Power draw by frequency: what am I missing?

[TLS Guy]

This is probably correct for symphony orchestras and baroque chamber music, but it's not true for modern pop music or some fusion jazz. (In my measurements it's also less true for Beethoven than Mozart, for example, but I digress.) I doubt you listen to Kanye West or Cardi B, but the bass line below 80Hz is powerful and nearly continuous. Fusion jazz often has strong synthesizer bass lines. I'm not a fan of the popular music either, but you and I are in the minority.



The point is not to confuse sensitivity with efficiency, which Swerd explained well, but that's not what Jerry asked initially. I figured he was knowledgeable enough to already know the difference between sensitivity and efficiency. As for the comment about the inefficiency of small sealed subs, true, but how is that relevant to his question?



Again, this is true for symphonies and chamber music, but many action movies (which I generally don't watch either) and pop music genres do have strong bass lines, and with a crossover at 80Hz, like most AVRs use, I'm certain powered subs of whatever cabinet strategy are indeed off-loading the primary amplifiers to some significant degree, and it's not "absolute nonsense". I had a few people who want to hear what their music sounds like on my primary system, and out of curiosity I've fired up my RTA software, so I'm not guessing about what's on a Kanye West album.

Again I think there are confounding issues here. I do master CDs from people from the pop culture from time to time, and the spectroscope shows no huge bass energy except for one case I remember.

However I think perversely subs may very well increase the load on a receiver.

If you have no sub and run a tower or bookshelf full range, the power is going to be excursion limited and sounds of distress from driver over excursion will be manifest and the volume knob turned down.

When you add a sub the drivers are offloaded, so that the speakers will be able to play much louder above sub range, were there is lots of power. So my suspicion is that a sub in many cases allows the user to make the receiver or power amps work harder.

 

[Irvrobinson]

Again I think there are confounding issues here. I do master CDs from people from the pop culture from time to time, and the spectroscope shows no huge bass energy except for one case I remember.

Oh really? Bluegrass?

I'll tell you what, stream some hip-hop or rap music from the internet (I'm an Amazon Prime Music member, so it doesn't cost me anything incrementally), and measure. Most of this stuff is mixed to be played in cars, and sub-80Hz energy is virtually non-stop. Unsurprisingly, most of the bass I measure and hear isn't James Larson subwoofer candy class stuff, but 40-80Hz stuff. Some albums have sub-30Hz energy, but that seems rare. Nonetheless, for most systems with an 80Hz crossover, 40Hz sounds will be 24db down from 80Hz levels.

I think perversely subs may very well increase the load on a receiver.

If you have no sub and run a tower or bookshelf full range, the power is going to be excursion limited and sounds of distress from driver over excursion will be manifest and the volume knob turned down.

I can see where this might be true for 2-way or small 3-way speakers, but it is unlikely to be true for larger 3-way or 4-way towers.

When you add a sub the drivers are offloaded, so that the speakers will be able to play much louder above sub range, where there is lots of power. So my suspicion is that a sub in many cases allows the user to make the receiver or power amps work harder.

Pure conjecture, and mostly based on your assessment of bass energy in symphonies and chamber music.

 

[JerryLove]

I chose that B&W 704 as an example mainly because it's two impedance graphs showed how it could be harder to drive in the bass frequencies below 300 Hz. I was surprised to see the rather large suckout in frequency response between 1 & 4 kHz, centered a bit higher than 2 kHz. This is quite a lot more than the usual polite BBC dip.

I do agree, based on my limited experience, that when speakers become difficult (in terms of swinging ohm values, for example) that does tend to show up in LF.

My take-away from that is "If your speaker is a difficult load, and your amp isn't up to the challenge, you'll most likely experience that in LF"; which is exactly what I ran into with my 801Ns.

Though I can see some similarities to my OP, that would be a separate issue none-the-less.

The frequency response graph you posted looks remarkably flat. Did you produce that graph?

I did not. That would be sourced from a once-prolific poster here who was a true believer in inert cabinets and controlled off-axis performance; as well as the joys of active crossovers.

How did you eliminate the floor bounce peaks & valleys?

I don't know; but I believe "outside on a pedestal" might have been involved.

 

[PENG]

I agree with irv, using powered subwoofers and XO higher at 80 Hz and above does take some burden off the power amp, but obviously it is contents dependent. Based on voltage and current measurements I took in the past when playing action movies, the differences in all cases were significant at 80 Hz XO and higher, not like 50% or even 30% iirc, but definitely significant. Even for classical music, one can definite notice the drop in the amp output for the same spl if the music has a lot of LF contents, 1812 overture for example. Speakers that has low impedance and large phase angles are tougher on amps, and as such, powered subs should help even more. I don't understand why we seem to be debating something that is real common sense.

 

[mtrycrafts]

... if the music has a lot of LF contents, 1812 overture for example. ...

