Power calculator ( Amp , AVR) - How much power do I need ?

[PENG]

Cool. So its essentially similar to AcousticFrontiers, fixing the terminologies we discussed earlier on this thread and making it more explicit .
View attachment 35396

I'm good - what I have been going back & forth with you guys for the past couple of days ;-) , as whats needed to product an accurate estimate of power requirement.

Looks like you're actually asking users to input Room gain in dB and Gain from multiple speakers in dB...rather than Yes/No for Room gain or #speakers and defaulting to certain gain(dB) as in AF & homestead .

Wont that make people scratch their heads ;-) I mean , the guys earlier(pogre, ryanasour) were arguing about incredible complexity and confusion with even asking the user basic inputs like speaker impedence and additional headroom (which are absolutely essential) !!

I'dnt be asking those gains in dB to end-users, but rather derive them , from more simpler questions ( #Speakers, Speakers placement) like what other online calculators do.

Also seems you're mixing both the peak SPL and Amp Power requirement into same - not sure about that , might confuse people,I think. I'd keep them separate, if possible.

Other than that, I'm good.. Now, if gene can agree to you hosting this calculator at AH - that'd be ideal !

Thanks for the feedback, it is not final yet so I can keep modifying to make it less confusing to beginners.

There is another potential confusing or controversial point. Acoustic Frontier seemed to think it is important to point out the build in 3 dB headroom in the peak spl calculators calculates. That is, if the calculated result is 100 W, it actually means 50 W average. I am not comfortable in going along with that thinking so unless someone can convince me, I won't consider it a build in headroom. My thinking is that, it is a matter of how one interprets THX's reference SPL of 85 dB average 105 dB peak. I interpret the 105 dB peak simply means they want to allow for the occasion "peak" SPL in movies that cause the SPL jump from 85 dB average to 105 dB average (note still average, but elevated by 20 dB, for may be a second or a little longer, hence called the "peaks"). Whereas Acoustic Frontier seemed to have interpreted that 105 dB peak as 102 dB average and that the extra 3 dB is simply due to the 2:1 peak to average ratio for a sine wave.

May be I am wrong and he is right, or vice versa, but I think it is better to error on the conservative side. What do you think?

 

[Pogre]

Wont that make people scratch their heads ;-) I mean , the guys earlier(pogre, ryanasour) were arguing about incredible complexity and confusion with even asking the user basic inputs like speaker impedence and additional headroom (which are absolutely essential) !!

I think you're missing our points.

It's doesn't have to be incredibly complex to confuse a newbie.

 

[PENG]

May be my question to AVUser in post#41 is not clear and I should add the following background info in the linked article (wrapped in quote)

Amplifier power / SPL calculator for home theater THX reference level

A question that frequently comes up when designing a home theater is 'how powerful an amplifier do I need?'. The answer to the question depends on a few things, in particular how sensitive your speakers are and how far you sit away from them. When we design a home theater we want it to meet...

www.acousticfrontiers.com

Finally we see the Equivalent RMS Watts which is roughly half the Peak Amplifier Watts. The reason for this conversion is that all amplifiers spec sheets list RMS Watts, as measured with a sine wave. A sine wave has a 3dB crest factor i.e. the difference between average and peak is 3dB. So we effectively derate the calculated Peak Amplifier Watts by 3dB to get the equivalent RMS specification. This is an important point, and nearly everyone we hear discussing this topic online fails to derate the amplifier power requirements. They make the mistake of equating peak SPL with RMS Watts. In the above example an amplifier with a specified RMS output of 683W would actually give us 108dB peaks at the listening position, which is 3dB more than we need. People using this approach will choose an amplifier roughly twice as powerful as they need.

So clearly the writer was counting on the 3 dB crest factor, the peak to average ratio of a sine wave that represents Power, the product of voltage and current to make his point about the 3 dB build in headroom that he thought we did not know about when using some of those online calculators such as Crown Audio's.

My question is, is the THX reference level of 85 dB + 20 dB peak based on the peak of the sine wave, or the average?

Before we answer the question, I suggest we consider at least the following points:

1). When we measure spl with a spl meter such as Radio Shack popular 33-2055, we set it to "C" slow right, so won't we be measuring the average? Likewise, when our AVR/AVP/AVC are calibrated to produce the reference 85 dB at volume "0" from our main mic position, that 85 dB would have to be "average", not "peak" right?

