New tidal

[Deleted member 108165]

Insider, I get the basic idea but this is beginning to go over my head a little *smile*

Specs for my amp: 70W @ 8ohm load and un-attentuated is capable of 106dB @ 2V input. Now I'm never going to need this kind of volume as I am more than happy with 75dB peak as loud enough for me, are you trying to say that my amp is only producing 10W because I've fitted attenuators to the inputs?

I've just performed an online calculation for the watts I would need for my situation:

Listening position 3m

Desired peak volume 75dB (with 15dB headroom?)

Speakers sensitivity 87dB

Result = 11W

Do I really need to remove my lovely attenuators? *wacko*

I do agree with you that these kind of volume pots are junk, more a case of marketing than decent engineering principles.

 

[insider9]

Insider, I get the basic idea but this is beginning to go over my head a little *smile*

I probably didn't explain it well enough, apologies.

Specs for my amp: 70W @ 8ohm load and un-attentuated is capable of 106dB @ 2V input. Now I'm never going to need this kind of volume as I am more than happy with 75dB peak as loud enough for me, are you trying to say that my amp is only producing 10W because I've fitted attenuators to the inputs?

No, what I'm saying is that a 7W amp with not attenuated signal would sound as loud as yours in that particluar scenario scenario. (10 times amp power is equivalent of 10dB volume difference)

Your amp is still outputing what it's supposed to you only make it harder as it's amplifying weaker signal.

I've just performed an online calculation for the watts I would need for my situation:

Listening position 3m

Desired peak volume 75dB (with 15dB headroom?)

Speakers sensitivity 87dB

Result = 11W

Do I really need to remove my lovely attenuators? *wacko*

I do agree with you that these kind of volume pots are junk, more a case of marketing than decent engineering principles.

Ok, do the same calculations but lower speaker sensitivity by 10dB. See how hard your amp is working by adding these attenuators.

All I'm saying that 10dB is a lot. 3-6dB would be much better. I believe attenuation can be beneficial just need to consider amp power and try and keep it low.

It's a crying shame that some manufacturers put volume pots like that.

 

[gasolin]

So it's hard for an amp with attenuators since it doesn't output max power at 10 or 11oclock?

A turntable is even harder for an amp since it in most cases it won't play,deliver max power as high as 12 oclock?

Attenuators is not just about volume range, it's also about sound quality

 

[insider9]

So it's hard for an amp with attenuators since it doesn't output max power at 10 or 11oclock?

A turntable is even harder for an amp since it in most cases it won't play,deliver max power as high as 12 oclock?

Attenuators is not just about volume range, it's also about sound quality

You're completely missing a point. What you set the volume pot to makes absolutely no difference whatsoever.

Please answer question in comment #49. Maybe you'll understand what I'm trying to say.

 

[insider9]

Proper gain structure is absolutely vital in any system and so is headroom. 10dB attenuation is simply too much in my opinion. You're removing headroom and limit amplifiers dynamic capability. Is that improving sound quality?

 

[gasolin]

Insider, I get the basic idea but this is beginning to go over my head a little *smile*

I probably didn't explain it well enough, apologies.

Specs for my amp: 70W @ 8ohm load and un-attentuated is capable of 106dB @ 2V input. Now I'm never going to need this kind of volume as I am more than happy with 75dB peak as loud enough for me, are you trying to say that my amp is only producing 10W because I've fitted attenuators to the inputs?

No, what I'm saying is that a 7W amp with not attenuated signal would sound as loud as yours in that particluar scenario scenario. (10 times amp power is equivalent of 10dB volume difference)

Your amp is still outputing what it's supposed to you only make it harder as it's amplifying weaker signal.

I've just performed an online calculation for the watts I would need for my situation:

Listening position 3m

Desired peak volume 75dB (with 15dB headroom?)

Speakers sensitivity 87dB

Result = 11W

Do I really need to remove my lovely attenuators? *wacko*

I do agree with you that these kind of volume pots are junk, more a case of marketing than decent engineering principles.

Ok, do the same calculations but lower speaker sensitivity by 10dB. See how hard your amp is working by adding these attenuators.

All I'm saying that 10dB is a lot. 3-6dB would be much better. I believe attenuation can be beneficial just need to consider amp power and try and keep it low.

It's a crying shame that some manufacturers put volume pots like that.

