ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1]   Go Down

Author Topic: How to measure?  (Read 3801 times)

Michael Gorecki

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 295
How to measure?
« on: September 30, 2014, 12:35:18 AM »

Alright so this is the last new thread I'll do for a little bit. I have another open post asking about speaker placement and alignment for a mismatched system. This post is in regards to obtaining measurements of those speakers. In the other post it was mentioned to use "Standard Drive Voltage" which I can assume is 1watt at 1meter but was presented as 2.83v at 1 meter. The question is " How do I get 1 watt?(2.83v)" and then I can assume that the measurement tone would be Pink Noise but I would like clarification as to the exact sound/tone.

I will not be charging the client for this service or setup of any of this gear. This is completely experimental to try and yield better results than I can with just using my ears.


Lastly, where can I get an open field to test speakers at?

Thanks for the help

Michael
Logged

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 9203
  • Atlanta GA
Re: How to measure?
« Reply #1 on: September 30, 2014, 07:54:24 AM »

Alright so this is the last new thread I'll do for a little bit. I have another open post asking about speaker placement and alignment for a mismatched system. This post is in regards to obtaining measurements of those speakers. In the other post it was mentioned to use "Standard Drive Voltage" which I can assume is 1watt at 1meter but was presented as 2.83v at 1 meter. The question is " How do I get 1 watt?(2.83v)" and then I can assume that the measurement tone would be Pink Noise but I would like clarification as to the exact sound/tone.

I will not be charging the client for this service or setup of any of this gear. This is completely experimental to try and yield better results than I can with just using my ears.


Lastly, where can I get an open field to test speakers at?

Thanks for the help

Michael
First of all you cannot apply 1 watt to a loudspeaker-at least not very easily and nobody actually does this-DESPITE what they say on the spec sheets.

If they did-the response would be varying much more than it does.

You only apply a VOLTAGE (typically 2.83V).  Since the impedance of the speaker changes with freq-so will the wattage-so at some freq you have much less than 1 watt and at other more than 1 watt.

2.83V is 1 watt into an 8 ohm load.  If you look at the actual impedance curve of a loudspeaker you will see that it is "kinda around 8 ohms maybe-depending".  Sometimes it is all a good bit above 8 ohms-but 8 is the closest standard value.

This goes for 4 and 16 ohms as well.  They are just "standard placeholders" and often do not actually represent the ACTUAL impedance of the loudspeaker-but it keeps it simple for people looking for simple answers to complex problems and helps them choose an amplifier size.

Typically pink noise is not used to get loudspeaker specs-but rather a swept tone is used.  Noise varies to much to get a constant value.

If you are trying to do system alignment-then who cares what the applied voltage/(resultant wattage) is-it will provide you with no information.  What you are looking for is COMPARATIVE-NOT ABSOLUTE measurements.

And in that case a 2 channel FFT (not swept sine) is often your fastest and easiest approach.

You don't have to have an open field-just an open area.  It also does not have to be "in the middle of nowhere" but you have to be able to recognize what reflections look like in a measurement and when to simply ignore then (for the sake of alignment).
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

Michael Gorecki

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 295
Re: How to measure?
« Reply #2 on: September 30, 2014, 11:39:48 AM »

Well that was definitely informative. I obviously need to get some books on electronics to understand more of this. You almost answered my question or you might have and I'm just ignorant to the terminology. How would I apply 2.83v to a speaker?

And I'm not sure exactly what measurements I need to take. I'm really just going to get out there and gather as much information as I can. I would just like to be gathering the right information from the beginning. The problem that I'm having is that the front loaded 218's are nullifying the 21" folded horns and I'm trying to get them to play nicely with each other.

Thanks for the info.

Michael


Sent from my iPhone using Tapatalk
Logged

Tim McCulloch

  • SR Forums
  • Hero Member
  • *
  • Offline Offline
  • Posts: 22187
  • Wichita, Kansas USA
Re: How to measure?
« Reply #3 on: September 30, 2014, 03:52:12 PM »

Well that was definitely informative. I obviously need to get some books on electronics to understand more of this. You almost answered my question or you might have and I'm just ignorant to the terminology. How would I apply 2.83v to a speaker?

And I'm not sure exactly what measurements I need to take. I'm really just going to get out there and gather as much information as I can. I would just like to be gathering the right information from the beginning. The problem that I'm having is that the front loaded 218's are nullifying the 21" folded horns and I'm trying to get them to play nicely with each other.

Thanks for the info.

