Post by Slinger on Jul 16, 2020 16:15:06 GMT
...of amplifier power?
Right, first the caveats, and a brief explanation. I received one of my regular emails from Graham Slee today - I'm assuming everyone knows of Graham Slee's audio kit? - but this one is a bit different. He wrote it for "fun," and as I enjoyed it I asked his permission to repost the email here. This was his reply:
You've been warned. 
Please don't waste your time telling me that "this" is wrong, because "that" isn't how "whatever" works. It was written for fun.
Slinger
Right, first the caveats, and a brief explanation. I received one of my regular emails from Graham Slee today - I'm assuming everyone knows of Graham Slee's audio kit? - but this one is a bit different. He wrote it for "fun," and as I enjoyed it I asked his permission to repost the email here. This was his reply:
Hi Paul,
It should be explained that I tried to write this in a non-technical way, and I'm no teacher, so if it sounds amateurish in any way, that's the reason.
I could see the story heading towards the ridiculous, but decided to continue to the "punch-line" anyway.
I hope to follow it up with another "amusing proof", if not more.
I hope not to be trolled. OK, provided you clarify the above, go on then.
Best wishes,
Graham
It should be explained that I tried to write this in a non-technical way, and I'm no teacher, so if it sounds amateurish in any way, that's the reason.
I could see the story heading towards the ridiculous, but decided to continue to the "punch-line" anyway.
I hope to follow it up with another "amusing proof", if not more.
I hope not to be trolled. OK, provided you clarify the above, go on then.
Best wishes,
Graham
To what extent are musical dynamics a function of amplifier power?
(I keep getting asked questions like this, with the comment "please explain in a non-technical way." The problem is, it's a technical subject!)
Dynamic range is the size of the space in which the music can exist without its extremes becoming distorted (or hidden by noise).
The packaging within which it is contained might be vinyl, compact disc, or some digital file.
There is a top beyond which the dynamic range will not go, and this must include transients. In vinyl, this is the limit of the tracking ability - about 5 times the maximum recorded level.
Some CDs and some digital files claim 10 times, but this cannot exist above the top, so the maximum recorded level has to be a tenth below the top.
Unfortunately, power has a square law. Its quantity doesn't double or half in the same way as voltage or current. The equivalent in power is 4 times, or a quarter, depending on if it's up or down.
We have to choose a mathematical function to express its range, such that it works with all units. That function turns out to be a law of nature anyway. Its use also relates to the way we hear loudness (or, in reverse, softness).
The unit is the decibel (dB).
This is where the engineer tends to lose the layman's attention because it is mathematics. We are naturally averse to mathematics (I know I am).
To work out any problem, we need to relate quantities, so we need to understand these quantities. By choosing the decibel, we can look at the research to understand the number of decibels we require.
It helps, therefore, to know things like the quietest number of decibels we can hear. For example, a tranquil room isn't completely quiet.
The noise might be 20dB (0dB being silent, and 1dB being the smallest change we can detect). Because dB's are a ratio, interchangeable to some extent, we must give these dB's a qualification. We are dealing with Sound Pressure Levels right now, so we call them dB SPL.
The loudest sounds we can stand comfortably are about 120 dB SPL, after which it becomes "painfully noisy."
To be comfortable, a dynamic range need not exceed 100 dB. Still, being purist, we wish to include sounds obliterated by the noise in our listening space - sounds which may not exist - so we demand 120 dB. 24-bit audio gives us 20 bits to cover this dynamic range. Each bit covers 6 dB, and 20 x 6 = 120 dB.
As I said above, some CD's and some digital files claim 10 times for transients, and 10 times equates to 20dB. So, our average listening level becomes 100 dB SPL (120 - 20 = 100).
Let's now take a look at speakers. If we have two speakers for stereo, then we have halved the power if we take one away. Half power is minus 3dB. So from each speaker, we require 97dB.
The speaker "sensitivity" is rated for a listening distance of 1 metre. Still, if our speakers are set apart 2 metres, our seating position is also 2 metres away. The sound pressure drops 6dB for each doubling of distance. We need to add 6dB to the 97dB, so we need 103dB per speaker at 1 metre.
If we look at our speaker's specification, we might find their sound pressure for 1 watt at 1 metre is 83dB (for example).
We therefore need to raise the power by 20dB (83 + 20 = 103). In power terms, 10dB is 10 times, so 20dB is 100 times. Therefore, we need 100 times 1 watt to reach 103 dB SPL to do the 100 dB SPL required for "reality" at our seating position. In other words, we need 100 watts, just to do the average maximum.
