ProSoundWeb Community

Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => Audio Measurement and Testing => Topic started by: Luis_Marquez on December 24, 2018, 02:18:02 PM

Title: How cabinets are measured
Post by: Luis_Marquez on December 24, 2018, 02:18:02 PM
New to taking measurements.I have some time this week to play around with my passive rig. Currently applying eq off ear and music source  but going to build some presets in the venue 360. My dirt backyard is ok in size. Working with REW software, berrymic, Art dual usbpre into X32rack. I’m having issues with the monitors in that don’t like the sound ,unnatural. No manufacture settings available. Goal is to have a fairly smooth passive rig and eq to taste at venue. What is the best cabinet position and mic location for taking measurements for mains, subs, and monitors?

Online search was vague. Got several articles indicating Cab on ground with mic 2m away. Does this apply to all cabs? My mains are three way so on ground in upright position or laying across?

Thanks and Mery Christmas’s to all
Title: Re: How cabinets are measured
Post by: Steve-White on December 25, 2018, 11:03:20 AM
Hello Luis.  I started a response to this and deleted it twice.  First off, my systems are all active with DSP's, everything including monitors, the DJ systems, even the system in my den that supports the TV.  The only passive stuff I have is my whole house music system and it's bi-amped with subs LOL!

Your questions cover lots of ground.  A suggestion would be to take it one step at a time - mains then monitors.  Read read read - you're off to a good start by posting here.  I'm fairly new to these forums, joined a year ago and have only put up a few posts.  The guys seem great - this is the real deal in here.

Have a look at this thread:  https://forums.prosoundweb.com/index.php/topic,169371.0.html

Read read read.  There's theory, then there's the application of the theory via knowledge gained using theory.  For any given problem, different engineers may come up with the same solution(s), sometimes not.  Each path may yield a good result, yet subtle differences.


The thread I referenced is a good read.
Title: Re: How cabinets are measured
Post by: Brian Jojade on December 26, 2018, 12:35:31 AM
Using measurement microphones to tune your system means that it will be tuned exactly where the mic was placed.  Different positions will mean different measurement results.  For FOH this can make it very difficult for a measurement mic to provide meaningful results unless you take measurements from several different positions and then come up with a compromise that makes everyone happy.

For monitors, it's a bit easier since you know roughly where the monitor will be compared to where the listener will be. And since there's only one listener, that's the only position that you really care about.

Now, you didn't specify which speakers you are dealing with.  Depending on the accuracy of the speaker, you may or may not be able to get decent sound out of it.  One major issue with speakers is that their response curves change between low and high volume.  If you use an EQ to adjust the level at low volume, then turn it up, your settings are meaningless.  Better speakers tend to have similar frequency response curves at more levels.  It's rare that you'll ever see that in any of the marketing materials though, which makes it pretty tough to know how they might react.

Like all things in audio it's going to come down to compromise.  Do as little as possible to the signal to make it sound as good as possible.  Rarely does the auto EQ's function end up being what you want the system to work like.  It's handy to give you an idea of what's happening, but when you see frequency bands next to each other at 20db discrepancies, that really says there's something either horribly wrong with the speakers (most aren't THAT bad) or the measurement system isn't doing what you need it to.  Learn to adjust by ear. It'll be much easier in the long run than trying to rely on tools that may or may not do the job!
Title: Re: How cabinets are measured
Post by: Mark Wilkinson on December 26, 2018, 12:37:37 PM
New to taking measurements.I have some time this week to play around with my passive rig. Currently applying eq off ear and music source  but going to build some presets in the venue 360. My dirt backyard is ok in size. Working with REW software, berrymic, Art dual usbpre into X32rack. I’m having issues with the monitors in that don’t like the sound ,unnatural. No manufacture settings available. Goal is to have a fairly smooth passive rig and eq to taste at venue. What is the best cabinet position and mic location for taking measurements for mains, subs, and monitors?

Online search was vague. Got several articles indicating Cab on ground with mic 2m away. Does this apply to all cabs? My mains are three way so on ground in upright position or laying across?

Thanks and Mery Christmas’s to all

Hi Luis, I continue to struggle with the questions you ask....but I guess i've come to the conclusion that getting rid of reflections is the 'name of the speaker measurement game'.

Subs are the easiest IMO,.... sub on ground, mic on ground. 
I like the 10m / higher voltage method, but will go as close as 2m if trying to measure distortion and keep noise floor low.
Otherwise 4m is the minimum.

