High GAIN=Higher gain=More compression?

[Viper Necklampy]

So, i didn't well speak what Gain level on specification on amplifier means exactly, maybe it act like a sort of compression on sound quality?
For example:
JDS Atom Labs 1.0W 32ohm: Low gain=x1.0/High Gain=x4.5
THX AAA 789 Unbalanced 1,8W 32ohm: Low gain=x0,33=-10dB / Medium Gain=x1.0=0dB / High Gain=x3.3=+10dB
SMSL Sp200 balanced 3W 32ohm: Low Gain=+6dB(x2.0) / High Gain=+18dB(x6,5)

See what i mean? So i assume jds atom is 0dB ''Boost?'' on low gain and around +12dB on High gain.
Am i talking right when i call this a ''Boost''? It is really a 'compression' or some sort of it? Like if when i go on youtube i see +5dB content, is more loud, more bassy and you could hear that is sounds like a compression, also lot of time with less dynamic range?
But anyway, without the ****** YouTube example, can i have an explanation of this 'Gain level' with boosted dB and gain multiplier (x2.0dB etc..)?
Is interesting, because i never see it explained and i'm curious

 

[PierreV]

The simplistic, practical way I see this (I am sure you'll get more detailed answers) is as follows

- gain multipliers are somewhat relative if you compare different devices.
- more gain -> more noise (with some nuances in the sense that an amplifier can be/are designed to operate at an optimal specific absolute gain level).

so, why more gain if it often leads to more noise you may ask? Well, simply because the loads vary a lot (say a sensitive IEM vs a 600 Ohms headset)

I imagine running a high gain multiplier on a sufficiently powerful amp can sound "compressed" and mostly distort on sensitive IEMs or even potentially blow them, but within range, it shouldn't matter too much in practice even if a higher noise floor intrinsically limits the maximum DR.

If you think in terms of digital camera, the sensor as a native gain value which is best (taking into accounts several things such as read noise, dark noise, quantization noise etc...) and the ISO setting is just a gain multiplier or divider for the convenience of the photographer

 

[Killingbeans]

See what i mean?

Not really...? There's a good chance I'm just speaking (well.. writing) out of my ass now, because I'm no expert, but I think you're only experiencing compression on YouTube as a downside of how everything is processed to minimise the use of internet bandwidth.

EDIT: Like somebodyelse points out below, talking about audio compression the way I just did actually doesn't makes sense. But if we were talking about "analog" radio stations and the way they use compression, it would.

More gain in an amplifier gives you higher voltage swing, but also raises the noise floor accordingly, and that's all there is to it?

It's a trade-off between power and noise.

 

[somebodyelse]

Gain is just a multiplication - what you get out is a multiple of what you put in. The ratio between the loud bits and the quiet bits (dynamic range) doesn't change.
Compression (in the audio sense not the data compression sense) reduces the dynamic range by applying less gain when things are loud than when they're quiet - at least that's the simplified explanation.
The 'loudness' effect (perceiving more bass when you turn up the volume) is because human hearing sensitivity isn't linear - read up on equal loudness contours for the details.

 

[Viper Necklampy]

Gain is just a multiplication - what you get out is a multiple of what you put in. The ratio between the loud bits and the quiet bits (dynamic range) doesn't change.
Compression (in the audio sense not the data compression sense) reduces the dynamic range by applying less gain when things are loud than when they're quiet - at least that's the simplified explanation.
The 'loudness' effect (perceiving more bass when you turn up the volume) is because human hearing sensitivity isn't linear - read up on equal loudness contours for the details.

So, if gain is a multiplier of real power, i guess it is a negative for SQ, like a compression is negative too..
So sp200 might have the better power by little, because 2000mw in single ended at just 2x gain (low gain) , while 789 do 1800mw 2x gain, and Atom just 1000mw at x4,5 gain. Well, THX amps might be much better than atom in power , at least on the card.

 

[NTomokawa]

i guess it is a negative for SQ

Not necessarily. It all depends on implementation. Shopping amplifiers based on gain is absolutely pointless.

 

[Viper Necklampy]

Not necessarily. It all depends on implementation. Shopping amplifiers based on gain is absolutely pointless.

Anyway, based on 3 response above, if i'm right, seems like higher gain (Multiplier gain) Is made to have more power, with of course some compromises. Good to know! Because it is not really measurable, apart to know if it is +6dB/x2, or x6 or so, just like power for head room and the benefit of sound quality is not measurable..
But you're telling it all depends on implementation? Like what, a 1000€ amp with 2x gain might have less compromises on higher gain than 400€ amp at almost same measurements? It's a thing of component? I don't get it honestly.

