enough to write "they have a negative gain (in dB)," to make the meaning more clear.  I may have read some antenna spec or heard a lecture sometime in my life in which negative gain (in dB) was used in that context not to refer to inversion, which may have caused the confusion.  Or maybe I came up with it myself.  In any case, sorry for the confusion.
 
Papers in control theory, signal processing (I guess that would include audio), etc. that I saw all refer to "negative gain" as inversion. 
 
So for the OP who asked a question on an audio forum where it's safe to say that we only care about the meaning in this context, we say that positive gain is where the input and output are of the same polarity, i.e. not inverted.  Negative refers to an inversion, which is a 180 degree phase shift.
 
If we let K = output / input (the ratio):
 
-infinity < K < -1:  negative gain, output has higher magnitude than input
K = -1:  negative gain, output is the same magnitude as the input
-1 < K < 0:  negative gain, output has lower magnitude than input
0 < K < 1: positive gain, output has lower magnitude than input
K = 1:  positive gain, output has the same magnitude as the input
1 < K < infinity:  positive gain, output has higher magnitude than input
 
Hopefully I also did not confuse anyone, and hopefully I'm still not confused myself!

 

[brconner]

If the gain is being described in dB then yes you can have negative gain that doesn't invert the signal. It's simpler to design circuitry using a gain ratio so I generally don't think of gain in dB. Decibels are a great way to compare signals though, especially in communications which is probably why  you found some articles having to do with antennas.