So do you know at what volume setting DougK's amp play max power wattage, how loud the music he is listening to is recorded?

If an amp has to work hard with low gain signal like a turntable where it only can play loud at 12 oclock (realtive to max power and speaker sensitivity) it must have been gift sent from heaven when they released the cdplayer in 1982

Do you also listen to the loudness war effected music that will make your no attenuated amp play max power at 9 oclock and enjoy it because you hardly ever have to move the volume knob before the music is loud? https://open.spotify.com/track/3JjZq20Kv6UphyyirAaY9A (look at the volume, you know in the right corner with the speaker icon)

 

[gasolin]

#203

https://www.harbeth.co.uk/usergroup/forum/the-science-of-audio/amplifier-matching-mismatching-and-clipping-a-curse/2174-how-loud-versus-how-far-you-turn-the-volume-control/page14

(funy how this thread started to be about tidal and now about volume gain,attenuators and generally power,watt,dynamic from an amp )

 

[andyjm]

An awful lot of misunderstanding on this thread. Just to clarify a few points:

Power - A power amplifier distorts all the time. This distortion gets worse as output power increases. The point at which it becomes objectionable is (in my opinion) the point at which it is producing its rated power. Unfortunately there is no real standard for this, but 0.1% THD+N seems a defendable position. Amazing how many amps have 1% THD+N as their threshold. Most listeners would have given up way before that.

Within reason, it is the output stages of an amp that define its power capability, the gain stages before the output are relatively irrelevant. As long as the intermediate stages aren't driven into clipping by too large an input signal, then it really doesn't matter if you attenuate the input and then boost the signal again prior to driving the output stages. A properly matched system won't require frantic twiddling of the volume control between sources to avoid blasting out your eardrums, and using attenuators may help in this regard. Again, attenuating an input so that the volume control is at a more useful point is also fine. An amp doesn't 'work harder' to produce rated power from a lower level signal.

However, this isn't the whole story. The curse of analogue is noise. An attenuator will decrease both the signal and the noise. Amplifying the signal again gets you back to where you started, but you are also amplifying the inherent input noise in the amp. If you have a long enough chain of attenuators followed by amps, then the output will be all noise and no signal.

So as pointed out above, a good design will have sensible levels and gain distributed through the system. Levels matched so that switching between inputs gives the same output sound level. Levels high enough that noise isn't drowning out the signal, but low enough so that intermediate gain stages aren't driven into clipping. Sequential gain / attenuation / gain is bad practice as it degrades signal to noise ratio.

It is quite an art to getting this right in an analogue design. I worked on a mixing desk once that had been poorly designed. It was very easy to have sensible input levels and sensible output levels, but have the intermediate stages distort as the buffer amps were driven into the supply rails and clipped.

 

[insider9]

Power - A power amplifier distorts all the time. This distortion gets worse as output power increases. The point at which it becomes objectionable is (in my opinion) the point at which it is producing its rated power. Unfortunately there is no real standard for this, but 0.1% THD+N seems a defendable position. Amazing how many amps have 1% THD+N as their threshold. Most listeners would have given up way before that.

Yes and more detailed than my reply (#50). Many amps quote far worse numbers than 1% THD+N as they can then quote higher power. Truly awful marketing practice that's desgined to mislead those who don't/won't read specs.

Also worth adding that how detectable distortion is to our ears depends on frequency and volume. Low bass notes in many listening rooms at average volumes are aften played at ~10% THD+N. It's not only not unusual, but sounds clean. That is of course not so much amp related but speaker and room. Far less distortion would be detectable in midrange.

Within reason, it is the output stages of an amp that define its power capability, the gain stages before the output are relatively irrelevant. As long as the intermediate stages aren't driven into clipping by too large an input signal, then it really doesn't matter if you attenuate the input and then boost the signal again prior to driving the output stages. A properly matched system won't require frantic twiddling of the volume control between sources to avoid blasting out your eardrums, and using attenuators may help in this regard. Again, attenuating an input so that the volume control is at a more useful point is also fine.

I was merely saying 10dB is quite a lot and would suggest 3-6dB instead if at all.

An amp doesn't 'work harder' to produce rated power from a lower level signal.

Consider this. When you attenuate signal, thus reducing headroom, amplifier will work closer to its rated power. As such its power consumption will be higher and it will likely produce more heat. I don't think it's unreasonable to say that this amp "works harder." You'll be paying larger energy bills "working harder"... But I agree it is an unfortunate phrase.

