Amount of power needed for dynamic peaks?

[Duke]

A 200W peak would not be anything like touching the voice coil with a 200W soldering iron.

A woofer's entire voice coil and metal former has considerable thermal inertia and a point source quick (peak) application of heat from a 200W soldering iron sitting at 300 degrees centigrade is a ridiculous analogy.

Please note that I only claimed validity to a "rough first approximation", in that a lot of heat is transferred virtually instantaneously. I do not dispute your more precise description.

That doesn't change your point on compression which is very much a valid point.

Thank you for acknowledging that.

 

[echopraxia]

Another thing I think many people neglect when considering amp power is the weighing of most music content towards the lower frequencies (at least on all but extremely bright music or speakers). 100db of white noise sounds almost unbearably loud (edit: it would kill your speaker and ears), but 100db tilted heavily towards bass is not bad at all.

For example, I was listening to a song fairly loudly just now that sounded very nice and “warm”, but not harsh at all. It was loud to the point where I could feel the impact of the bass drum’s peaks physically impacting through couch. Turning on my SPL meter, and it was “only” 80db! But that was A-weighted. Switching to C-weighting and it suddenly become 90db. But this was a slow moving average. Switching to peak mode, it measured 100db at times (hence the couch shaking).

80db from the perspective of hearing damage (quite safe). Yet >100db of peak power! That‘s a HUGE difference.

So I think a lot of people may also underestimate their power needs (especially if you don’t have subwoofers) because you may either be used to A-weighted SPL (because that more closely models hearing pain and damage) not realizing how much bass power we can painlessly enjoy, or that we also often forget how much headroom we need for dynamic peaks as well.

 

[restorer-john]

100db of white noise sounds almost unbearably loud, but 100db tilted heavily towards bass is not bad at all.

I'd be using pink noise unless you have stocks of spare tweeters.

 

[echopraxia]

I'd be using pink noise unless you have stocks of spare tweeters.

Lol.. yeah I should say I haven’t actually tried that, and never would want to. Shows even more so just how much our power demands tends to concentrate in the lower frequencies.

 

[CDMC]

Probably, I was too bad with words.
What you can do is
1)find out dbtp value
2)find out gains of your amp/preamp/volume control
3)knowing your DAC specifications (manufacturers usually give V @ 0 dBFS) calculate voltage at the speaker.
4)if you don't know gains or specifications you can use a voltmeter - use a sine wave generator (be careful) at 50 Hz and measure Vrms at the speaker input and then you'll know relation between digital level and voltage.

You'll need to be careful with calculations as manufacturers and software makers use different definitions of dbfs.

Ahh, thanks. I can do this.

 

[aac]

Ahh, thanks. I can do this.

https://ffmpeg.org/ffmpeg-filters.html#Examples-137
This command can be used with ffmpeg for finding true peaks
ffmpeg -nostats -i filename.wav -filter_complex ebur128=peak=true -f null -

Also there is this
http://www.cieri.net/Documenti/JBL/Technical Notes/JBL Technical Note - Vol.1, No.16.pdf

 

[jlo]

Ironically this is incorrect, but keeps being restated by authorities that should know better. The reference calibration is not 85db at -20dbfs, but 83 db at -20dbfs IN A LARGE THEATER. It was revised more than 30 years ago.
https://www.digido.com/portfolio-item/level-practices-part-2/

No, SMPTE reference calibration level is 85dB(C) SPL at main listening position for a pink noise at about -22dBFSD (rms 22-22k, SMPTE2095-1) with a 12dB crestfactor.
That means that max level with wideband noise would be about 85+(22-12) =95dB(C) SPL.
The max level for sub is more difficult to assess : RTA level of sub is 10dB above RTA of front speakers that gives about 5dB(C) higher SPL so ref level is near 90dB(C) SPL with reference pink noise. As programs may have very different sounds/bandwidth/... in LFE track, it seems more secure to count with a +20dB (instead of +10dB for front speakers) so max 105dB(C) SPL at listening position.