Michael


Sent from my iPhone using Tapatalk

If you don't own a signal generator, find a WAV file of a 60Hz sine wave.  Hook up your voltmeter to the output of an amplifier and turn up the mixer until you get a reading of 2.8v.  Disconnect the meter and hook up the speaker to the amp.  The choice of 60Hz is based on most common meters not having a wide frequency bandwidth, and for the USA/Canada/Mexico meters are designed to work best at AC mains frequency.  If you have a wider bandwidth, true RMS meter, use it.

The real issue is this conglomeration of equipment is not relative sensitivity (hell, you can probably match levels close enough by ear alone), but that none of it was designed to work with the other products from a TIME standpoint.  Time is phase, time is chronology, time is distance, time is frequency.  All subwoofer designs create some type of acoustic filter (that can ultimately be modeled as an electrical circuit), and filters require TIME to do their business.  Additionally, any electrical filters used will add their own time-variant components).

What we're measuring for is to see how close or far apart in time 2 sources are, and over how much of those device's operating bandwidth that difference occurs.  Note that their relative positions in time can/will vary with frequency.

A much simpler approach would be to return the folded horn speakers and trade them for more of the front loaded subs, or vice versa.  Why?  Because I'm not confident that you'll be able to learn the WHYS of measurement and alignment over the internet, in 6 Easy Lessons«.  It's not a failing of Michael G, it's that this stuff takes a really long time to learn.  You can be making measurements in about 2 hours after gathering the equipment, software, and watching a couple of tutorials, but you'll be making bad measurements and still not really know what you're looking at.  I made hundreds of invalid measurements before I learned what they look like on screen, and what was stuff I could safely ignore for the task at hand.  I still make bad, noisy or otherwise sketchy measurements but now I'm much less likely to act upon them  8)
Logged
"Practicing an art, no matter how well or badly, is a way to make your soul grow, for heaven's sake. Sing in the shower. Dance to the radio. Tell stories. Write a poem to a friend, even a lousy poem. Do it as well as you possible can. You will get an enormous reward. You will have created something."  - Kurt Vonnegut

Art Welter

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1938
Re: How to measure?
« Reply #4 on: September 30, 2014, 04:03:54 PM »

How would I apply 2.83v to a speaker?

And I'm not sure exactly what measurements I need to take. I'm really just going to get out there and gather as much information as I can. I would just like to be gathering the right information from the beginning. The problem that I'm having is that the front loaded 218's are nullifying the 21" folded horns and I'm trying to get them to play nicely with each other. 
Michael ,

Tim already answered your first question, I 'll explain what you need to do to to get the measurements to align a system using different types of speakers that don't have presets you can easily adopt.

An example below shows the response of two subs, one a bass reflex (BR),  the other a tapped horn (TH), both tested using Smaart (a dual FFT system that can do frequency, phase and time tests) with the mic laying on the ground (ground plane, half space measurement) the same distance (2 meters) in front of the cabinets,  both using BW 25 and 125 filters.

Filters and DSP processing add delay, but both cabinets using the same filters result in the same delay latency. The path length sound must travel through the TH delays the sound emitted from it's mouth compared to the front loaded BR. Smaart has a feature that can measure delay time, then apply delay to the test signal so that it  "lines up" with the delayed output of the loudspeaker at the test distance. Without correct time alignment, phase readings will be inaccurate. Blowing wind can result in hundreds of degrees of phase shift above 10 kHz, but fortunately the long wave lengths of subwoofer frequencies are fairly immune to that problem, though wind noise can still affect frequency response, and phase is intimately linked to frequency response. Use wind screens, wait until gusts settle before hitting "save".

Looking at the delay time we had Smaart apply, we see the TH has a delay of 26.52ms, the BR 8.62ms. We can see in the phase response that the TH lags the BR 180 degrees at 125 Hz, due to it being connected reverse polarity compared to the BR. Because of the polarity reversal, the TH lags by one cycle at the crossover point plus the path length difference, a total of 17.9ms. The wavelength of 125 Hz is 8ms, if we correct the polarity reversal (1/2 wavelength), to time align the two cabinets at 125 Hz would requires the BR to be delayed by 13.9ms, the difference between the acoustic center of the BR and TH. Actually, because of the high frequency content used to determine the delay in Smaart, and the rapidly changing upper phase response in the TH, the measured time difference is off,  it should only be around 9ms. There are a lot of little details that take a lot of testing to figure outů

Back to looking at the phase response, it is obvious that the BR and TH are different, at 40 Hz there is a 120 degree difference, at 125Hz a 180 degree difference. Reversing the TH polarity (a 180 degree phase change at all frequencies) would make their phase line up at 125 Hz, but the BR would still lag the TH by 60 degrees at 40 Hz. For signals to coherently sum, they must be within 1/4 wavelength, no more than a 90 degree phase difference. 60 degrees is less than 90 degrees, so if we time/phase align the upper response, the LF response still will be decent. Because the wavelengths at the upper end of a sub's pass band are much shorter than at the low end, small time alignment differences will result in large phase differences. In large systems, one can only align time and phase in one position.

