Measuring and understanding THD and THD+N for real world tracks - Page 4 - Home Theater Forum and Systems - HomeTheaterShack.com

Old 02-10-10, 10:54 PM
Senior Shackster
Bill

Join Date: Sep 2009
Location: Austin, TX
Posts: 676
re: Measuring and understanding THD and THD+N for real world tracks

Quote:
jmcomp124 wrote: View Post
... Assume that for a real world movie track, if you are able to measure the Max Voltage(rms) at the subwoofer preamp output (MV=0 and the whole movie is sampled) and you are able to bound that voltage to the amp sensitivity setting being the ceiling (1.4v), then you can prevent the amp from clipping (assume 0 attenuation in the gain).
Is this assumption correct?
Looking around on the web, I think your understanding of amplifier sensitivity is correct, that it is the voltage that will drive the amp to maximum power, for a given gain setting and output impedance.

I think the problem with determining a priori where clipping will occur arises with the impedance of the speaker. Amplifiers are designed using assumptions about the load of the speaker, but in the real world the impedance of the speaker is not constant. So it may draw more current at some frequencies than at others, and the amplifier clips when its output current limit is reached. At least, that's what I've gathered reading about satellite speakers, and I assume the same is true of the sub woofer even over its limited range.

So the calculations give you a best case limit, but you might begin clipping before getting to that point. If you want to guarantee no clipping, you need to account for all the variables, or leave a conservative cushion. From your description of your amplifiers, with a 3600w limit, I would have thought that they provided so much power that for any calibration to reference level you would get no where near their limits. This exercise should be largely academic for you, right?

Bill
laser188139 is offline

Old 02-10-10, 11:06 PM Thread Starter
Senior Shackster
Jai

Join Date: Feb 2007
Location: Portland, OR, USA
Posts: 181
re: Measuring and understanding THD and THD+N for real world tracks

Quote:
laser188139 wrote: View Post
Greg simplified slightly. If the test tone measured at 75dB, then the LFE output should never exceed 115dB at reference levels. But the 5 or 7 satellite channels contribute to the sub output below their crossovers. So the theoretical maximum for film is usually given as 121dB. The chances of all the channels having 115dB/105dB of bass content at the same time are small.

Of course, this was calibrated at your listening position. If you are sitting more than 1m from the speakers, even more SPL is needed, as you are going to lose 6dB every time the distance doubles.

When looking at amplifier power, of course you have to recognize that the impedance of the speaker is not really constant with respect to frequency. If the impedance of the speaker drops to 4ohms instead of 8ohms at the tuned frequency, then it will need 4 times as much amplifier power for the same SPL. Which is why most people regard the calculation of amplifier power for a given sound level as an estimated, to be tempered by empirical performance. In other words, if the sub bottoms out or the amplifier clips, don't drive it quite that hard.
Thanks Bill. I have noticed this with redirected bass wherein it goes louder than 115dB, in fact closer to 120dB which matches your description. You bring up a very crucial point regarding impedance. It's been at the back of my mind. A significant variation in impedance will sink all these assumptions I am making. It's been a while since I looked at my WinISD models, but now that I have the box, is there an easy way to measure impedance changes with movie tracks? I think I am asking for too much and maybe just stick to the traditional approach of using the Clip LEDs which I already have dialed in.
My current setup has a 6dB in-line attenuator with a Rane PE-17 that is used as a subsonic filter at 17Hz. Without the attenuator, the Rane begins to clip at close to reference levels. For the best signal to noise ratio I have both input/output gain at max. This is another reason why I want to keep the Vrms at less than 2v.
jmcomp124 is offline
Old 02-10-10, 11:11 PM Thread Starter
Senior Shackster
Jai

Join Date: Feb 2007
Location: Portland, OR, USA
Posts: 181
re: Measuring and understanding THD and THD+N for real world tracks

Quote:
laser188139 wrote: View Post
Looking around on the web, I think your understanding of amplifier sensitivity is correct, that it is the voltage that will drive the amp to maximum power, for a given gain setting and output impedance.

I think the problem with determining a priori where clipping will occur arises with the impedance of the speaker. Amplifiers are designed using assumptions about the load of the speaker, but in the real world the impedance of the speaker is not constant. So it may draw more current at some frequencies than at others, and the amplifier clips when its output current limit is reached. At least, that's what I've gathered reading about satellite speakers, and I assume the same is true of the sub woofer even over its limited range.