Mains?  I think the same technique....speaker on ground, mic on ground.  I like 10m here too, but will drop to 3-4m if need be. 
I think going closer than 3m creates too much triangulation distance/differences to align HF/VHF overlap well. 
Sometimes I lay the boxes on their side, sometimes stand them up...but either way I tilt them towards perpendicular to the mic.

Never done any monitors... that might be the one time I'd be happy with box on the ground and the mic up in the air where the talent's ear is..

I hope more folks will chime in with their techniques....
Title: Re: How cabinets are measured
Post by: Luis_Marquez on December 26, 2018, 04:13:45 PM
Thanks for your replies.
Steve. The cabs to be measured are MI grade Carvin trx12n’s, trx 153n’s, and trx subs. Local bands, and church festival. Coming back after being several years out. Working on building the mix and want to get comfortable with what I have. Step up down the road. Will probably end up going powered as this has been the trend.
      I will continue read,read, read and thanks for the link. I did read it when it was first posted. Will go over it serveral time to digest. Btw wish you many more birthday cakes

Brian- That is something I did notice in walking the horizontal coverage. The response did change closer to the outer coverage. So will take multiple measurements and find a compromise.

Mark- Looks like you use 10m mic distance as reference for mains /subs. In the sm80 thread, Mr Ivan Beaver also mentioned subs on ground with mic at 10m. I remember reading somewhere that cabs were raised 12ft from the ground and measured at that height. I guess only possible If you have equipment to raise the cabs safely.
Title: Re: How cabinets are measured
Post by: Steve-White on December 30, 2018, 02:06:40 PM
Thanks for your replies.
Steve. The cabs to be measured are MI grade Carvin trx12n’s, trx 153n’s, and trx subs. Local bands, and church festival. Coming back after being several years out. Working on building the mix and want to get comfortable with what I have. Step up down the road. Will probably end up going powered as this has been the trend.
      I will continue read,read, read and thanks for the link. I did read it when it was first posted. Will go over it serveral time to digest. Btw wish you many more birthday cakes

Brian- That is something I did notice in walking the horizontal coverage. The response did change closer to the outer coverage. So will take multiple measurements and find a compromise.

Mark- Looks like you use 10m mic distance as reference for mains /subs. In the sm80 thread, Mr Ivan Beaver also mentioned subs on ground with mic at 10m. I remember reading somewhere that cabs were raised 12ft from the ground and measured at that height. I guess only possible If you have equipment to raise the cabs safely.

Ok, I kind of have a feel for what you are doing now, what you are working with.  As I stated, I now use DSP's on everything - started ~10 years ago, set up my den music system, tri-amped and never looked back.

Methodology is important leading to consistency.  I always setup a baseline on systems doing near field (1-2m) measurements on individual components with the DSP's and tweak all the crossover parameters and touch-up with the output channel parametric eq's, set limiters, etc.  Then, once I have a baseline on tops and subs separately, I set the system up as it will be used - tops up on tripods or stacked - however it will be used.

Establishing a consistent & solid baseline and tuning with a building block approach is the point here.  Rather than starting from scratch each time you setup.  Too easy to get completely lost, especially on a re-entry trajectory you are on.
Title: Re: How cabinets are measured
Post by: Luke Geis on December 30, 2018, 06:13:32 PM
Not knowing exactly how you have your system setup, I will give a few words of advice.

1. Try to measure the system as it would be as you normally set it up.

2. Measure 1 half ( or 1/3rd if going with subs on aux ) of the system at a time.

3. Keep in mind that going for a totally flat/linear response is often an effort in futility, so just focus on areas of the response that have problems. Note that large holes in the response, or commonly recurring dips and peaks may be comb filtering or reflections and cannot be fixed with EQ.

4. With monitors, I find that less is more. They are meant to be loud, not sound good. Placing the measurement mic about where the performer will be is what I do. And then I ONLY get rid of stuff that is peaking or standing out. I typically roll the lows off somewhere between 140-200hz, going for what sounds natural and NOT TOO FULL OR BOOMY. I try and not do any real corrective EQ until a problem actually arises. There is often only 1-2 frequencies that will feedback and once those are stabilized trying to go any louder is often not possible. Just because the fader is going up does not mean it is actually louder.

5. A good rule of thumb for any part of the system is that if you are using more than 6 bands of EQ, you need to stop, re-assess and see what you can do to reduce the # of cuts to keep it under 6. Once you start cutting beyond about 6 bands, you are going backward, not forwards.