Care must be used when applying delay, if too much is added, the phase response can look smooth, but the LF could be lagging by an entire wavelength. Precise time/phase alignment can be tested by reversing polarity of the delayed system, the resulting cancellation dip will be deepest when the delay is set exactly right.

The same alignment process must be done with the upper cabinets in the crossover region to deliver "punch" and impact .
Easy as that ;).

Art
Logged

Mario Pollio

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 90
Re: How to measure?
« Reply #5 on: September 30, 2014, 05:56:17 PM »

Michael ,

Tim already answered your first question, I 'll explain what you need to do to to get the measurements to align a system using different types of speakers that don't have presets you can easily adopt.

An example below shows the response of two subs, one a bass reflex (BR),  the other a tapped horn (TH), both tested using Smaart (a dual FFT system that can do frequency, phase and time tests) with the mic laying on the ground (ground plane, half space measurement) the same distance (2 meters) in front of the cabinets,  both using BW 25 and 125 filters.

Filters and DSP processing add delay, but both cabinets using the same filters result in the same delay latency. The path length sound must travel through the TH delays the sound emitted from it's mouth compared to the front loaded BR. Smaart has a feature that can measure delay time, then apply delay to the test signal so that it  "lines up" with the delayed output of the loudspeaker at the test distance. Without correct time alignment, phase readings will be inaccurate. Blowing wind can result in hundreds of degrees of phase shift above 10 kHz, but fortunately the long wave lengths of subwoofer frequencies are fairly immune to that problem, though wind noise can still affect frequency response, and phase is intimately linked to frequency response. Use wind screens, wait until gusts settle before hitting "save".

Looking at the delay time we had Smaart apply, we see the TH has a delay of 26.52ms, the BR 8.62ms. We can see in the phase response that the TH lags the BR 180 degrees at 125 Hz, due to it being connected reverse polarity compared to the BR. Because of the polarity reversal, the TH lags by one cycle at the crossover point plus the path length difference, a total of 17.9ms. The wavelength of 125 Hz is 8ms, if we correct the polarity reversal (1/2 wavelength), to time align the two cabinets at 125 Hz would requires the BR to be delayed by 13.9ms, the difference between the acoustic center of the BR and TH. Actually, because of the high frequency content used to determine the delay in Smaart, and the rapidly changing upper phase response in the TH, the measured time difference is off,  it should only be around 9ms. There are a lot of little details that take a lot of testing to figure outů

Back to looking at the phase response, it is obvious that the BR and TH are different, at 40 Hz there is a 120 degree difference, at 125Hz a 180 degree difference. Reversing the TH polarity (a 180 degree phase change at all frequencies) would make their phase line up at 125 Hz, but the BR would still lag the TH by 60 degrees at 40 Hz. For signals to coherently sum, they must be within 1/4 wavelength, no more than a 90 degree phase difference. 60 degrees is less than 90 degrees, so if we time/phase align the upper response, the LF response still will be decent. Because the wavelengths at the upper end of a sub's pass band are much shorter than at the low end, small time alignment differences will result in large phase differences. In large systems, one can only align time and phase in one position.

Care must be used when applying delay, if too much is added, the phase response can look smooth, but the LF could be lagging by an entire wavelength. Precise time/phase alignment can be tested by reversing polarity of the delayed system, the resulting cancellation dip will be deepest when the delay is set exactly right.

The same alignment process must be done with the upper cabinets in the crossover region to deliver "punch" and impact .
Easy as that ;).

Art

Fantastic response, I benefited from this also. Thank you very much for taking the time to explain this. I want to take the smaart seminar to learn more and do more hands on stuff, but the only class near me is booked and there are no future classes scheduled in the forceable future in my area.
Logged

ProSoundWeb Community

Re: How to measure?
┬ź Reply #5 on: September 30, 2014, 05:56:17 PM ┬╗


Pages: [1]   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.036 seconds with 21 queries.