 

[pjug]

Depends entirely on what you're listening to. Crest factor of highly compressed music (e.g. electronic, rock/metal, modern pop and rap) is generally 10-11dB at maximum.

Highly dynamic music (film scores, high dynamic range classical, jazz, etc) might be more like 15-20dB crest factor, but peak level is often lower.

So, figure peaks will need 10ish times the power of your average level... Factor that in with your speakers' sensitivity and you're in the ballpark. I think unless you're listening at massively loud volumes, 100wpc is probably plenty. My Focal Shape 65s are 80w low and 25w high (they're active, biamped) and can get far too loud to comfortably listen to long before the amps run out of steam.

I am not following how you get the 10ish number. If crest factor is 15-20dB, then the you voltage to keep from clipping follows from this. Then power is proportional to voltage squared. So 10ish would be too low. Also, you can find tracks with crest factors on the order of 25dB, so it depends very much on whether you feel like you need to be able handle these extreme cases.

 

[DonH56]

V = volts, W = watts (power), x = multiplier

x V = 10^(dBV/20)
x W = 10^(dBW/10)

1 dB = 1.122x V = 1.259x W
3 dB = 1.413x V = 1.995x W
6 dB = 1.995x V = 3.981x W
10 dB = 3.162x V = 10x W
15 dB = 5.623x V = 31.62x W
17 dB = 7.079x V = 50.12x W (this was typical peak-to-average ratio in music found many years ago)
20 dB = 10x V = 100x W
30 dB = 31.62x V = 1000x W

HTH - Don

 

[CDMC]

No, SMPTE reference calibration level is 85dB(C) SPL at main listening position for a pink noise at about -22dBFSD (rms 22-22k, SMPTE2095-1) with a 12dB crestfactor.
That means that max level with wideband noise would be about 85+(22-12) =95dB(C) SPL.
The max level for sub is more difficult to assess : RTA level of sub is 10dB above RTA of front speakers that gives about 5dB(C) higher SPL so ref level is near 90dB(C) SPL with reference pink noise. As programs may have very different sounds/bandwidth/... in LFE track, it seems more secure to count with a +20dB (instead of +10dB for front speakers) so max 105dB(C) SPL at listening position.

This is the information I have:


“Appendix 2: SMPTE Practice
All quoted monitor SPL calibration figures in this paper are referenced to -20 dB FS. The “theatre standard”, Proposed SMPTE Recommended Practice: Relative and Absolute Sound Pressure Levels for Motion-Picture Multichannel Sound Systems, SMPTE Document RP 200, defines the calibration method in detail. In the 1970’s the value was quoted as “85 at 0 VU” but as the measurement methods became more sophisticated, this value proved to be in error. It has now become “85 at -18 dB FS” with 0 VU remaining at -20 dBFS (sine wave). The history of this metamorphosis is interesting. A VU meter was originally used to do the calibration, and with the advent of digital audio, the VU meter was calibrated with a sine wave to -20 dB FS. However, it was forgotten that a VU meter does not average by the RMS method, which results in an error between the RMS electrical value of the pink noise and the sine wave level. While 1 dB is the theoretical difference, the author has seen as much as a 2 dB discrepancy between certain VU meters and the true RMS pink noise level.
The other problem is the measurement bandwidth, since a widerange voltmeter will show attenuation of the source pink noise signal on a long distance analog cable due to capacitive losses. The solution is to define a specific measurement bandwidth (20 kHz). By the time all these errors were tracked down, it was discovered that the historical calibration was in error by 2dB. Using pink noise at an RMS level of -20 dBFS RMS must correctly result in an SPL level of only 83 dB. In order to retain the magic “85” number, the SMPTE raised the specified level of the calibrating pink noise to -18dB FS RMS, but the result is the identical monitor gain. One channel is measured at a time, the SPL meter set to C weighting, slow. The K-System is consistent with RP 200 only at K-20. I feel it will be simpler in the long run to calibrate to 83 dB SPL at the K-System meter’s 0 dB rather than confuse future users with a non-standard +2 dB calibration point.
It is critical that the thousands of studios with legacy systems that incorporate VU meters should adjust the electrical relationship of the VU meter and digital level via a sine wave test tone, then ignore the VU meter and align the SPL with an RMS-calibrated digital pink noise source.”

