Hi all.
I have built a home PA system from the ground up and I am wondering how to best go about objectively measuring the output of the system.
A have a reasonable background in electrical engineering and audio, but I have never embarked on any kind of measurement project.
I am guessing I need a measurement mic of some kind? What else? What software? What procedure?
Many Thanks,
Psin3t
I have built a home PA system from the ground up and I am wondering how to best go about objectively measuring the output of the system.
A have a reasonable background in electrical engineering and audio, but I have never embarked on any kind of measurement project.
I am guessing I need a measurement mic of some kind? What else? What software? What procedure?
Many Thanks,
Psin3t
You could use a Sabine Work Station with a calibrated Beyringer Electret microphone for accurate result as an SPL measurement. Stand the microphone 1metre away but inline with each speaker and inject white noise through your system. Repeat for all loudspeakers. The values will differ slightly depending on the surrounding area, room, outside due to reflections.
Or,
Check out the SPL sensitivity of each loudspeaker element, manufacturers specification sheet and multiply by the power feeding them, in Watts.
Or,
Check out the SPL sensitivity of each loudspeaker element, manufacturers specification sheet and multiply by the power feeding them, in Watts.
The most straightforward answer I discovered in my research:
Purchase and download AudioTool for Android or iPhone (<$10).
Purchase a Dayton Audio IMM-6 Calibrated Measurement Microphone (<$50).
Download the microphone calibration file for your mic serial number.
Rename it to *.cal and load the file onto your phone AudioTool directory.
In AudioTool, select the "Load Cal" option from the Menu, and choose the Dayton calibration file from the list. The calibration data will be loaded, power summed to the 1/3 octave bins used by AudioTool, and saved in your Preferences - there is no further need to load the file whenever you start AudioTool, unless you change the calibration method or alter the calibration values manually. Insert microphone.
You are now setup for semi-pro realtime audio analysis of an SPL up to 127 dB for 1% THD @ 1000 Hz.
For under $60.
http://www.daytonaudio.com/index.php/imm-6-idevice-calibrated-measurement-microphone.html
https://play.google.com/store/apps/details?id=com.julian.apps.AudioTool&hl=en
You are welcome, Psin3t.
Psin3t.
Purchase and download AudioTool for Android or iPhone (<$10).
Purchase a Dayton Audio IMM-6 Calibrated Measurement Microphone (<$50).
Download the microphone calibration file for your mic serial number.
Rename it to *.cal and load the file onto your phone AudioTool directory.
In AudioTool, select the "Load Cal" option from the Menu, and choose the Dayton calibration file from the list. The calibration data will be loaded, power summed to the 1/3 octave bins used by AudioTool, and saved in your Preferences - there is no further need to load the file whenever you start AudioTool, unless you change the calibration method or alter the calibration values manually. Insert microphone.
You are now setup for semi-pro realtime audio analysis of an SPL up to 127 dB for 1% THD @ 1000 Hz.
For under $60.
http://www.daytonaudio.com/index.php/imm-6-idevice-calibrated-measurement-microphone.html
https://play.google.com/store/apps/details?id=com.julian.apps.AudioTool&hl=en
You are welcome, Psin3t.
Psin3t.
Last edited:
For a PA system, I would have thought a more appropriate set of measurements would be:
Does the system destroy itself when I apply various forms of abuse?
Does the system shut down minimally to protect itself when abused?
Down the system restart automatically when the abuse/r is removed?
Survival to keep the audience entertained/informed is surely way above any other performance requirement.
Is voice intelligible when required?
Is the audio output entertaining?
Is it loud enough to be heard at the far end of the auditorium?
Is it too loud at the near end that might attract the attentions of environmental heath operatives?
I would think that the minutae of frequency response comes way down the list of requirements.
Does the system destroy itself when I apply various forms of abuse?
Does the system shut down minimally to protect itself when abused?
Down the system restart automatically when the abuse/r is removed?
Survival to keep the audience entertained/informed is surely way above any other performance requirement.
Is voice intelligible when required?
Is the audio output entertaining?
Is it loud enough to be heard at the far end of the auditorium?
Is it too loud at the near end that might attract the attentions of environmental heath operatives?
I would think that the minutae of frequency response comes way down the list of requirements.
Last edited:
My initial suggestion for measuring a PA system would be to take it outdoors away from buildings for measurement because in general the system would be used in many different environments so initially you want to correct response anomolies with the speakers themselves. But in this case if this system will never move then you can do it in place, and a nice tool set to accomplish that is a software package called Room EQ Wizard, a calibrated test mic like the Dayton, and a USB interface with phantom power for the mic.
With that you start the process by setting crossovers and time and phase align them as best you can.. this is where a PA processor comes in, and then you tackle amplitude response anomolies.
With that you start the process by setting crossovers and time and phase align them as best you can.. this is where a PA processor comes in, and then you tackle amplitude response anomolies.
When you say measuring output, that can mean 2 things really:
Output as in obtaining information on how loud it can go, which is a very complicated question, or you're asking literally how to perform measurements of the sound output of the system.
Assuming you mean the former, I think you can sort most speakers into two groups: those that can sustain damage from overpowering before they sound unacceptably bad, and those that sound unacceptably bad before sustaining damage from overpowering. A lot of prosound subwoofers will cook or shred themselves before they sound downright objectionable, whereas I think a lot of tops will sound pretty terrible before they get damaged.