However, this isn't the whole story. The curse of analogue is noise. An attenuator will decrease both the signal and the noise. Amplifying the signal again gets you back to where you started, but you are also amplifying the inherent input noise in the amp. If you have a long enough chain of attenuators followed by amps, then the output will be all noise and no signal.

So as pointed out above, a good design will have sensible levels and gain distributed through the system. Levels matched so that switching between inputs gives the same output sound level. Levels high enough that noise isn't drowning out the signal, but low enough so that intermediate gain stages aren't driven into clipping. Sequential gain / attenuation / gain is bad practice as it degrades signal to noise ratio.

Properly designed gain structure is important. It's as bad attenuating too much signal without a good reason as amplifying it at multiple stages. I was simply questioning reducing over half the headroom, but it turned into "what time is on the dial debate"

It is quite an art to getting this right in an analogue design. I worked on a mixing desk once that had been poorly designed. It was very easy to have sensible input levels and sensible output levels, but have the intermediate stages distort as the buffer amps were driven into the supply rails and clipped.

I've used a pro amp once that had a quirk in design that unless you set its volume pots to max it produced detectable distortion. I wasn't happy with that solution, even though I was using a preamp earlier in signal chain.

 

[Deleted member 108165]

Thank you both Insider and andyjm. It's not a question of explaination, you have both explained it incredibly well... more a case of me being a little thick on this occasion *smile* Many thanks to both of you for your time and patience, I now understand. If I ever upgrade my amp I will look for one with a better designed volume pot.

Cheers guys, much appreciated *good*

Gasolin, apologies for thread hijack

 

[gasolin]

I've used a pro amp once that had a quirk in design that unless you set its volume pots to max it produced detectable distortion. I wasn't happy with that solution, even though I was using a preamp earlier in signal chain.

Behringer A500

 

[insider9]

I've used a pro amp once that had a quirk in design that unless you set its volume pots to max it produced detectable distortion. I wasn't happy with that solution, even though I was using a preamp earlier in signal chain.

Behringer A500

That's right!

 

[insider9]

It's only something to bear in mind. Some brands are better than others in this regard. Marantz isn't the worst, but far from the best. Esoteric I've got here now has 0.5 dB steps, but you don't have to spend £10k on an amp to get this. Densen I used before also had 0.5 dB steps and not a volume pot but mechanical relays that clicked with each volume change.

I'd still make the decision on how the amp sounds and would have to live with whatever the volume pot is.

Questioning attenuators wasn't really to stick it in to manufacturers, or make anyone feel bad about their gear. Just trying to understand why they're being used. Now that I know if you're able to swap them for smaller value ones you'd lose some of the pot movement but potentially improve the performance. Could result in better dynamics particularly transient reproduction. Attenuation can be beneficial but too much is almost always detrimental.

Simple test if you fancy it. Listen to some well recorded, high dynamic range, drumming with and without attenuators. Try to match the volume as much as possible before you listen. Might need a different track to set this up. If you can't hear a difference you're all good.

 

[Deleted member 108165]

Cheers Insider, I'll give it a bash, just gotta find a suitable track... I'm not a headbanger, more into pop music. The only one that springs immediately to mind would be Telegraph Road by DS.

There may some merit in your thoughts as LP playback has always sounded infinitely better than any digital replay on my amp.

 

[davedotco]

Andyjm covered a couple of these points, but pretty much everything else is nonsense.

An amplifier does not have to 'work harder' if you use an attenuator on the input, that is just plain wrong and the suggestion that this reduces the output available is nonsense.

The errors and misconceptions are so widespread that I actually thought that insider9 was setting up an epic troll.

We have discussed all this many times before in a number of threads, best to look them up.

Gasolin, one of the best ways to understand this is to think of it like this.

You have a 100watt power amplifier which may be separate or part of an integrated design. This amplifier is fully 'on' at all times, ie it's 'gain' is fixed, the output that it produces is directly proportional to the input.

Typically a power amp meeds an input of 0.5 to 1 volt to give full output, most digital sources output 2volts (Red Book Standard), often more, so in reality, the volume control is an attenuator itself. Using an external plug in attenuator simply changes the range of attenuation. There in no practical downside to this in normal use, any possible increase in noise will be way below audibility, the amplifier workload remains unchanged and there is no reduction in the amplifiers output capability, oh, and you get your volume control range back.