I am not a huge fan of the whole baseline or pre-set EQ idea myself. Every venue and setup is just a little bit different and a linear PA in space #1 will not be linear in space #2. If there are inherent characteristics of the PA that are easily able to be heard and measured, yeah sure, nail those down, but to me measuring a PA is something done at the job site and handled on the spot. Obviously, if the PA always has a large hump around 250hz or something, that should be addressed.

The Carvin TRX stuff isn't bad by any means, but it does require proper DSP and setup to work right. My trick has been to find what works well in the system and build on that. It may not sound great, but instead of trying to make a PA something it's not, it pays to work with the strengths and build around that.
Title: Re: How cabinets are measured
Post by: Frank Koenig on December 31, 2018, 01:41:10 PM
This is a big subject that we hammer on here periodically but I'll try to "Powerpointize" my ideas on it for reference and discussion.

*Anechoic measurements are the most useful for most purposes. They can be achieved three ways:

   1. Use a good, sufficiently large-for-the-wavelength, anechoic chamber. Few of us have this luxury.

   2. Suspend both speaker and mic in the air and window the impulse response (IR) to remove reflections (pseudo free-field).
      The frequency resolution that can be achieved this way is limited by how far away from reflecting surfaces (the ground)
      we can get.

   3. Place both speaker and mic on the ground and window the IR to remove all reflections other than that from the
      ground (ground-plane). This allows greater frequency resolution, needed at low frequencies, but fails to give accurate
      results at high frequencies where the source size becomes comparable to the wavelength (typically around 400 Hz).
     This transition frequency can be determined by comparing the pseudo free-field and ground-plane measurements
      in the vicinity of the transition.

*A frequency resolution of at least 1/6 octave is necessary for most purposes. More is better -- we can always smooth.

*All measurements must be substantially in the far field of the source. As a rough guide, 3m is good for mains, 10 m is good for individual subs (not huge sub arrays).

*It is useful to statistically combine (average, for example) measurements across angles for any source that is directional. The angular spatial sampling rate needs to be considered. 5 deg is typically sufficient for mains. Subs, that are more-or-less omnidirectional, can be measured on-axis alone with good results. If settings are being developed for a set of speakers it is prudent to combine measurements across individual units. This represents a lot of work. Statistically combining multiple phase measurements is non-trivial when the variability in measurement distance becomes comparable to the wavelength.

*It may be useful to consider the total acoustic power, integrated over all angles (as we would measure in a reverberation chamber with an RTA), in addition to the sound pressure at a set of points. As most speakers have highly variable directivity at higher frequencies, this curve will look different from the within-pattern SPL response. Without a reverberation chamber (integrating sphere) it is difficult to measure but can be approximated from multiple SPL measurements. The voicing that sounds good subjectively in a reverberant space might be a compromise between SPL and total power. I think most of us start with a flat-average SPL response as our "setting" and compensate for the rest by ear using low-Q filters.

*Once you have a good setting for a system, don't let anyone, including yourself, mess with it. This is your reference or baseline. The idiosyncrasies of different rooms should be compensated by ear (or Caterpillar). This is likely somewhat controversial and perhaps not applicable to complex fixed installations. But then we're getting into the realm of room-acoustic rather than speaker measurement.

*For portable systems in the field, measurement is useful for system health check and sub alignment, and probably not much else. Sub alignment can be done with a tape measure if you know the reference delay for the system.

Personal note: Once I started using monitors that were actually dialed more-or-less flat, "ringing out" (wringing out ;) ) for GBF pretty much became a thing of the past. Never need it outdoors and very rarely indoors.

--Frank
Title: Re: How cabinets are measured
Post by: Chris Grimshaw on January 02, 2019, 06:16:38 AM

4. With monitors, I find that less is more. They are meant to be loud, not sound good.

I find good-sounding monitors mean they don't need to be that loud for the artist to hear what they need. Clarity and volume are correlated, sure, but other factors are at play.


To the OP,
Measuring speakers can be a tricky business - that's why the people that are good at it can make money out of it!

Anyway. The key is practice, and figuring out which bit of the measurement is an artifact of the environment, and which are the products of the speaker itself. After that, you'll need to learn what to do now you understand the graphs.

I'd recommend learning the basics (and testing the measurement setup) indoors where you're not pushed for time or worried about weather or annoying the locals.
Take a speaker that you don't mind abusing a little, and do some measurements. Try these:

- Close to the cone
- Around 1' away, on and off axis
- Around 10' away, on and off axis again.

Note how each measurement has an amount of "fuzz", and that "fuzziness" generally increases as you move the mic away from the speaker. That's from reflections creeping into the measurement.