So the calculations give you a best case limit, but you might begin clipping before getting to that point. If you want to guarantee no clipping, you need to account for all the variables, or leave a conservative cushion. From your description of your amplifiers, with a 3600w limit, I would have thought that they provided so much power that for any calibration to reference level you would get no where near their limits. This exercise should be largely academic for you, right?

Bill
You are right, it is largely academic. Usually when I do such an exercise, I stumble upon other ideas. Thanks for the knowledge transfer. My previous post about the Rane also is one reason why I am thinking along these lines.
Similar to dialogue normalization, it would be nice if movie tracks are encoded with metadata that says what the peak demand of the LFE track in that movie is. However, then it would have to account for redirected bass and things get way too complex. I am glad at least someone here is connecting with the strange ideas in my mind without getting irritated with my questions.
Thanks again!
jmcomp124 is offline
Old 02-10-10, 11:27 PM
Senior Shackster
Bill

Join Date: Sep 2009
Location: Austin, TX
Posts: 676
re: Measuring and understanding THD and THD+N for real world tracks

You're welcome, Jai.

Quote:
jmcomp124 wrote: View Post
... is there an easy way to measure impedance changes with movie tracks? ...
If you really want to get into measuring speaker impedance curves, my guess is your google skills are as good as mine. Fascinating as the subject is, I don't think I'm going down that path with you.

Have fun,
Bill
laser188139 is offline
Old 02-10-10, 11:37 PM Thread Starter
Senior Shackster
Jai

Join Date: Feb 2007
Location: Portland, OR, USA
Posts: 181
re: Measuring and understanding THD and THD+N for real world tracks

Quote:
laser188139 wrote: View Post
You're welcome, Jai.

If you really want to get into measuring speaker impedance curves, my guess is your google skills are as good as mine. Fascinating as the subject is, I don't think I'm going down that path with you.

Have fun,
Bill
I understand and no problem. In utopia, I would like a box that measures driver excursion in real time and modulate input voltage accounting for amplifier limits also. For now, I am not going down that path you don't want to tread either.
jmcomp124 is offline
Old 02-11-10, 10:26 AM Thread Starter
Senior Shackster
Jai

Join Date: Feb 2007
Location: Portland, OR, USA
Posts: 181
re: Measuring and understanding THD and THD+N for real world tracks

Quote:
laser188139 wrote: View Post
Looking around on the web, I think your understanding of amplifier sensitivity is correct, that it is the voltage that will drive the amp to maximum power, for a given gain setting and output impedance.

I think the problem with determining a priori where clipping will occur arises with the impedance of the speaker. Amplifiers are designed using assumptions about the load of the speaker, but in the real world the impedance of the speaker is not constant. So it may draw more current at some frequencies than at others, and the amplifier clips when its output current limit is reached. At least, that's what I've gathered reading about satellite speakers, and I assume the same is true of the sub woofer even over its limited range.

So the calculations give you a best case limit, but you might begin clipping before getting to that point. If you want to guarantee no clipping, you need to account for all the variables, or leave a conservative cushion. From your description of your amplifiers, with a 3600w limit, I would have thought that they provided so much power that for any calibration to reference level you would get no where near their limits. This exercise should be largely academic for you, right?

Bill
Bill,
Pondering through this, it appears that the impedance variation may be a non-issue in my particular case because I use the LMS woofers that claim linear BL. This suggests an almost linear impedance since B will be mostly uniform with current being the larger variable but since BL is linear then impedance must be. Even if it does vary, variations may not be large so, gating the input voltage to 1.4v (amp sensitivity) using the loudest subwoofer track in the movie, should work. I will post results as I get some time to do these experiments. A friend of mine in my labs has equipment to measure peak voltage in fine granularity and he offered to contribute. I will post whatever I find, if folks here are interested?.
Thanks,
-Jai
jmcomp124 is offline
Old 02-11-10, 10:37 AM
Elite Shackster

Greg

Join Date: Nov 2007
Location: NY
Posts: 2,601
re: Measuring and understanding THD and THD+N for real world tracks

Quote:
laser188139 wrote: View Post
Greg simplified slightly. If the test tone measured at 75dB, then the LFE output should never exceed 115dB at reference levels. But the 5 or 7 satellite channels contribute to the sub output below their crossovers. So the theoretical maximum for film is usually given as 121dB. The chances of all the channels having 115dB/105dB of bass content at the same time are small.
Thanks Bill, I hadn't thought of that.
Quote:
Of course, this was calibrated at your listening position. If you are sitting more than 1m from the speakers, even more SPL is needed, as you are going to lose 6dB every time the distance doubles.
But if the 75dB had been calibrated at the LP, the additional SPL the sub was putting out to achieve that was already accounted for?