2). Now if we add 20 dB on top to allow for the "peak", wouldn't that "peak" literally mean the spl recorded during the highest level measurable by a spl meter or program/mic in the same way, that is, based on something like "C" weighting slow? In other words, we would still be talking "average" over a second or so, and that 20 dB "peak" simply elevate the average (over say 1 second +/-...) from 85 dB average to 105 dB average, sort of saying 85 dB is the prolonged average spl while 105 dB is the short duration/or "peak average"?? Yes I know "peak average" sounds very weird, but I use for lack of a better expression that I can think of right at this moment.

3). "peaks" in movies are likely set by the low frequencies often produced by subwoofers capable of 115 dB at 1 meter distance anechoic. If we look at the 80 Hz sine wave, the crest peak would happen twice (positive and negative) per cycle and the duration of one complete cycle = 1/80 or 0.0125 second. That is, if we are counting on the amplifier's output voltage sine wave peak, you will get the 105 dB spl for a very tiny fraction of the 0.025 second duration that is for the whole ONE cycle of the wave!!

So it seems to me THX's 20 dB peak likely means the average of at least one half cycle of a sine wave during the maximum level recorded in a movie, but not the 3 dB peak of that same one sine wave that last for a few milliseconds. In fact, I doubt any of us has a spl meter that can measure that 3 dB peak resulted from the crest top of a sine wave. "C" weighting fast apparently is not faster than 0.125 second, that's not even fast enough to capture the average of one single 80 Hz sine wave.

I search the THX website extensively for a clear definition and measurement methodology for the 105 dB peak SPL to no avail, but I did find something sort of related under the heading TASA certification. See below:

THX FAQ

Frequently asked questions and answers for everything THX related.

www.thx.com

To receive TASA compliance certification, the final measurement must not exceed the TASA upper volume limit of 85 Leq(m). Leq(m) measures decibels over time, creating an average, whole number.

Am I out to lunch on this, regarding AF's claim I quoted above, or may be he is?

And I would appreciate it if someone can provide a link to the methodology for measuring the THX required 85 dB SPL and 20 dB peak. Without that, all we know from their website is something like:

THX Certified Receivers reproduce studio Reference Level, 85dB SPL with 20dB of headroom.

 

[AVUser001]

I think you're missing our points.

It's doesn't have to be incredibly complex to confuse a newbie.

Not missing at all. Think you're missing my points.

This is exactly what I was highlighting to PENG as feedback to the revised excel calculator he came up, thats essentially same/similar to AF....but with potentially added complexity and technical questions like Room gain in DB and Speaker gain in DB etc...more technical questions than either homestead or AF or Crown,which will confuse the newbie ( and some here) He's also combined SPL and Peak calculator into one...

and yet you guys maintain silence when he asked for feedback ;-) Cant be too defensive , when presented with constructive criticism ;-)

See my suggestions in the thread to simplify that further, but still retaining the essential questions.

But keep in mind, being too simplistic and missing critical inputs will lead to wrong reults , plain and simple. The Crown calculator , which I linked earlier , is a lot better than homestead,..but stil misses the essential impedence input . Something is not better than nothing , always, esp when it leads to wrong results.

this is exacly what PENG is doing with the calculator - asking the right essential questions to arrive at the right answer. Not assuming and misleading like the homestead.

 

[AVUser001]

May be my question to AVUser in post#41 is not clear and I should add the following background info in the linked article (wrapped in quote)

Amplifier power / SPL calculator for home theater THX reference level

A question that frequently comes up when designing a home theater is 'how powerful an amplifier do I need?'. The answer to the question depends on a few things, in particular how sensitive your speakers are and how far you sit away from them. When we design a home theater we want it to meet...

www.acousticfrontiers.com

So clearly the writer was counting on the 3 dB crest factor, the peak to average ratio of a sine wave that represents Power, the product of voltage and current to make his point about the 3 dB build in headroom that he thought we did not know about when using some of those online calculators such as Crown Audio's.

My question is, is the THX reference level of 85 dB + 20 dB peak based on the peak of the sine wave, or the average?