 

[insider9]

It's probably because I'm talking about signal attenuation whereas you're talking about input sensitivity matching

 

[davedotco]

It's probably because I'm talking about signal attenuation whereas you're talking about input sensitivity matching

I am really talking about gain. Modern amplifiers have more than enough, mostly due to high input sensitivity.

I am struggling to think of any digital based system where there is any downside to using attenuators as described by the OP.

And the suggestion that an amplifiers output is reduced by such attenuation or that an amplifier has to 'work harder' in such a situation simply does not make any sense to me.

If you want to outline your thinking, please feel free to do so.

 

[insider9]

Max amp output will be only achieved with full strength signal as close to 0 dBFS.

If your max signal is -10 dBFS (since you've attenuated 10dB) your amp output will be lowered. It will still produce max power, but this will translate to lower SPL.

I've mentioned earlier that if you overload the input there may be benefits of attenuation. But that, although attenuation is only sensitivity matching in real terms. If you lower 2V to 1V and amp still can provide max power with 400mV, you didn't really attenuate the signal.

 

[davedotco]

Max amp output will be only achieved with full strength signal as close to 0 dBFS.

If your max signal is -10 dBFS (since you've attenuated 10dB) your amp output will be lowered. It will still produce max power, but this will translate to lower SPL.

I've mentioned earlier that if you overload the input there may be benefits of attenuation. But that, although attenuation is only sensitivity matching in real terms. If you lower 2V to 1V and amp still can provide max power with 400mV, you didn't really attenuate the signal.

I really do not get this!

If the amp output is 'lowered', how can it still produce 'max power' and if it is still producing max power how can that 'translate to lower spl'? It can't, what you are saying makes no sense.

Whatever the level of the incoming signal, attenuated or not, as long as it is higher than the input sensitivity, the amp will produce full output, all that changes is the setting of the volume control.

 

[insider9]

I never disputed the last paragraph. I suspect this is where the confusion is coming from.

All I was discussing was signal attenuation. You're not actually attenuating signal if you're changing voltage above the threshold of amplifier input sensitivity.

 

[davedotco]

I never disputed the last paragraph. I suspect this is where the confusion is coming from.

All I was discussing was signal attenuation. You're not actually attenuating signal if you're changing voltage above the threshold of amplifier input sensitivity.

Just trying to be clear, both for me and anyone else reading this thread.

So using the -10dB 'pads' as in the OP's original post is not 'attenuation' if the lowered signal is above the threshold of the amplifier sensitivity. I understand what you are saying though I dissagree with the terminalogy you are using.

Given that most amplifiers have far more gain than is necessary, i do not think I have ever come across a case where 'padding' the input has ever reduced the input level below the threshold of amplifier sensitivity.

This is why I consider your earlier comments about the pads being detrimental to the performance of the amplifier to be 'nonsense', not going to happen with any modern amplifier with a digital source outputing 2 volts or more.

 

[insider9]

Just trying to be clear, both for me and anyone else reading this thread.

So using the -10dB 'pads' as in the OP's original post is not 'attenuation' if the lowered signal is above the threshold of the amplifier sensitivity. I understand what you are saying though I dissagree with the terminalogy you are using.

That's fine, there's no question that my inability to put my point across is what caused confusion.

Given that most amplifiers have far more gain than is necessary, i do not think I have ever come across a case where 'padding' the input has ever reduced the input level below the threshold of amplifier sensitivity.

Yes, true. In which case no signal attenuation would have taken place.

This is why I consider your earlier comments about the pads being detrimental to the performance of the amplifier to be 'nonsense', not going to happen with any modern amplifier with a digital source outputing 2 volts or more.

Comment was regarding signal attenuation overall, not pads themselves. - 10dB pads with 2V signal will effectively present 200mV (1/10th) to amplifier. As long as the amplifer input sensitivity is no more than 200mV then in regards to signal there has been no attenuation as seen by the amplifier. It's still full signal which in digital audio would be refered to as 0 dBFS.

Assume Leema Tucana II that I've reviewed here are the specs.

Sensitivity for maximum output (CD input): 565mV

Sensitivity for maximum output (other inputs): 311mV

With that particular amp you would be reducing headroom by sending 200mV singal to it, particularly with a CD input. Worth mentioning that Tucana can adjust gain on individual inputs up and down, so you'd never need to use pads with it.