-Greg

Don't worry... nothing new here, I've already made that mistake. Trust me.
glaufman is offline
Old 02-11-10, 10:56 AM
Senior Shackster
Bill

Join Date: Sep 2009
Location: Austin, TX
Posts: 676
re: Measuring and understanding THD and THD+N for real world tracks

Quote:
glaufman wrote: View Post
... But if the 75dB had been calibrated at the LP, the additional SPL the sub was putting out to achieve that was already accounted for?
You are probably right, that this does not enter into Jai's question of the maximum voltage limit where clipping would occur in the amp. I was focused on the testbed environment, where we were calibrating 75dB at REW to match a particular voltage, and confounded the two 75dBs.

I think it can enter the question of staying within the limits of the sub's physical speaker. But there Jai has suggested real time measurement of cone excursion.

Last edited by laser188139; 02-11-10 at 03:19 PM.
laser188139 is offline
Old 02-11-10, 03:05 PM Thread Starter
Senior Shackster
Jai

Join Date: Feb 2007
Location: Portland, OR, USA
Posts: 181
re: Measuring and understanding THD and THD+N for real world tracks

Quote:
glaufman wrote: View Post
Thanks Bill, I hadn't thought of that. But if the 75dB had been calibrated at the LP, the additional SPL the sub was putting out to achieve that was already accounted for?
I am not sure that is the case. I think the test tones are discrete to each channel. If it did contain redirected bass tones, then we would hear something in the mains during subwoofer tones due to the way crossovers are designed.
jmcomp124 is offline
Old 02-11-10, 07:52 PM Thread Starter
Senior Shackster
Jai

Join Date: Feb 2007
Location: Portland, OR, USA
Posts: 181
re: Measuring and understanding THD and THD+N for real world tracks

Quote:
glaufman wrote: View Post
But if the 75dB had been calibrated at the LP, the additional SPL the sub was putting out to achieve that was already accounted for?
I probably misunderstood your question. I think this question was tied to Bill's comment below...

"Of course, this was calibrated at your listening position. If you are sitting more than 1m from the speakers, even more SPL is needed, as you are going to lose 6dB every time the distance doubles"

The additional SPL the sub or any other channel would have to do, would be relative to their distance from the primary listening position. My point is that, the subwoofer test tone, does not account for the redirected bass from the mains or other channels. As a simple test, if you set your mains to full band and the center and surrounds to the lowest crossover possible and measure the SPL (C-weighted) at the listening position, you will find that it is the same as when the crossover for all main and surround channels are set to say 100Hz. This proves that the subwoofer test tone does not account for the redirected bass from movie tracks that can push it 6dB higher. So with that clarified, there are some movies recorded hot as someone (Ricci?) pointed out earlier. So you can be even 10dB off by simple extrapolation with internal test tone calibrations. If this reasoning is not correct, please let me know.
Thanks,
-Jai
jmcomp124 is offline

 Bookmarks

 Tags measruing , real , thd+n , tracks , understanding , world

Message:
Options

## Register Now

Random Question
Random Question #2

User Name:
 PLEASE READ BELOW PRIOR TO ENTERING AN EMAIL ADDRESS! ATTENTION! YOU MUST ACTIVATE YOUR ACCOUNT!Activation requires you reply to an email we will send you after you register... if you do not reply to this email, you will not be able to view certain areas of the forum or certain images... nor will you be able download software. AN INVALID EMAIL ADDRESS WILL CAUSE YOUR ACCOUNT TO BE DELETED! See our banned email list here: Banned Email List We DO NOT respond to spamcop, boxtrapper and spamblocker emails... please add @hometheatershack DOT com to your whitelist prior to registering or you will get nowhere on your registration. Email Address:
OR

## Log-in

Human Verification

In order to verify that you are a human and not a spam bot, please enter the answer into the following box below based on the instructions contained in the graphic.