Next, EQ the low-frequency response so that, when the mic's close to the cone, it measures flat (NB - best done with sealed speakers). You might need 20dB of boost, which is fine for playing around at home but should be avoided when setting up the PA system.

Now the LF response is nice and flat, take the mic out into the room again. Measure in a few different positions in the room. Note how the once-flat low-frequency response is now anything but. Also note how the response changes according to location.

It's stuff like that which we can't really EQ out. Everyone's getting a different response, so "fixing" one might make the others worse.

Sidenote - I'm quite lucky at the moment - the living room gives a peak at around 45Hz pretty consistently. Knocking that down a bit benefits pretty much every seat. Other peaks and dips do occur and I have to live with that. There's another one at 15Hz, which is useful - means my sub runs flat to 15Hz without needing loads of power to get there.


Okay, so I'll assume you've got to the point where you've measured everything you can with a small speaker indoors. You've EQ'd it, can spot reflections vs speaker effects, etc.

Lets get the PA system out.

I'm gonna run through setting up a tri-amped system (2-way mid-high, plus sub(s)). Some of this will apply, some of it won't.

LF/MF crossover is fairly easy: the cabinets are basically omnidirectional, and there's limited scope to damage anything so long as you're sensible.
Pick a crossover point that's suitable for the cabinets in play (ie, you might run a 15" down to 70Hz to meet a 15" sub that's only reinforcing a bit, or you might run that 15" down to 120Hz if it has a couple of 18"s to do the heavy lifting).
100Hz is usually a safe bet for most cabinets, but it's something you get a feel for.

Throw an LR24 crossover at it, and see what you get. LR24 crossovers are useful in a few ways (they sum flat, offer plenty of driver protection).
Next, invert either the LF or MF band, and run the sweep again. By tweaking the delays, you'll be able to get a nice deep and symmetrical notch. Set cabinets back to normal polarity and you should get a seemless blend between the cabinets.


MF/HF crossover is more tricky.
This time, you've got to consider how much stick the compression driver/horn combo can take, as well as off-axis response.
Things to consider:
- As the wavelength approaches that of the diameter of the cone, the cone speaker will start to show narrowing directivity.
- We want the off-axis response to be smooth.
- As you go higher in frequency, cone drivers will start to have a fairly rough response. Some drivers are smoother at the top end than others.
- You don't want to drive the compression driver/horn lower than the manufacturer says is safe, unless you're doing something odd. Example - I'm running a 1" driver down to 1.6kHz, but it's only keeping up with a 6" midbass. ie, it's not gonna go that loud. If you tried to run a 1" driver down to 1.6kHz and keep up with 2x15"s, you're gonna be replacing diaphragms most gigs.
- As you move the mic around, you'll be closer to one driver and further from another. Your phase curves will move accordingly.

So, a 15" 2-way cabinet with a 60-degree horn might have a higher crossover point than the same with a 90-degree horn.

Armed with that information, you might look online and find the HF unit is fine down to 1kHz, the horn loads down to 700Hz, and the midbass driver starts getting rough at 2kHz.
That means you'll have a band of 1kHz up to a little under 2kHz to get things lined up in terms of matching the directivity of the cone and horn. Matching them up nicely will get you a nice smooth off-axis response, which is a Good Thing. Have you ever listened to a 15" 2-way box with a tiny 1" HF driver? - Even if it sounds good directly in front of the cabinet, moving a couple of steps to the side will yield a very different sound.


When it comes to EQing things, here are some general guidelines: (some are more set in stone than others - it's up to you to decide which ones are more flexible in your case)

- Avoid boosting below the port tuning frequency.
- Every 3dB of boost means the driver is receiving 2x the power in that area. 6dB = 4x power, 10dB = 10x power.
- A narrow notch in the response is rarely worth the power required to boost it to flat, and usually only occurs in one mic position anyway (see note above about making sure the EQ is beneficial in more than just one seat)
- Narrow peaks sound worse than narrow notches. If some seats get a peak and others don't, it's often best to bring the peak down and give some people a notch in the response.
- Don't worry much about >15kHz. Your HF driver may roll off at 16kHz, and be 10dB down at 20kHz. At live music volumes, nobody will notice. EQing 20kHz back to flat isn't going to be noticed, but will mean the HF driver gets that bit warmer.


Apologies for the long post. I hope some of it comes in handy.

Chris

PS - +1 on everything Frank said above. Good info there.
Title: Re: How cabinets are measured
Post by: Jerome Malsack on January 10, 2019, 10:22:18 PM
http://carvinimages.com/manuals/76-15300b-trx2_new_11.11.2008.pdf

The 153 has a 60 degree x 40 degree horn  so the highs are very tight in the center.   mids and bass will be well spread unless your getting a power alley.

