How much amplifier power do you realy need?

[Confused]

I found the video per the link below pretty much by accident when searching for something rather different.  I have to be honest, I found this to be utterly gobsmacking, I had always suspected that amps needed a little more power than simple maths based on required dB(A) and speaker efficiency would suggest, but nothing like this.  The video of the amp's display is a little blurry, but's lets just say that I has mentally parked the decimal point in the display a factor of 10 backwards until it dawned on me what was really going on.

I would say it is well worth 10 minutes of your time to watch.

https://www.youtube.com/watch?v=bRMR9JZ1m0s&app=desktop

So what exactly are the amplifiers measuring?  As mentioned in the video it is the peak / tranient power that is captured, so not a continuous average.   The amps are CH Precision A1's.  This from the CH Precision manual:

Power monitoring. Each power amplifier board is equipped with a DSP that monitors the instantaneous output voltage and current of each M1 channel. Both values are sampled at around 100 kHz, ensuring peak values are properly detected. This circuitry has several purposes: give the user a feedback of the peak power fed to the loudspeakers, and detect malfunctions such as short-circuits or amplifier damage.

As many will know, Harbeth's are not the most efficient speakers you can buy, in fact rather the opposite, the 40.1's are rated at a lowly 85dB/w/1m.

This is another video showing the CH Precision power meters.  It is not clear what speakers are used here, but the peak levels are clearly lower than are seen on the Harbeth's.  (but still hit the odd spectacular peak)

https://www.youtube.com/watch?v=UP8i8F62OlM

I have to say that when I first viewed the Harbeth clip I was staggered by what I was seeing, this was not what I expected at all.  Thinking about it though, it does make sense of some of my past experiences.  For example, I first listened to the KEF Blades with a Devialet Expert D240.  This appeared to drive the Blades with ease, which is what you might expect.  Later, I heard the Blades with a D170.  It was terrible.  This was not some subtle step back in performance, but a case of the amp not really working with the speaker, dynamics vanished, the bass was a mess.  Not good to listen to at all.  I remember being puzzled by this at the time.  The Blades are rated with a sensitivity of 91dB/w/1m.  So for normal listening levels surely you only need 10 or 20 Watts or something?  Why should the drop from 240 to 170 make any difference?  Based on simple maths re dB(A) levels and power, it did not make sense.  However, looking at those figures that the CH Precision amps are displaying, it would seam possible that the D240 was coping with the Blades, but the D170 struggling, with audible consequences.

It also reminds me a Devialet event at Oxford Audio, when Devialet's Chief designer Mathieu Pernot stated that his perfect amplifier design would have "infinite power".  This seamed slightly absurd to me at the time and I was talking to him about this later in the day.  I was making a real world point about the rated power capability of typical speakers.  In response Mathieu pointed out that there is no agreed method or protocol for measuring manufacturer quoted speaker power ratings, no agreed standard if you like, so manufacturers tend to give figures that are more like "recommended amplifier power ratings.  He then went on to explain that a speaker "rated" for say 50w to 400w, could easily take instantaneous peaks of power way over the 400w "maximum" with ease.  OK, try putting that kind of wattage through continuously and heat will generate, voice coils will melt or seize, but very high transient peak power levels are not an issue.  For me, knowing (based on maths) that a typical speaker could run at well over 100dB(a) with less than 20W, Mathieu's explanation seamed a little theoretical.  Now I have seen the above video, it makes perfect sense.  I have always though that higher power amplifiers provide a certain sense of "ease" to the presentation, but I could not really rationalise the science behind what was just a subjective view.
  
Now I have seen the above video, many things make a lot more sense. 

For those interested, this is a link to the track used in the Harbeth video:

https://www.youtube.com/watch?v=XhuJxdaU87I

Playing this at home, I would have to say that it does include some rather extreme bass content, which combined with the Harbeth's low efficiency does perhaps go some way towards explaining the rather high power figures on that CI amp.  I can also see that my habit of using one or two electronic tracks when auditioning kit might have some merit, I can't see any other music genres providing quite the speaker / amplifier challenge of something like the Pan Sonic track linked above.  (plus, I quite like electronic music) 

Although even with miserable YouTube 128kbs streaming, it does sound pretty awesome with my little Devialet providing the thundering bass power.  

[Confused]

14 hours ago, audiobomber said:

IMO this is a simplistic and deceptive video. Watts per channel is next to useless as a measure of an amp's ability to drive a loudspeaker. Current is what really matters, not voltage. I will forever recall hearing a 40W amp blow away a 200W amp in a store demo decades ago. A real eye-opener.