Putting aside the 2db difference in methods, the bigger variable is room size. An 83 or 85db reference level is still way too high for small rooms (i.e less than 15,000 cubic feet). The levels need to be adjusted down based on room size.

 

[jlo]

This is the information I have:

But those infos are not up to date or a bit wrong.
See https://screenexcellence.com/downloads/SMPTE_TC_25CSS_B_1_oct_2014.pdf
and here https://ieeexplore.ieee.org/document/7395532

 

[CDMC]

But those infos are not up to date or a bit wrong.
See https://screenexcellence.com/downloads/SMPTE_TC_25CSS_B_1_oct_2014.pdf
and here https://ieeexplore.ieee.org/document/7395532

Sweet, thank you for the links. What is your position on the adjustment for room size, as the SMPTE article only seems to address large rooms (their smallest room which is a dubbing stage is 54' x 32', or 16.5m x 9.7m for those in the rest of the world)? To me this seems to be the big elephant in room (pun intended) that is not being addressed in the home theater arena. 85 db reference level is simply too loud and needs to be adjusted down for room size. I understand that it complicates the issue, which may put manufactures off, but it seems like they would be better starting with a reference of 78 or 80 db which would cover room sizes of up to 10,000 cubic feet (i.e. 12' x 26' x 32' or larger than virtually 99% of domestic rooms). This would at least get people within 2-3db of an accurate target level.

 

[aac]

Sweet, thank you for the links. What is your position on the adjustment for room size, as the SMPTE article only seems to address large rooms (their smallest room which is a dubbing stage is 54' x 32', or 16.5m x 9.7m for those in the rest of the world)? To me this seems to be the big elephant in room (pun intended) that is not being addressed in the home theater arena. 85 db reference level is simply too loud and needs to be adjusted down for room size. I understand that it complicates the issue, which may put manufactures off, but it seems like they would be better starting with a reference of 78 or 80 db which would cover room sizes of up to 10,000 cubic feet (i.e. 12' x 26' x 32' or larger than virtually 99% of domestic rooms). This would at least get people within 2-3db of an accurate target level.

Just set the volume at a comfortable dialogue level, and you have your calibration, no need to bother with all those numbers.
But numbers that people "calibrate" to are too high, that is right.
Movie theaters also crank volume incredibly high sometimes, maybe its modern mastering, I don't know. I stopped going to movie theaters because of that.

 

[jlo]

What is your position on the adjustment for room size, as the SMPTE article only seems to address large rooms

I agree that some movies are too loud : normally the volume control of movie theaters should be set at 7. Due to many clients complaining about too loud levels, the control is often set at 5 or even lower (1 graduation of a Cinema processor is 3.33dB if I remember). Knowing this, some sound engineers mix louder to compensate !
And you are right, in small rooms, lower levels are generally prefered. So as AAC just said, ajust dialog level at your convenience

Back to main subject, I've seen audio tracks with about 30dB crestfactor.

 

[RichB]

The good news is that digital signals have a finite volume represented by 0 dBFS. If you compute the maximum volume on your AVR, you can know that the amplifier will not be clipping.
That does not mean that your speakers are not distorting and your ears are enjoying the experience.

The AHB2s have accurate clipping indicators and I fully expected to bridge them taking them from 180 WPC to 480 WPC into 4 ohms. However, I never saw them clip when listening to music at what I will call rationally loud. That is loud, fun, but with some regard to preserving my hearing. Prior to owning these amps, I would not consider an amplifier with less than 200 WPC into 8 Ohms or 350 into 4 Ohms. Rationally, there is a 3 dB difference here.