 

[patate91]

You have to check for gain with an active setup.

Let say you use to different amplifiers with different gain. Tweeters and woofers have different sensitivity. You'll need to l"evel match" everything.

Look for gain structure.

 

[Blumlein 88]

Okay look gain is simply multiplication. If the input is 1 volt, the output wtih 12 db of gain is 4 volts. Waveform is the same no compression. If the input had a signal at 1 volt and some unrelated noise at -60 db, the noise is at .001 volt. Apply 12 db of gain and the output signal is 4 volts and unrelated noise is now .004 volt. The same ratio between signal and noise. Gain basically doesn't change that.

Compression is actually compression and "make up gain." Let us say we set our compression to be 4:1 compression for the top 20 db of the total range available. For signal levels below -20 db nothing happens. As signals pass above -20 db the level is reduced by our 4:1 ratio. So the maximum signal level is not 0 db anymore, but -15 db because the upper 20 db has been compressed to fit into only 5 db of range. This would make the peak signal levels lower, so make up gain is applied. This pushes the max compressed signal of -15 db upward by 15 db so it can peak at 0 db again. In the process it has applied gain to all the signal including stuff below -20 db. So your weak signals have been pushed up, making low level detail louder and easier to hear. It also raises the noise floor upward by 15 db. By the time you have done all this, you've reduced total dynamic range by 15 db, squashed that range into a smaller dynamic range, and made the average level higher than it otherwise would be. So it sounds louder. Taken to extremes compression can take a recording with a few loud portions with a dynamic range of 80 db, and squash it into loud music that has only 10 db dynamic range and no actual quiet portions.

 

[Viper Necklampy]

Okay look gain is simply multiplication. If the input is 1 volt, the output wtih 12 db of gain is 4 volts. Waveform is the same no compression. If the input had a signal at 1 volt and some unrelated noise at -60 db, the noise is at .001 volt. Apply 12 db of gain and the output signal is 4 volts and unrelated noise is now .004 volt. The same ratio between signal and noise. Gain basically doesn't change that.

Compression is actually compression and "make up gain." Let us say we set our compression to be 4:1 compression for the top 20 db of the total range available. For signal levels below -20 db nothing happens. As signals pass above -20 db the level is reduced by our 4:1 ratio. So the maximum signal level is not 0 db anymore, but -15 db because the upper 20 db has been compressed to fit into only 5 db of range. This would make the peak signal levels lower, so make up gain is applied. This pushes the max compressed signal of -15 db upward by 15 db so it can peak at 0 db again. In the process it has applied gain to all the signal including stuff below -20 db. So your weak signals have been pushed up, making low level detail louder and easier to hear. It also raises the noise floor upward by 15 db. By the time you have done all this, you've reduced total dynamic range by 15 db, squashed that range into a smaller dynamic range, and made the average level higher than it otherwise would be. So it sounds louder. Taken to extremes compression can take a recording with a few loud portions with a dynamic range of 80 db, and squash it into loud music that has only 10 db dynamic range and no actual quiet portions.

So gain-multiplication means just multiply the noise (distortion). Do, does ultra-transparent tech like thx, 'doesn't care' noise multiplication, as it have ultra low level of distortion or 'noise'? So i guess even 12-20x instead of 2x or 6x gain would be ok with a very black background of noise?

 

[DonH56]

Gain determines how much output signal there is for a given input.

If the gain is 10 and you input 1 V you get 10 V out. If the gain is 20 and you input 1 V you get 20 V out. If the output clips at 15 V, then for a gain of 10 you do not reach clipping (max output), and for gain of 20 you'll clip before the input reaches max so you'll need to turn down the volume. Gain is just one parameter and by itself does not really tell you anything; you need to know the signals going in and going out and the characteristics of the entire signal chain.

Gain affects noise, distortion, and the base signal. Everything put in gets amplified (multiplied) by the gain. Gain alone won't change the signal-to-noise ratio (though real gain stages do add noise and distortion; usually very little so you can neglect it).

HTH - Don

 

[AtomAmp]

Hi. I am new in this forum. Thank you for accepting my registration.

This topic picked my curiosity. I have a practical question which we encounter when shopping for headphone amplifiers:

If the headphone amplifier specs says it has low gain (+0db, unity) and high gain (say +9db) and it has 500mW maximum power at 32 ohms, assuming standard 2V. Is it correct to say that if I don't want to use high gain, so I use low gain, then I get only 62.5 mW maximum power at 32 ohms, assuming my math is correct (500/8) ? Then if i use 300 ohm headphones I get only around 6 mW maximum power (62.5/[300/32])?