Boy, you said it. I have the TELARC version of it, authentic period cannons recorded separately. The sub amps that has a meter has no problem hitting 300 Watts while the rest is less than 10 watts.

 

[Verdinut]

There's also the drum in the "Dies irae" from the Verdi Requiem on the Telarc label as well.

How about the "Lacrimosa" from Berlioz Grande Messe des morts (Requiem) ? We are really due to have a new release of Berlioz Requiem on Blu-ray!

 

[JerryLove]

Because common sense is rarely right?

Of course moving some of the sound production from amp #1 to amp #2 means a lower demand on amp #1. You could do the same thing with bi-amping. But again, that would simply suggest that amp #1 wasn't up to the task in question anyway.

Yes. The classical piece notorious for the use of cannon as a musical instrument and widely understood to be incredibly demanding of LF output is incredibly demanding of LF output. Meanwhile: Israel "IZ" Kamakawiwoʻole's "Somewhere Over the Rainbow" doesn't appear to drop below 80Hz at all.

Let's imagine that the 1-2 octaves in question did consume 30% of the overall power consumption. That would lower the ability of the amp at that moment by, what? 1-2db?

 

[PENG]

Because common sense is rarely right?

Of course moving some of the sound production from amp #1 to amp #2 means a lower demand on amp #1. You could do the same thing with bi-amping. But again, that would simply suggest that amp #1 wasn't up to the task in question anyway.

Yes. The classical piece notorious for the use of cannon as a musical instrument and widely understood to be incredibly demanding of LF output is incredibly demanding of LF output. Meanwhile: Israel "IZ" Kamakawiwoʻole's "Somewhere Over the Rainbow" doesn't appear to drop below 80Hz at all.

Let's imagine that the 1-2 octaves in question did consume 30% of the overall power consumption. That would lower the ability of the amp at that moment by, what? 1-2db?

Right, it depends, so blanket statements can be misleading..

 

[Verdinut]

Because common sense is rarely right?
Let's imagine that the 1-2 octaves in question did consume 30% of the overall power consumption. That would lower the ability of the amp at that moment by, what? 1-2db?

Just perhaps enough to make it clip!

 

[JerryLove]

Just perhaps enough to make it clip!

The odds of that seem low. *

It's basically the argument in favor of bi-amping.

*note: I'm not saying clipping is or is not rare. I'm suggesting that the odds that your amp is within 2db of it's clipping range is relatively low. I think anything less than a 2x power factor is ignorable.

 

[Verdinut]

The odds of that seem low. *

It's basically the argument in favor of bi-amping.

*note: I'm not saying clipping is or is not rare. I'm suggesting that the odds that your amp is within 2db of it's clipping range is relatively low. I think anything less than a 2x power factor is ignorable.

You were referring to the possibility of 30% of the amplifier's power being drawn by the low frequencies on some peaks, but the possibility is close to more than 75% being drawn for the lows in some recordings which represents more than 6dB.

Note that with classical music, the dynamic range is somettimes way wider than any pop or action film recording. As an example, one audioholic reported a dynamic range of 35dB on an opera recording. That would mean a power demand increase of over 3, 000 times that of the lowest sound level.

 

[JerryLove]

You were referring to the possibility of 30% of the amplifier's power being drawn by the low frequencies on some peaks, but the possibility is close to more than 75% being drawn for the lows in some recordings which represents more than 6dB.

This is moving the goal-posts. I was responding to Peng's statement:
"I took in the past when playing action movies, the differences in all cases were significant at 80 Hz XO and higher, not like 50% or even 30% iirc, but definitely significant. "

Now you would like to make a new claim. Your new claim is that LF peaks draw 3x as much power has peaks everywhere else on the audio spectrum (75% of total draw).

That circles back to my original post. Can you support the claim that LF peaks 6db higher than all other frequencies?

Note that with classical music, the dynamic range is somettimes way wider than any pop or action film recording. As an example, one audioholic reported a dynamic range of 35dB on an opera recording. That would mean a power demand increase of over 3, 000 times that of the lowest sound level.

Unless hearing loss is your goal, that means that the quiet moments are more quiet.

Dynamic range is irrelevant to peak power consumption. Assuming sensitivity is fixed, only peak volume matters. Can you support that the area between a typical sub crossover and roll-off on reasonably typical speakers has peaks "more than 6db" higher than any other point in the spectrum? Not on some specific piece (like that 1812 overture) but as a rule.

 

[Verdinut]

This is moving the goal-posts. I was responding to Peng's statement:
"I took in the past when playing action movies, the differences in all cases were significant at 80 Hz XO and higher, not like 50% or even 30% iirc, but definitely significant. "

Now you would like to make a new claim. Your new claim is that LF peaks draw 3x as much power has peaks everywhere else on the audio spectrum (75% of total draw).

That circles back to my original post. Can you support the claim that LF peaks 6db higher than all other frequencies?


Unless hearing loss is your goal, that means that the quiet moments are more quiet.