One rule-of-thumb is that the speaker sensitivity rating 1 watt at 1m anechoic can be a good starting point for 2 channel music and room reinforcement. I found this to be the case for the Salon2s in my room. 2.83 volts produced 86 dB at my listening position.

Here is the measurement-based method that I used assess the power required for two speakers in my room.

The following is required:

• A Multi-meter to measure AC voltage
• A 440 Hz full scale (0 DBFS) WAV, FLAC, MP3 file.

To measure:

1. Set your AVR/Processor to low level
2. Attach the multimeter to the + and - of your speaker terminals (I used clip to banana but you'll have to find a good method for your speakers)
3. Play the full scale tone (I used 440Hz, 1kHzm and 2kHz tones which have similar results).
4. Adjust the level until the reading is as close to 2.83 volts as possible.
5. Write down this volume level.
6. Make sure your are playing the tone from 1 or 2 speakers (but not in surround mode).
7. Measure the SPL at your listening position with a sound pressure meter. I don't think I would trust a phone.
8. Write down this DB level.
9. Consult your speaker documentation. If you think it is closer to 8 Ohms this is approximately 1 watt. The Salon2s were mostly 4 Ohms so that doubles the power, or 2 watts.
10. Make a spreadsheet with your typical listening levels for music and movies.

I played both speakers with the tone for my tests since this is typical. If you played 1 speaker and agree then double the SPL measured. For me, 2.83 volts = 2 watt and stereo playback of 86 dB. Which was the rated sensitivity for the Salon2s.

Here is a spreadsheet with my typical listening levels, process volume, and maximum power:

The AHB2 clipping indicators support the above power levels.
I watched The Accountant last night and turned up to the maximum I found comfortable. During the shootout it was uncomfortably loud and there were no clipping indicators.

For most movies in my system, loud "cinematic" listening -16. That means the maximum power required is about 64 watts.
Even then, for most of the film, I am using 1/4 of a watt.

The SPL calculator computes the following:

My speakers are more than 2 feet from the side walls but the rears are with so I chose in corners (although that sounds bad ).
The calculated results is 104 dB at 180 watts (the rated power of the AHB2 into 4 Ohms).
My spreadsheet and measurements compute that 128 watts which is very close.

A high-quality amplifier with low distortion and linear performance are still required (IMO) for the Salon2s but, clearly I have enough power and my amplifiers do not clip in my system.

I have 2 friends with Salon2s and Benchmark AHB2 amps. Two of us bi-amp, the other bridges.

- Rich

 

[echopraxia]

The good news is that digital signals have a finite volume represented by 0 dBFS. If you compute the maximum volume on your AVR, you can know that the amplifier will not be clipping.
That does not mean that your speakers are not distorting and your ears are enjoying the experience.

The AHB2s have accurate clipping indicators and I fully expected to bridge them taking them from 180 WPC to 480 WPC into 4 ohms. However, I never saw them clip when listening to music at what I will call rationally loud. That is loud, fun, but with some regard to preserving my hearing. Prior to owning these amps, I would not consider an amplifier with less than 200 WPC into 8 Ohms or 350 into 4 Ohms. Rationally, there is a 3 dB difference here.

One rule-of-thumb is that the speaker sensitivity rating 1 watt at 1m anechoic can be a good starting point for 2 channel music and room reinforcement. I found this to be the case for the Salon2s in my room. 2.83 volts produced 86 dB at my listening position.

Here is the measurement-based method that I used assess the power required for two speakers in my room.

The following is required:

• A Multi-meter to measure AC voltage
• A 440 Hz full scale (0 DBFS) WAV, FLAC, MP3 file.