Before we answer the question, I suggest we consider at least the following points:

1). When we measure spl with a spl meter such as Radio Shack popular 33-2055, we set it to "C" slow right, so won't we be measuring the average? Likewise, when our AVR/AVP/AVC are calibrated to produce the reference 85 dB at volume "0" from our main mic position, that 85 dB would have to be "average", not "peak" right?

2). Now if we add 20 dB on top to allow for the "peak", wouldn't that "peak" literally mean the spl recorded during the highest level measurable by a spl meter or program/mic in the same way, that is, based on something like "C" weighting slow? In other words, we would still be talking "average" over a second or so, and that 20 dB "peak" simply elevate the average (over say 1 second +/-...) from 85 dB average to 105 dB average, sort of saying 85 dB is the prolonged average spl while 105 dB is the short duration/or "peak average"?? Yes I know "peak average" sounds very weird, but I use for lack of a better expression that I can think of right at this moment.

3). "peaks" in movies are likely set by the low frequencies often produced by subwoofers capable of 115 dB at 1 meter distance anechoic. If we look at the 80 Hz sine wave, the crest peak would happen twice (positive and negative) per cycle and the duration of one complete cycle = 1/80 or 0.0125 second. That is, if we are counting on the amplifier's output voltage sine wave peak, you will get the 105 dB spl for a very tiny fraction of the 0.025 second duration that is for the whole ONE cycle of the wave!!

So it seems to me THX's 20 dB peak likely means the average of at least one half cycle of a sine wave during the maximum level recorded in a movie, but not the 3 dB peak of that same one sine wave that last for a few milliseconds. In fact, I doubt any of us has a spl meter that can measure that 3 dB peak resulted from the crest top of a sine wave. "C" weighting fast apparently is not faster than 0.125 second, that's not even fast enough to capture the average of one single 80 Hz sine wave.

I search the THX website extensively for a clear definition and measurement methodology for the 105 dB peak SPL to no avail, but I did find something sort of related under the heading TASA certification. See below:

THX FAQ

Frequently asked questions and answers for everything THX related.

www.thx.com

To receive TASA compliance certification, the final measurement must not exceed the TASA upper volume limit of 85 Leq(m). Leq(m) measures decibels over time, creating an average, whole number.

Am I out to lunch on this, regarding AF's claim I quoted above, or may be he is?

And I would appreciate it if someone can provide a link to the methodology for measuring the THX required 85 dB SPL and 20 dB peak. Without that, all we know from their website is something like:

THX Certified Receivers reproduce studio Reference Level, 85dB SPL with 20dB of headroom.

@PENG I undertood your question and know what you're talking about (ie the confusion ;-)) - I've read his article before..., juggling between meetings.. so I will keep it short,

In short, I think he's mixing up the 3db crest factor of a pure sine wave , with a 20db crest factor for music.

So for a pure sine wave, you have the 3dB crest factor ( difference between the RMS( more appropriately the continous average) and peak, as shown below

but for dynamic music with so many peaks & dips , my intepretation of THX recommended 20db headroom , is actually to account for the 20db crest factor for music.

This is what the Crown calculator does too - without assuming anything - explicitly asks the end-user to enter the headroom(dB) over the avg listening level , as it depends on program material.

So we plan for 20db crest factor for music ..., 85db avg , 105db avg peak (or peak avg ;-)) - Both reference level.

+3db , on top of that - is purely a comfort headroom at the amp, inline with some of the recommendations of 'whatever you think you need , double it" philosophy (you see this in ASR too , based on speaker failures). Dont need this mandatorily - as its going to double the power again [ Did I do it for my amp purchase..you bet I did ;-) - like I said - no compromises when it came to power..,but thats a personal choice ] ...

The amp dynamic peak specs over and on top of the RMS(which we factored 20db anyway), will take care of those fractional seconds anyway..so really not a necesssity to add +3db and double the power again ...., atleast in the calculator. If some wants to maintain a +3db , as additional headroom for safety - thats upto them.

 

[PENG]

So its essentially similar to AcousticFrontiers, fixing the terminologies we discussed earlier on this thread and making it more explicit .