We now must cover the 20dB in which the transients "live" to take the listening experience to the full dynamic range of 120dB. If you've been following me so far, you'll know that 20dB in power terms is 100 times, and so we need 10,000 watts!
On a 60 amp supply fuse, we might be able to draw 14.4 kW just, but 10 kW per channel might be pushing it?
Perhaps we need to rethink all these "musical dynamics being a function of amplifier power" business? I think somebody might have misunderstood.
Until next time, thanks for reading.
Graham
PS. If the above appears a little "tongue in cheek", it's intentional
Copyright © 2020 HiFi System Components Ltd, All rights reserved.
(I keep getting asked questions like this, with the comment "please explain in a non-technical way." The problem is, it's a technical subject!)
Dynamic range is the size of the space in which the music can exist without its extremes becoming distorted (or hidden by noise).
The packaging within which it is contained might be vinyl, compact disc, or some digital file.
There is a top beyond which the dynamic range will not go, and this must include transients. In vinyl, this is the limit of the tracking ability - about 5 times the maximum recorded level.
Some CDs and some digital files claim 10 times, but this cannot exist above the top, so the maximum recorded level has to be a tenth below the top.
Unfortunately, power has a square law. Its quantity doesn't double or half in the same way as voltage or current. The equivalent in power is 4 times, or a quarter, depending on if it's up or down.
We have to choose a mathematical function to express its range, such that it works with all units. That function turns out to be a law of nature anyway. Its use also relates to the way we hear loudness (or, in reverse, softness).
The unit is the decibel (dB).
This is where the engineer tends to lose the layman's attention because it is mathematics. We are naturally averse to mathematics (I know I am).
To work out any problem, we need to relate quantities, so we need to understand these quantities. By choosing the decibel, we can look at the research to understand the number of decibels we require.
It helps, therefore, to know things like the quietest number of decibels we can hear. For example, a tranquil room isn't completely quiet.
The noise might be 20dB (0dB being silent, and 1dB being the smallest change we can detect). Because dB's are a ratio, interchangeable to some extent, we must give these dB's a qualification. We are dealing with Sound Pressure Levels right now, so we call them dB SPL.
The loudest sounds we can stand comfortably are about 120 dB SPL, after which it becomes "painfully noisy."
To be comfortable, a dynamic range need not exceed 100 dB. Still, being purist, we wish to include sounds obliterated by the noise in our listening space - sounds which may not exist - so we demand 120 dB. 24-bit audio gives us 20 bits to cover this dynamic range. Each bit covers 6 dB, and 20 x 6 = 120 dB.
As I said above, some CD's and some digital files claim 10 times for transients, and 10 times equates to 20dB. So, our average listening level becomes 100 dB SPL (120 - 20 = 100).
Let's now take a look at speakers. If we have two speakers for stereo, then we have halved the power if we take one away. Half power is minus 3dB. So from each speaker, we require 97dB.
The speaker "sensitivity" is rated for a listening distance of 1 metre. Still, if our speakers are set apart 2 metres, our seating position is also 2 metres away. The sound pressure drops 6dB for each doubling of distance. We need to add 6dB to the 97dB, so we need 103dB per speaker at 1 metre.
If we look at our speaker's specification, we might find their sound pressure for 1 watt at 1 metre is 83dB (for example).
We therefore need to raise the power by 20dB (83 + 20 = 103). In power terms, 10dB is 10 times, so 20dB is 100 times. Therefore, we need 100 times 1 watt to reach 103 dB SPL to do the 100 dB SPL required for "reality" at our seating position. In other words, we need 100 watts, just to do the average maximum.
We now must cover the 20dB in which the transients "live" to take the listening experience to the full dynamic range of 120dB. If you've been following me so far, you'll know that 20dB in power terms is 100 times, and so we need 10,000 watts!
On a 60 amp supply fuse, we might be able to draw 14.4 kW just, but 10 kW per channel might be pushing it?
Perhaps we need to rethink all these "musical dynamics being a function of amplifier power" business? I think somebody might have misunderstood.
Until next time, thanks for reading.
Graham
PS. If the above appears a little "tongue in cheek", it's intentional
Copyright © 2020 HiFi System Components Ltd, All rights reserved.
Please don't waste your time telling me that "this" is wrong, because "that" isn't how "whatever" works. It was written for fun.
Slinger