To measure:

1. Set your AVR/Processor to low level
2. Attach the multimeter to the + and - of your speaker terminals (I used clip to banana but you'll have to find a good method for your speakers)
3. Play the full scale tone (I used 440Hz, 1kHzm and 2kHz tones which have similar results).
4. Adjust the level until the reading is as close to 2.83 volts as possible.
5. Write down this volume level.
6. Make sure your are playing the tone from 1 or 2 speakers (but not in surround mode).
7. Measure the SPL at your listening position with a sound pressure meter. I don't think I would trust a phone.
8. Write down this DB level.
9. Consult your speaker documentation. If you think it is closer to 8 Ohms this is approximately 1 watt. The Salon2s were mostly 4 Ohms so that doubles the power, or 2 watts.
10. Make a spreadsheet with your typical listening levels for music and movies.

I played both speakers with the tone for my tests since this is typical. If you played 1 speaker and agree then double the SPL measured. For me, 2.83 volts = 2 watt and stereo playback of 86 dB. Which was the rated sensitivity for the Salon2s.

Here is a spreadsheet with my typical listening levels, process volume, and maximum power:
View attachment 78477

The AHB2 clipping indicators support the above power levels.
I watched The Accountant last night and turned up to the maximum I found comfortable. During the shootout it was uncomfortably loud and there were no clipping indicators.

For most movies in my system, loud "cinematic" listening -16. That means the maximum power required is about 64 watts.
Even then, for most of the film, I am using 1/4 of a watt.

The SPL calculator computes the following:

View attachment 78479

My speakers are more than 2 feet from the side walls but the rears are with so I chose in corners (although that sounds bad ).
The calculated results is 104 dB at 180 watts (the rated power of the AHB2 into 4 Ohms).
My spreadsheet and measurements compute that 128 watts which is very close.

A high-quality amplifier with low distortion and linear performance are still required (IMO) for the Salon2s but, clearly I have enough power and my amplifiers do not clip in my system.

I have 2 friends with Salon2s and Benchmark AHB2 amps. Two of us bi-amp, the other bridges.

- Rich

This reminds me -- I think a lot of "amp power anxiety" would be alleviated if more amps had clipping indicator lights (and clipping protection) like the Benchmark AHB2. For example, most of my worry of amp power comes not from any lack of loudness from my system (it gets more than loud enough), but just worries that I might unintentionally be clipping either slightly (at loss of sound quality) or extremely during movie peaks (at potential damage to tweeters). This is mostly a theoretical concern I know, given data like that you've provided, but it still would be nice to have a clipping indicator on more amps.

 

[richard12511]

Ironically this is incorrect, but keeps being restated by authorities that should know better. The reference calibration is not 85db at -20dbfs, but 83 db at -20dbfs IN A LARGE THEATER. It was revised more than 30 years ago.

https://www.digido.com/portfolio-item/level-practices-part-2/

The second part is that you need to adjust the levels based on the size of your listening room. When doing so the proper reference level for most domestic rooms ends up at 76-78 db c weighted per speaker using 500-2000hz band limited pink noise. If you look at most of the home theater guys you will see they constantly say that the reference level it too high and they listen 5-10 db below. That is because even the manufactures and industry gurus that should know better and be discussing it always miss the room size adjustment.

https://www.soundonsound.com/techni...ect-studio-reference-monitoring-levels?page=3

Translating this to usage, for most home theaters, the Dolby/THX reference max level should be 96-98 db per speaker and 106-108 for subwoofers. I find that for music listening on highly dynamic music 105db peaks are not out of the question. For compressed music, peak levels are much lower.

For me personally, it doesn’t matter if the standard was revised or if it only applies to larger rooms, as it’s still the target I used, but as a general guideline, it’s good to know.

Something I’ve never really understood (maybe you can explain) is why does the same spl sound quieter in large rooms. Going strictly by the measured spl at my ear, shouldn’t 105db always sound like 105dB. I’ve heard the 75dB home reference before, but I’ve always calibrated for 85dB as 75dB doesn’t sound loud enough to me.

