or Connect
AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Bi-amping B&W CM10
New Posts  All Forums:Forum Nav:

# Bi-amping B&W CM10 - Page 8

Quote:
Originally Posted by kiwi2

Quote:
Originally Posted by whoaru99

If you listened to music with less dynamic range, say 10dB, the max/peak SPL of the system doesn't change.

Max and peak are different things. Max may be an average of 0.2 of a second. Peak may be a brief moment that happens much quicker than that. Peak refers to single information in the recording itself.

Hmmm.

In electronics peak refers to an instantaneous (meaning something that may conceptually last for zero time, IOW a point in time) maximum.

Maximum means just that - something that in some sense is maximized.

For example an power line can have a maximum voltage (usually meaning RMS voltage) of 120.0 volts. Maximum in this sense means that we never observe it any higher.

If the the power line voltage is a sine wave, then the corresponding peak voltage is that 120 volts multiplied by the square root of two or 169.70562748477140585620264690516... volts.

### AVS Top Picks

And to obsfucate the topic even more, has anyone taken into account the level of SPL when using those special, peaky audiophile speaker cables: the Volcano, the K2, and the most, peakiest of all, the Everest.

OMG, I just imagined biamping with the Everest. Just the thought. What a rush of adrenaline.
Quote:
Originally Posted by arnyk

Hmmm.

Again, I will refer to this discussion where someone put a voltmeter across the terminals and compared the reading against a SPL meter. Peak volts was higher (in dB terms) than the max of the SPL meter.

A recording was something made in the real world. And in the real world you can have very short but very high peaks like from the leading edge hit as an instrument is hit, stuck, plucked, whatever.
Quote:
Originally Posted by whoaru99

I agree they can be different.

What's your point relative to the calculator or isn't that your point?

Even max SPL is only an average. Short peaks can still exist on top of max SPL.

It takes an oscilloscope or special detector to measure instantaneous peaks. A voltmeter won't do it, at least most won't (you need one with wide bandwidth and a peak hold feature; there are some but they are not your average Radio Shack or other bargain meter). And, there is the problem of converting volts into acoustical output; there's no easy way to confirm XX volts is actually outputting YY dB SPL due to the (in general) unknown transfer function from terminals to sound (acoustic output). Apples and oranges.

Max to me is max, be it instantaneous or longer in duration. If you want to say "maximum average value" then add headroom to a peak value I understand what that means. Specs such as THX determine a max (peak, whatever) value above a defined reference level and so headroom is already taken into account. I doubt folk listen at 105 dB and then want another 10 dB of headroom (I am sure there are some). At some point the argument becomes academic. Well, until you blow something up... I typically listen around 70'ish dB and the system can output 105 dB max peaks. Enough for me.
Quote:
Originally Posted by kiwi2

Quote:
Originally Posted by arnyk

Hmmm.

Again, I will refer to this discussion where someone put a voltmeter across the terminals and compared the reading against a SPL meter. Peak volts was higher (in dB terms) than the max of the SPL meter.

A recording was something made in the real world. And in the real world you can have very short but very high peaks like from the leading edge hit as an instrument is hit, stuck, plucked, whatever.

However, the ear doesn't notice clipping or other forms of distortion until they have been going on longer than instantaneous.

There is no general agreement that I can find about even how to express the prevalence of clipping.

One way is the percentage of time that the music is clipped, and some say 10%. I've experimented with this, and it depends on the music.

Another is the number of dB shaved off by clipping, and some say 3 dB. I've experimented with this, and it depends on the music.

But, whatever it is, it isn't 1 microsecond, 1% or something fleeting like that. In moderate amounts clipping seems to music sound a little brighter... In large amounts it makes music sound like mud.
Quote:
Originally Posted by DonH50

Max to me is max, be it instantaneous or longer in duration. If you want to say "maximum average value" then add headroom to a peak value I understand what that means. Specs such as THX determine a max (peak, whatever) value above a defined reference level and so headroom is already taken into account. I doubt folk listen at 105 dB and then want another 10 dB of headroom (I am sure there are some). At some point the argument becomes academic. Well, until you blow something up... I typically listen around 70'ish dB and the system can output 105 dB max peaks. Enough for me.

Thank you.
Quote:
Originally Posted by DonH50

It takes an oscilloscope or special detector to measure instantaneous peaks.

My tool of choice is a mutichannel pro grade audio interface running at 24/96.

Matching up with a AVR output takes a calibrated attenuator that can take the ca. 40-100 volt signals from the power amps down to the 1-6 volts that audio interfaces swing with.
Quote:
A voltmeter won't do it, at least most won't (you need one with wide bandwidth and a peak hold feature; there are some but they are not your average Radio Shack or other bargain meter).

IME the capture time of even Flukes is far from instantaneous. True RMS is desirable, but it also tends to slow peak response down.

No, most analog voltmeters are not the tool of choice. Oscilloscopes are better but you need some kind of recording or holding facility because transients are not forever. A 12 bit USB oscilloscope adaptor for a PC seems like a good tool and they are not prohibitively expensive any more.