Then you set limiters accordingly. RMS limiters are great for things like subs if you have equipment that offers that functionality. Peak limters to protect CD diaphragms, etc. Then define what distortion level you would consider unacceptable. Understand that THD% is not a measure of how objectionable distortion to a listener. You can have two speakers with flat frequency response and broadband 10% THD distortion, and one can sound pleasant and the other downright ugly. It all depends on how that distortion manifests and in what ways the output is deviating from the input.
Also understand that depending on the music you play, your potential output can be severely reduced. EDM type music requires a shitload more headroom in the bass and subbass. If you don't have enough subwoofers, your overall volume is limited by your subwoofers. Compare that to other genres like rock where the entire mix might be hipassed at 55hz by your mix engineer and you might, for example, be limited by how loud the compression drivers can go before they get shrill or screechy or however you want to describe the awful sounds they can make.
I'm a mechanical engineer, and I like specifications, but there's a reason the sound industry has the sayings about bringing "enough rig for the gig". Output potential of a sound system is an incredibly complicated question, so there is no simple way to distill an answer. It's far better to get a feel for your system's capabilities and realize that if you have any doubts about its ability to get loud enough at clean levels for a certain task, it might not be suitable.
Output as in obtaining information on how loud it can go, which is a very complicated question, or you're asking literally how to perform measurements of the sound output of the system.
Assuming you mean the former, I think you can sort most speakers into two groups: those that can sustain damage from overpowering before they sound unacceptably bad, and those that sound unacceptably bad before sustaining damage from overpowering. A lot of prosound subwoofers will cook or shred themselves before they sound downright objectionable, whereas I think a lot of tops will sound pretty terrible before they get damaged.
Then you set limiters accordingly. RMS limiters are great for things like subs if you have equipment that offers that functionality. Peak limters to protect CD diaphragms, etc. Then define what distortion level you would consider unacceptable. Understand that THD% is not a measure of how objectionable distortion to a listener. You can have two speakers with flat frequency response and broadband 10% THD distortion, and one can sound pleasant and the other downright ugly. It all depends on how that distortion manifests and in what ways the output is deviating from the input.
Also understand that depending on the music you play, your potential output can be severely reduced. EDM type music requires a shitload more headroom in the bass and subbass. If you don't have enough subwoofers, your overall volume is limited by your subwoofers. Compare that to other genres like rock where the entire mix might be hipassed at 55hz by your mix engineer and you might, for example, be limited by how loud the compression drivers can go before they get shrill or screechy or however you want to describe the awful sounds they can make.
I'm a mechanical engineer, and I like specifications, but there's a reason the sound industry has the sayings about bringing "enough rig for the gig". Output potential of a sound system is an incredibly complicated question, so there is no simple way to distill an answer. It's far better to get a feel for your system's capabilities and realize that if you have any doubts about its ability to get loud enough at clean levels for a certain task, it might not be suitable.
Last edited:
Are you going for a system in a particular space measurement, or a free-field, system alone check? There are lots of different software solutions for the former, as it is a standard problem, and they all lie. They lie for perfectly good reasons, not like the cheap systems' vast exaggeration of their capacities, but if you move the test microphone a foot, interesting differences creep in - multiplexing mics and doing an average over the hall is just a different lie, if a somewhat more realistic one.
In the case of the latter, you are not going to be able to get an anechoic chamber big enough to handle bass frequencies. Below - shall we say 80 Hz, folks? - are going to completely outweigh speaker characteristics. Just sitting a cabinet on the ground, or on a stand holding it a metre off completely modifieshow it performs 'down there'. I do recommend a fresh snowfield with a foot deep powder, aiming at snow covered fir trees as a decent equivalent to a chamber, but getting the gear in in the evening is quite disagreeable, and various crows and things can really mess up your measurements (I got a herd of Swiss cows with cowbells who were curious, and came and investigated, despite me explaining to them in three languages they shouldn't be out in such weather). Getting loud enough that you can ignore wildlife risks avalanche.
If you can cadge time on a chamber - university or manufacturer, probably - they'll already have all their gear set up, and be better at using it (normally) than you are likely to get in your project. Even without the subwoofer frequencies, knowing how smooth the comprehension frequencies are can be very useful.
In the case of the latter, you are not going to be able to get an anechoic chamber big enough to handle bass frequencies. Below - shall we say 80 Hz, folks? - are going to completely outweigh speaker characteristics. Just sitting a cabinet on the ground, or on a stand holding it a metre off completely modifieshow it performs 'down there'. I do recommend a fresh snowfield with a foot deep powder, aiming at snow covered fir trees as a decent equivalent to a chamber, but getting the gear in in the evening is quite disagreeable, and various crows and things can really mess up your measurements (I got a herd of Swiss cows with cowbells who were curious, and came and investigated, despite me explaining to them in three languages they shouldn't be out in such weather). Getting loud enough that you can ignore wildlife risks avalanche.
If you can cadge time on a chamber - university or manufacturer, probably - they'll already have all their gear set up, and be better at using it (normally) than you are likely to get in your project. Even without the subwoofer frequencies, knowing how smooth the comprehension frequencies are can be very useful.
White noise? Seems like a bad idea. Pink noise, OK.
As stated in many of the posts above, so you know what it is you are trying to measure? You can get a measurement mic for not a lot of money, and some of the software is free. You'll need a soundcard that has XLR input and phantom power for most measurement mics. The little iPhone mic sold by Parts Express is not bad, and there are some good iPhone apps.
Again, what is it you want to know about your system?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.