Thanks. I'm obviously not an engineer.

 

[Doodski]

Oops misread and deleted post.

 

[twsecrest]

Hi. I am new in this forum. Thank you for accepting my registration.
This topic picked my curiosity. I have a practical question which we encounter when shopping for headphone amplifiers:
If the headphone amplifier specs says it has low gain (+0db, unity) and high gain (say +9db) and it has 500mW maximum power at 32 ohms, assuming standard 2V. Is it correct to say that if I don't want to use high gain, so I use low gain, then I get only 62.5 mW maximum power at 32 ohms, assuming my math is correct (500/8) ? Then if i use 300 ohm headphones I get only around 6 mW maximum power (62.5/[300/32])?
Thanks. I'm obviously not an engineer.

The gain setting is not really that big a deal, it's somewhat more a personal preference situation.
Think of the gain setting as more a loudness boost (pre-voltage boost)
It takes the same amount of voltage to get a given loudness for a headphone, the gain setting does not change that particular factor.

 

[Blumlein 88]

Hi. I am new in this forum. Thank you for accepting my registration.

This topic picked my curiosity. I have a practical question which we encounter when shopping for headphone amplifiers:

If the headphone amplifier specs says it has low gain (+0db, unity) and high gain (say +9db) and it has 500mW maximum power at 32 ohms, assuming standard 2V. Is it correct to say that if I don't want to use high gain, so I use low gain, then I get only 62.5 mW maximum power at 32 ohms, assuming my math is correct (500/8) ? Then if i use 300 ohm headphones I get only around 6 mW maximum power (62.5/[300/32])?

Thanks. I'm obviously not an engineer.

If you mean 2 volt input standard and with unity gain you get two volts out, then with 32 ohms, you'll get 125 mW.

2 volts / 32 ohms= .0625 amps. So 2 volts x .0625 amps= .125 watts or 124 mW.

2 volts / 300 ohms= .00667 amps. So 2 volts x .00667 amps = .0133 watts or 13.3 mW.

 

[AtomAmp]

The gain setting is not really that big a deal, it's somewhat more a personal preference situation.
Think of the gain setting as more a loudness boost (pre-voltage boost)
It takes the same amount of voltage to get a given loudness for a headphone, the gain setting does not change that particular factor.

Thank you for your response. I agree with you that it takes the same voltage to drive a certain headphone to a certain volume level, whether it is on low gain or high gain. But my question I guess is if I am on low gain and I am at the max volume setting, can I still get the power swing (or perhaps voltage swing) needed in the song? Otherwise stated, if on low gain the max power on 300 ohm headphones is around 6mw per my example, and I am at the max volume setting, does it mean the headphone amplifier cannot give my 300 ohm headphone another 1mw or so?

 

[Blumlein 88]

Thank you for your response. I agree with you that it takes the same voltage to drive a certain headphone to a certain volume level, whether it is on low gain or high gain. But my question I guess is if I am on low gain and I am at the max volume setting, can I still get the power swing (or perhaps voltage swing) needed in the song? Otherwise stated, if on low gain the max power on 300 ohm headphones is around 6mw per my example, and I am at the max volume setting, does it mean the headphone amplifier cannot give my 300 ohm headphone another 1mw or so?

No the amp can't give more. In the case of the 300 ohm load, the voltage will be the limiting factor. If you switched to high gain, and turn down the volume the amp can give you more into the 300 ohm load. How much more depends upon the max voltage the amp can put out.

 

[AtomAmp]

If you mean 2 volt input standard and with unity gain you get two volts out, then with 32 ohms, you'll get 125 mW.

2 volts / 32 ohms= .0625 amps. So 2 volts x .0625 amps= .125 watts or 124 mW.

2 volts / 300 ohms= .00667 amps. So 2 volts x .00667 amps = .0133 watts or 13.3 mW.

Wow, thank you. So my assumptions are obviously wrong. But let me try to take this in.

Please remove from my assumption the 2v input. I just assumed it based on standard specs I guess.

So in reverse order, what is the voltage input of 500mW at 32 Ohms at unity gain?

 

[RayDunzl]

So in reverse order, what is the voltage input of 500mW at 32 Ohms at unity gain?

http://www.sengpielaudio.com/calculator-ohm.htm

 

[AtomAmp]

No the amp can't give more. In the case of the 300 ohm load, the voltage will be the limiting factor. If you switched to high gain, and turn down the volume the amp can give you more into the 300 ohm load. How much more depends upon the max voltage the amp can put out.

Ok. But what about the other aspect of it in current? It cannot give more current?