 

In the video they strapped the amp channels. Monoblocking is for sound reinforcement, not for home stereo; lots of voltage, but compromised current capability, thus inferior SQ. If they had vertically biamped instead, the sound would have improved by leaps and bounds, and the amps would still not have run out of steam.  

 

How about the infamous Apogee Scintilla, which only a handful of amps could handle? One of the most famous pairings was the 25wpc Classe. Most high-powered amps totally crapped out with this 1-ohm load.

 

I have two stereo amps from the same manufacturer rated at 50Wpc into 8 ohms. My speakers are 89dB/w/m, and 4 ohms. Driving the speakers with one stereo amp sounds fine, even at the highest levels I care to listen. Adding the second amp in vertical biamp mode increases sound quality markedly, with improved bass control, soundstage, instrument separation and detail. Watts per channel does not explain the difference in these two setups.

So how were the amps delivering 750 watts if they were not capable of producing current?  For a given voltage watts and amps (current) are directly proportional to each other.  Don't get me wrong, I fully agree that current matters, and that "wpc" ratings are often highly misleading, but @mansr is right, there are simple relationships between current, impedance and power. (and volts for that matter)  If you have lots of current, you will by definition have lots of watts, assuming you have some volts too of course.

 

The point being, I agree with you re current, and bi-amping can be a better option than bridging, depending on the speakers impedance characteristics, but the video still has some interest I think, in that the amps are showing some surprisingly high power figures, in what seamed a fairly normal "real world" situation.

[Confused]

As an open question, how does amplifier power, current demand and so on relate to the dB "A" weighted scale?

 

The Laptev Sea track used in the video is has a lot of low bass content, right down to frequencies more felt that heard, so not just below 100Hz, but dipping below 20Hz.  Looking at the a weighted scale (see graph below), these low frequencies "sound" -30dB down to human hearing at say 40Hz, even more as you drop to 30 or 20Hz.  -30dB is a lot!  So if you take the power needed to reproduce 1000Hz at 90dB(a), then you would need fully 1000 times the amplifier power for an equally loud sounding sound at 40Hz.  I am thinking aloud a little bit here, but could this be a factor in what we are seeing in the video?

 

Or to put this another way, it is easy to say I have speakers with an efficiency of 90dB 1 w / 1 m, so I do not need more than say 10 w for my normal listening level of 90dB at 3m, but if the bass frequencies need an order of magnitude more power to reach 90dB(a), how exactly does this work?

 

 

[Confused]

3 hours ago, audiobomber said:

Double the output voltage is available in monoblock mode, current capability is halved because the amp sees half the load impedance. An 8 ohm speaker load becomes 4 ohms, and if it dips to say 3 ohms in the bass like the Nautilus 802, the amps see 1.5 ohms at those frequencies. This is a punishing load, and most amps will not be able to supply sufficient current to deliver full power.

 

Audio Engineer's Reference Book 2.8.3 (2)

"The amplifier may run out of ability to provide the highest peak instantaneous output currents required, even though it may have something in hand for peak voltages. This is called current clipping."

 https://books.google.ca/books?id=XOvf30iChsYC&pg=SA2-PA118&lpg=SA2-PA118&dq=amplifier+"current+clipping"&source=bl&ots=DdxLBLbZWb&sig=ACfU3U3y7TE1ekpKonvy99i6-bsCr7EbBg&hl=en&sa=X&ved=2ahUKEwiJ993K3LTjAhVGOs0KHZCDCRcQ6AEwB3oECAkQAQ#v=onepage&q=amplifier "current clipping"&f=false

 

I would have liked to see an SPL reading in the video. At one point one of the men is shouting very loudly to the other, from inches away, and has to repeat himself to be heard. Clearly these were ear-damaging exposures that I would avoid. Lowering the level by a few dB would radically decrease the power requirements and make better music. Of course that was not the agenda.

 

Out of interest, how loud would you estimate the SPL’ s to be?   OK, there is much speculation here, but from about the 6:00 min mark in the video you can just about make out the voices, not what they are saying, when the camera and microphone are about three to four metres away from those talking.  After 7:00 min when the camera closes up on one attendees smart phone to show the track name, you can clearly hear his voice and the camera is maybe half a metre away.  At one point, three of the guys are gathering around very close to the right hand speaker, which to me makes me think that that were running sub 100dB, loud but nothing crazy.

 

From this, I would estimate something a little over 90dB, with peaks under 100dB. This is just a guesstimate though, it is very difficult to judge.  I too would have likes to see an SPL reading, it is easy enough to do with a smart phone app or similar, so it is a very disappointing omission.