Dynamic range is irrelevant to peak power consumption. Assuming sensitivity is fixed, only peak volume matters. Can you support that the area between a typical sub crossover and roll-off on reasonably typical speakers has peaks "more than 6db" higher than any other point in the spectrum? Not on some specific piece (like that 1812 overture) but as a rule.

I would refer you to my Post #26 above for examples. The bass frequencies in those classical work sequences take a lot more than 6db over that demanded by the mid and high frequencies. I didn't measure them but I would bet that they take at least 90% of all the amplifiers' power on peaks.

As for Dynamic range, I disagree with you. Peak power is exactly dependent on dynamic range demands and the power supply of the amplifier has to have sufficient rail voltage, amperage and capacitance for the headroom to meet that demand.

 

[JerryLove]

I would refer you to my Post #26 above for examples.

Can you support the claim, not on some specific piece (like that 1812 overture),but as a rule?

I can just as easily point at my previous "Somewhere Over the Rainbow" reference to show that LF is practically non-existent.

Indeed: this peak comparison kind-of needs to be cross-song/movie... at least for the vast majority of material that one is likely to listen to; otherwise we have "your amp is too small" as the underlying problem. No point fixing "the 1812 overture" if the same SPL is hit by a cymbal in some other popular piece where the same power is needed, but at a different frequency.

The bass frequencies in those classical work sequences take a lot more than 6db over that demanded by the mid and high frequencies. I didn't measure them but I would bet that they take at least 90% of all the amplifiers' power on peaks.

If you have not measured them, how can you be sure that there are no peaks above 80Hz within 10db of the peaks from 40-80Hz?

As for Dynamic range, I disagree with you. Peak power is exactly dependent on dynamic range demands and the power supply of the amplifier has to have sufficient rail voltage, amperage and capacitance for the headroom to meet that demand.

OK. Give me the math.

I have two songs. Song A has a 50db range and Song B has a 100db range. I play both songs such that the peak volume is 90db @1m on my speakers (which are 90db @1m @1w sensitive). [if it matters: assume they are super-flat ohm loads at 4Ohm with -3db roll-offs at 40Hz and 25KHz].

If needed: you can assume they are the exact same recording and I've used dynamic compression to reduce the dynamic range on one.

How much more rail voltage, amperage, and capacitance will I need for Song B than Song A?

 

[mtrycrafts]

... No point fixing "the 1812 overture" if the same SPL is hit by a cymbal in some other popular piece where the same power is needed, but at a different frequency.
...

I don't think you could stand 300 watts going to the cymbal section in your room.

 

[mtrycrafts]

...


OK. Give me the math.

I have two songs. Song A has a 50db range and Song B has a 100db range. I play both songs such that the peak volume is 90db @1m on my speakers (which are 90db @1m @1w sensitive). [if it matters: assume they are super-flat ohm loads at 4Ohm with -3db roll-offs at 40Hz and 25KHz].

If needed: you can assume they are the exact same recording and I've used dynamic compression to reduce the dynamic range on one.

How much more rail voltage, amperage, and capacitance will I need for Song B than Song A?

Same. You postulated 90 dB peaks, the high point in the range. So, one is at 40 dB for the quietest passage, the other is -10 dB.

 

[JerryLove]

I don't think you could stand 300 watts going to the cymbal section in your room.

That would depend on a number of factors, including speaker sensitivity etc.

There's got to be data on peaks across the spectrum for common examples somewhere... I'm not sure how to best phrase that google query.

Same. You postulated 90 dB peaks, the high point in the range. So, one is at 40 dB for the quietest passage, the other is -10 dB.

I agree. But Verdinut has asserted that we need to take dynamic range into account when determining peak amp performance. So that question would be at him or someone who supports his assertion.

 

[PENG]

That would depend on a number of factors, including speaker sensitivity etc.

There's got to be data on peaks across the spectrum for common examples somewhere... I'm not sure how to best phrase that google query.


I agree. But Verdinut has asserted that we need to take dynamic range into account when determining peak amp performance. So that question would be at him or someone who supports his assertion.

As I said many times about many times about power requirements, it depends, and I have no problem with any claims, but not general/blanket statements as though such claims were facts.

Let me try to phrase that Google query and check my book marks.

 

[PENG]

First attempt to Google on dynamic range, music genre and frequency spectrum resulted in the following article that @JerryLove and @Verdinut may be interested in reading.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4753356/

 

[PENG]

Here's a huge data base (nothing about frequency, just DNR) that people uploaded so I assume we need to take them with a grain of salt. If I remember right the link to this database had been posted before so many of you might have visited the site already.

http://dr.loudness-war.info/

It looks like the 20 dB peak THX requirement for movies would work for music as well. One of my 1812 CD has 18 dB, Symphony fantastic 20 dB. I scanned more than 30 pages and only saw 20 and 21 dB a few times, all orchestral classical. There were some non classical ones with high numbers too.