 

[RichB]

This reminds me -- I think a lot of "amp power anxiety" would be alleviated if more amps had clipping indicator lights (and clipping protection) like the Benchmark AHB2. For example, most of my worry of amp power comes not from any lack of loudness from my system (it gets more than loud enough), but just worries that I might unintentionally be clipping either slightly (at loss of sound quality) or extremely during movie peaks (at potential damage to tweeters). This is mostly a theoretical concern I know, given data like that you've provided, but it still would be nice to have a clipping indicator on more amps.

I agree. All amps and even AVRs should have clip indicators. There is a possibility for amps, but I don't hold out much hope for AVRs. Although, Denon is makes great strides with the ability to turn off amps when using external amplification.

ATI amps have clip indicators which is nice. AVRs should have them but I think that they are, for the most part, not in the amp business so clip indicators are not implemented.

The AHB2 has separate clip indicators per channel and distinguishes between voltage and current clipping and shuts down when sustained distortion exceed 1%. The first watt and below is where most listening is done so noise and distortion at low levels are most important, even while, amps with sufficient power should be available the users system and preferences.

I probably did not need to do the power analysis with the AHB2s but I was also concerned that during peak they would clip. I became curious when I was pushing the system beyond what I find enjoyable with movies and they were not clipping. The analysis provided validation.

At one time, I had ATI AT6000s (300/450 WPC into 8/4 ohms) and still worried about clipping. The 7 channel amps must be freight shipped. Now, I suspect the AT4000s would have been more than enough.

- Rich

 

[echopraxia]

I agree. All amps and even AVRs should have clip indicators. There is a possibility for amps, but I don't hold out much hope for AVRs. Although, Denon is makes great strides with the ability to turn off amps when using external amplification.

ATI amps have clip indicators which is nice. AVRs should have them but I think that they are, for the most part, not in the amp business so clip indicators are not implemented.

The AHB2 has separate clip indicators per channel and distinguishes between voltage and current clipping and shuts down when sustained distortion exceed 1%. The first watt and below is where most listening is done so noise and distortion at low levels are most important, even while, amps with sufficient power should be available the users system and preferences.

I probably did not need to do the power analysis with the AHB2s but I was also concerned that during peak they would clip. I became curious when I was pushing the system beyond what I find enjoyable with movies and they were not clipping. The analysis provided validation.

At one time, I had ATI AT6000s (300/450 WPC into 8/4 ohms) and still worried about clipping. The 7 channel amps must be freight shipped. Now, I suspect the AT4000s would have been more than enough.

- Rich

It also just occurred to me... I wonder if similar clipping vs SPL tests can just be done via REW and a calibration mic. I know this would also be testing the speaker's distortion vs SPL capabilities, but that's actually a good thing anyway (since end-to-end results is what ultimately matters). I'm just not so sure a sinewave sweep would be really useful here, vs broad spectrum noise (and I'm not sure if it's possible to measure distortion from the latter).

 

[RichB]

It also just occurred to me... I wonder if similar clipping vs SPL tests can just be done via REW and a calibration mic. I know this would also be testing the speaker's distortion vs SPL capabilities, but that's actually a good thing anyway (since end-to-end results is what ultimately matters). I'm just not so sure a sinewave sweep would be really useful here, vs broad spectrum noise (and I'm not sure if it's possible to measure distortion from the latter).

When discussing dynamic performance, it may be difficult to measure momentary loss of linearity (not considered distortion) and THD. This is where I suspect power supply headroom of stand-alone amplifier has the advantage over an AVR.

After having many amplifiers, I decided to move away from large transformer (and transformer hum) to a SOTA design with few if any flaws. It worked, I am done buying amps. If I am ever presented with clip indicators, they can be bridged. So far, bridging is not required and all 5 AHB2s (in two systems) have been flawless.

- Rich