Quote:
And, there is the problem of converting volts into acoustical output; there's no easy way to confirm XX volts is actually outputting YY dB SPL due to the (in general) unknown transfer function from terminals to sound (acoustic output). Apples and oranges.

If you want to make accurate measurements of SPL you will have a mic calibrator, and once you have one of those relating volts to SPL is actually pretty easy.

Quote:
Max to me is max, be it instantaneous or longer in duration. If you want to say "maximum average value" then add headroom to a peak value I understand what that means. Specs such as THX determine a max (peak, whatever) value above a defined reference level and so headroom is already taken into account. I doubt folk listen at 105 dB and then want another 10 dB of headroom (I am sure there are some). At some point the argument becomes academic. Well, until you blow something up... I typically listen around 70'ish dB and the system can output 105 dB max peaks. Enough for me.

To put things into perspective, the max SPL for professional gear is often spec'd at 116 dB.
Quote:
Originally Posted by arnyk

A 12 bit USB oscilloscope adaptor for a PC seems like a good tool and they are not prohibitively expensive any more.

I have one of these through work. A Picoscope 4423, 4 channnel 20MHz.

Posted this in some other thread, but here's a trace of turn-on power draw from one of my amps.

Hey Arny,

Good points, of course!

Old analog guy that I am, I didn't consider the obvious digital interface, duh!

I am sure there are cheaper meters, but the one I had in mind I was using at work for precision measurements, an HP (or, Agilent) DMM running around \$10k USD. I was not thinking of the Fluke handhelds; some of the top models are (or were, have not looked recently) worthy contenders but as you say still too slow especially in RMS mode (not the mode to use for peak detection, as you noted). Peak detection is not a priority for most voltmeters, analog or digital, and their bandwidth typically targets power circuits and not audio. There are some specialized audio meters but be prepared to spend a grand or more.

A DSO makes capturing transient peaks easy.

A mic calibrator is a good idea; mine disappeared ages ago and I have not had the urge to replace it. They are not all that expensive so long as you do not need to go too low in frequency or too high in amplitude. My SPL meter is not calibrated (cheapo Rat Shack). I rarely feel the need to know actual amplitude with any real accuracy, especially since my primary line of work for the past few decades has not been audio installations. I do have a decent measurement mic I use (Earthworks M30) and it includes a calibration chart and curve for frequency response, but IIRC amplitude is just a single-point measurement.

116 dB is about twice as loud as my 105 dB, which is already louder than I can stand. However, professional systems usually target a much larger area than my little room. And, I am not into Kiss (140 dB peaks at a live show, good grief)...

Onwards - Don
Here is an interesting discussion in the DIY section that asks "What causes a speaker to be dynamic" and the general consensus is to have at least 10dB headroom on top of max SPL...

http://www.avsforum.com/t/1509494/what-causes-a-speaker-to-be-dynamic

Of course that isn't only amplifier power but also the driver's ability to handle x amount of power without running into compression as well.
I dont believe that reference is at all directed to amp power. It is with respect to speaker/driver capability.
I also don't see the relevance to a discussion on passively biamping speakers.
Quote:
Originally Posted by kiwi2

Here is an interesting discussion in the DIY section that asks "What causes a speaker to be dynamic" and the general consensus is to have at least 10dB headroom on top of max SPL...

http://www.avsforum.com/t/1509494/what-causes-a-speaker-to-be-dynamic

Of course that isn't only amplifier power but also the driver's ability to handle x amount of power without running into compression as well.

I just read the thread and I see not even a little hint about amplifier compresion, only about the incredible power levels required to hit insane SPL levels with average efficiency drivers..

Seems like me to be a case of reading what one wishes to read! ;-)

The thread is all about driver compression. Perhaps you can point out the 7 or so posts agreeing that amplifier power is relevant in order to find the consensus the above post seems to be claiming.

This is a typical comment:

http://www.avsforum.com/t/1509494/what-causes-a-speaker-to-be-dynamic#post_24155670

""a loose rule of thumb where compression starts to become a factor, even if slightly, at like 1/3 or 1/2 a speakers rated wattage handling?""

All true - the high output systems used for live sound and high performance HT are typically in the 96-100 dB/W range, and that cuts the amplifier power required by factors of 4 to 10. With all that, driver compression is still an issue.

Driver compression is well-described in the thread and may even be asssumed to be its topic.

The best, most linear drivers are horrifically nonlinear near full output compared to amps, and almost nobody actually uses them. They use the junior parts. A good compression driver runs \$200-400 and a matching waveguide maybe 1/4-half that. That gives you a 3-way speaker containing maybe \$1,000 worth of raw components for a projected system street 1 piece price of about \$6K. The pros use these systems by the dozens.
Edited by arnyk - 1/6/14 at 6:39am
New Posts  All Forums:Forum Nav:
Return Home
Back to Forum: Audio theory, Setup and Chat

### AVS Top Picks

AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Bi-amping B&W CM10