I guess so, though mine is totally transparent and one can compare with just about any calculator including Crown's on the www. except the AF one. The AF one has something hidden, making its calculated values a touch out of wack, but it is in the decimal places so it matters not.

What annoyed me most is his arbitrary use of "average" and "RMS" that went beyond the simple common error of use the wrong terminology. Using the wrong terminology is one thing, more people do it than not anway.. It is annoying because it is so obvious that he knows the math, but didn't bother to try not mix the two up in one short paragraph that even involve the math. If you read that paragraph, I bet you would agree that it is downright confusing, they way he wrote it..

Looks like you're actually asking users to input Room gain in dB and Gain from multiple speakers in dB...rather than Yes/No for Room gain or #speakers and defaulting to certain gain(dB) as in AF & homestead .

Good point, so I removed the multiple speakers gain option. I really don't think it is a good idea anyway because it would be so variable depending on other factors, and besides, THX's reference level is for one speaker at a time so may as well do things by the book. One potential source of confusion eliminated! I may add a note to remind people the calculate SPL is based on just one speaker so in the real world one may expect a few dB louder when watching a movie.

Room gain is a little different, we know its pretty much always there, though it also would vary quite a bit depending on the room, speakers and placement. It also affects much more on the low frequency band and not much to nothing on the high frequencies. To minimize confusion and complexity, I would note that one can enter 0 to 3 dB and will be on the conservative side. Even if someone enters 3 dB, I would be comfortable to say it would most likely error on the safe side.

Also seems you're mixing both the peak SPL and Amp Power requirement into same - not sure about that , might confuse people,I think. I'd keep them separate, if possible.

It wasn't my intention to "mix...", but jut to be inclusive so one doesn't (me initially of course as I have been the only one using it for last few years) have look for another calculator. So I re-organize the presentation to make it almost like a separate one but still on the same sheet.

Other than that, I'm good.. Now, if gene can agree to you hosting this calculator at AH - that'd be ideal !

It is still work in progress, and hoping for further inputs for improvements. In the mean time I will try to find to make it work without Excel, though it should probably work with Google sheet. I am going to send it to my Good phone and Galaxy pad to see if it would work properly.

 

[AVUser001]

Yeah, not really you how he got mixed in and in exactly what way but I am pretty sure that was his own misconception about "3 dB" hidden treasure he alluded to. I wish he was right obviously but not much chance of that at all. By the way, personally I would always recommend 10 dB on top of the calculated value. I know I know that's just me and I can afford it.. That's what I meant by "well below the amplifier's output limit, i.e. at least 10 dB below during peaks, when I use that to qualify my claim that well designed class AB amps (only those that measured well..) would sound the same in a blind comparison test.

Correct, not really you , but he mixed up crest factors of regular sine wave and music ..
Many people think a pure sine wave with normal 3db crest factor applies to music too which has erratic peaks & dips..as much as 20db , which is what THX appropriately factors that we need to incorporate for the recommended power. I know it sounds too complicated, but hopefully my explanation made sense.

Do you really mean, you add +10db on top of +20db peaks that THX recommends (ie a total of +30db headroom) ..that'll be 1000 times more power than your average listening level !! ? Are you sure about that ....
To me, +20db sounds OK- that's 100 times more power than whats needed for average listening level

 

[AVUser001]

What annoyed me most is his arbitrary use of "average" and "RMS" that went beyond the simple common error of use the wrong terminology. Using the wrong terminology is one thing, more people do it than not anway.. It is annoying because it is so obvious that he knows the math, but didn't bother to try not mix the two up in one short paragraph that even involve the math. If you read that paragraph, I bet you would agree that it is downright confusing, they way he wrote it..

Yeah, I understood what he was trying to get at..., if you just use his arrived peak value (without downconverting to so called RMS by subtracting 3db crest factor of a pure sine wave), we'd be fine.., He was off by 3db.

Good point, so I removed the multiple speakers gain option. I really don't think it is a good idea anyway because it would be so variable depending on other factors, and besides, THX's reference level is for one speaker at a time so may as well do things by the book. One potential source of confusion eliminated! I may add a note to remind people the calculate SPL is based on just one speaker so in the real world one may expect a few dB louder when watching a movie.

Yes, just 1 speaker - to be on the safe side and align with THX.