Does this make sense at all?

 

[davedotco]

Just trying to be clear, both for me and anyone else reading this thread.

So using the -10dB 'pads' as in the OP's original post is not 'attenuation' if the lowered signal is above the threshold of the amplifier sensitivity. I understand what you are saying though I dissagree with the terminalogy you are using.

That's fine, there's no question that my inability to put my point across is what caused confusion.

Given that most amplifiers have far more gain than is necessary, i do not think I have ever come across a case where 'padding' the input has ever reduced the input level below the threshold of amplifier sensitivity.

Yes, true. In which case no signal attenuation would have taken place.

This is why I consider your earlier comments about the pads being detrimental to the performance of the amplifier to be 'nonsense', not going to happen with any modern amplifier with a digital source outputing 2 volts or more.

Comment was regarding signal attenuation overall, not pads themselves. - 10dB pads with 2V signal will effectively present 200mV (1/10th) to amplifier. As long as the amplifer input sensitivity is no more than 200mV then in regards to signal there has been no attenuation as seen by the amplifier. It's still full signal which in digital audio would be refered to as 0 dBFS.

Assume Leema Tucana II that I've reviewed here are the specs.

Sensitivity for maximum output (CD input): 565mV

Sensitivity for maximum output (other inputs): 311mV

With that particular amp you would be reducing headroom by sending 200mV singal to it, particularly with a CD input. Worth mentioning that Tucana can adjust gain on individual inputs up and down, so you'd never need to use pads with it.

Does this make sense at all?

A 10dB pad does not reduce the voltage to 1/10th. That would require a 20dB pad.

A 10dB pad will reduce a 2 volt signal to approx 632 mvolts, so perfectly ok in this example.

 

[insider9]

Just trying to be clear, both for me and anyone else reading this thread.

So using the -10dB 'pads' as in the OP's original post is not 'attenuation' if the lowered signal is above the threshold of the amplifier sensitivity. I understand what you are saying though I dissagree with the terminalogy you are using.

That's fine, there's no question that my inability to put my point across is what caused confusion.

Given that most amplifiers have far more gain than is necessary, i do not think I have ever come across a case where 'padding' the input has ever reduced the input level below the threshold of amplifier sensitivity.

Yes, true. In which case no signal attenuation would have taken place.

This is why I consider your earlier comments about the pads being detrimental to the performance of the amplifier to be 'nonsense', not going to happen with any modern amplifier with a digital source outputing 2 volts or more.

Comment was regarding signal attenuation overall, not pads themselves. - 10dB pads with 2V signal will effectively present 200mV (1/10th) to amplifier. As long as the amplifer input sensitivity is no more than 200mV then in regards to signal there has been no attenuation as seen by the amplifier. It's still full signal which in digital audio would be refered to as 0 dBFS.

Assume Leema Tucana II that I've reviewed here are the specs.

Sensitivity for maximum output (CD input): 565mV

Sensitivity for maximum output (other inputs): 311mV

With that particular amp you would be reducing headroom by sending 200mV singal to it, particularly with a CD input. Worth mentioning that Tucana can adjust gain on individual inputs up and down, so you'd never need to use pads with it.

Does this make sense at all?

A 10dB pad does not reduce the voltage to 1/10th. That would require a 20dB pad.

A 10dB pad will reduce a 2 volt signal to approx 632 mvolts, so perfectly ok in this example.

*dash1*

Good! At least now we know where I've gone wrong. Sorry, you're right, voltage is calculated differently to power. Thanks Dave!

 

[Vladimir]

What about how well music is recorded, produced, mastered? Why the obsessions with MINISCULE and many times IMAGINERY differences of sound quality between formats? That's like Oenophiles talking about wine bottles 1000 times more often than about the actual wine.

Listening to garbage at 10000000THz / 10000000000TBit is still listening to garbage. I find a lot of music I like unlistenable because it sounds horrendously bad on revealing gear and played loud. I don't care if its FLAC or 192 mp3. If anything the less resolving format improves it.

Is paying for lossless streaming worth the money? I've come to the conclusion that most music is made in every aspect to sound like *****t to a discernable listener. Therefore, pay for lossy streaming and download in hi-res (or buy CD/SACD) the rare gems worth listening with a critical ear. TBH not even all classical sounds that bloody amazing as well. So I apply this rule on everything.