Room gain is a little different, we know its pretty much always there, though it also would vary quite a bit depending on the room, speakers and placement. It also affects much more on the low frequency band and not much to nothing on the high frequencies. To minimize confusion and complexity, I would note that one can enter 0 to 3 dB and will be on the conservative side. Even if someone enters 3 dB, I would be comfortable to say it would most likely error on the safe side.

Yes Room gain needs to be factored in. My point really was not to ask the end-user to estimate the room gain(db) and plug it into the calc. Most will be lost.
I'd simply ask - similar to AF & homestead - more in the layman terms "Do you place them close to wall" . If Yes , add 3db internally ..If no , make it 0. I see AF adds in 6db gain if yes. This will keep it simple.

It wasn't my intention to "mix...", but jut to be inclusive so one doesn't (me initially of course as I have been the only one using it for last few years) have look for another calculator. So I re-organize the presentation to make it almost like a separate one but still on the same sheet.

Cool , yes two different calculators on the same page ( assuming this is going on AH ) ..like how Crown does.

 

[PENG]

Correct, not really you , but he mixed up crest factors of regular sine wave and music ..
Many people think a pure sine wave with normal 3db crest factor applies to music too which has erratic peaks & dips..as much as 20db , which is what THX appropriately factors that we need to incorporate for the recommended power. I know it sounds too complicated, but hopefully my explanation makes sense.

Do you really mean, you add +10db on top of +20db peaks that THX recommends (ie a total of +30db headroom) ..that'll be 1000 times more power than your average listening level !! ? Are you sure about that ....
To me, +20db sounds OK- that's 100 times more power than whats needed for average listening level

Oh no, of course not, not another 10 on top of the already generous 20 lol!! I meant for example, in my room I need less than 0.25 WPC for my listening level so if I want to compare amps, even if I crank it up to 0.5 WPC, the amps will still be running at more than 20 dB below their rated output.

By the way, I don't think the AF gentlemen mixed up the THX requirement/sine wave crest/music crest as such. I think he simply forgot you cannot, and I doubt THX would either, count on that momentarily peak and call it 105 dB peak. I believe THX meant for the 105 dB peak to be the SPL recorded during the movies scenes when the average SPL was at maximum level, hence the term peak, so that "peak" may last a second or even a few seconds but it is the "maximum average" As I said, for lack of better words I knew it would hard to someone to know what I am talking about.

As to the music crest, that's a whole different story, too complicated to even start discussing, let alone debating on that one. Suffice to say for now, that one has to remember, music waves are sine waves, they don't look like it at all obviously only because of the very rich harmonics. If you do a FFT on the wave you can resolve it to a series of sine waves of different harmonics, odd and even, a whole lot of them.. That's what the amplifier would amplify, all those sine waves all at once. May be not as much as you have, but I did research that topic quite in depth. I could still be dead wrong, and if so I can be convinced and admit it if and when I realize it. Until then, let's save this super interesting topic for another day.

And, thanks again for your and others input on a better power/spl calculator that can accept impedance input, very valuable indeed..

 

[AVUser001]

Oh k , you gave me a heart-attack, for a sec and I was thinking something happened to PENG ;-) Ok , +20db it is..and done.

Yes music is still a series of sine waves ..but unilke a perfect "pure" test tone(uniform amplitute,intervals etc) , has a varying crest factor - with different order harmoics - with resultant all over the place, dependent on the recording - thats what I mean. The ratio of peak to avg - crest - can be +5-20db(or more).
like this..

For now, like you said , we're good with the power calculations, with the changes discussed above - unless someone else has any other concerns.

 

[PENG]

I spent some time improving my calculator. Please do me a favor one more time, and see if this newest version is better, now that I believe I have improved it based on inputs from some of you.

I tested it on both Anroid and iPhone, Google sheet and Numbers that are Apps build in to those smart phones can do the job so you don't need Excel.

Changes are made to:
- allow various specified nominal impedance values
- user can choose sensitivity in either dB/W or dB/2.83 V but not both
- allow entering room gain
- allow entering desired extra amp power headroom (on top of the headroom manufacturer specified)
- fields not used (no input data) will remain blank, instead of those error signs
- attempts to enter data where/when it is not allowed will get a pop up error message
- links to some useful articles are included