Title: Re: How cabinets are measured
Post by: lindsay Dean on January 11, 2019, 02:45:00 PM
I can always tell when a non-musician/vocalist is running sound when he says "the monitors are not supposed to sound good just be loud"
wait?......😳....what?
Title: Re: How cabinets are measured
Post by: Luke Geis on January 11, 2019, 09:12:34 PM
Except I am a musician and singer with 27 years experience playing in bands, et all.

Monitors can be made to sound good subjectively, but I contend that a singer that really wants A LOT will have one that doesn't sound good no matter what you do. The distortion and harmonics alone is one reason. The second is that by the time it's that loud, A LOT of stage wash from the mic will be in the speaker as well. Don't even get me started on intermodulation and its effects.

The number of tricks employed in order to make a stable monitor that gets up to a true 130db SPL is astronomical. Some of the tricks will make your head spin. The truth is, a truly loud monitor mix will not sound very good by most subjective measures. Now that isn't to say that you cannot have a loud and good sounding monitor mix, but when I say it is meant to be loud, not sound good, it is standing on the shoulders of an objective goal, not a subjective measure. A monitor mix that sounds like an angel resting its wings on your shoulders, that doesn't get loud enough, is a worthless monitor mix from an OBJECTIVE standpoint. A monitor that isn't raped to death by corrective and musical EQ, WILL be louder than one that is.

This is why I say monitors are meant to be heard, not sound good. Once you have the needed SPL from the monitor, then you can work on making it sound pretty. My experience anyway, has lead me to that fact. I will say that better monitors tend to sound better and maintain stability too. So it's not so much that no F$%*s are given about the quality of sound, but that if I have to choose between loud or pretty, I am going with loud and then see how pretty it can be made after I have the needed level.
Title: Re: How cabinets are measured
Post by: Mark Wilkinson on January 25, 2019, 12:29:22 PM
Saw Meyer's M-Noise in LSI a few days ago.   https://m-noise.org/

In the 'Real World SPL' link found down the page under M-Noise Education,
the speaker on the ground / mic on the ground technique is shown starting at about 1:40. 

I know many folks are aware of this technique....just thought the link might help make it clearer to those that aren't.

This 'lay speaker on side, tilt to perpendicular to mic pointing into hard surface ground',
has been giving me results that match pretty well indoors to outdoors....surprisingly close really.
I'd rather be more than 1m back though, as recommended in the vid...always feel like too much triangulation is going on up close for decent timing adjustments.
Title: Re: How cabinets are measured
Post by: Ivan Beaver on January 25, 2019, 03:54:46 PM
Saw Meyer's M-Noise in LSI a few days ago.   https://m-noise.org/

In the 'Real World SPL' link found down the page under M-Noise Education,
the speaker on the ground / mic on the ground technique is shown starting at about 1:40. 

I know many folks are aware of this technique....just thought the link might help make it clearer to those that aren't.

This 'lay speaker on side, tilt to perpendicular to mic pointing into hard surface ground',
has been giving me results that match pretty well indoors to outdoors....surprisingly close really.
I'd rather be more than 1m back though, as recommended in the vid...always feel like too much triangulation is going on up close for decent timing adjustments.
The idea behind M noise is a good one.  A lot of thought was put into the noise.  It makes sense to me, much more than simple pink noise. 

I am glad that Meyer is releasing the wav file and encouraging others to use it.

But putting a full range speaker on the ground is not generally what I consider the best way to measure them, it gives a false sense of the low freq output, vs up in the air where they would normally be located.

I would "argue" that different types of measurements should/could be done different ways, in order to get a real world usable measurement.

Some cabinets should be measured MUCH further away than 1m to get a usable number that will translate to a modeling program.  While other cabinets are fine at 1m.

Some cabinets will give greater than usable numbers at 1m, and other cabinets will give lower than usable numbers.  It depends on a number of different factors involved.

BTW, I did a quick measurement using the M noise and the same NTI XL2 meter on a non Meyer loudspeaker (that is suitable for being measured at 1m).

I got 9.5dB higher peak output than the normal loudspeaker specs, so they are correct, maybe the spec sheet needs to change.

Title: Re: How cabinets are measured
Post by: Mark Wilkinson on January 25, 2019, 04:23:34 PM
The idea behind M noise is a good one.  A lot of thought was put into the noise.  It makes sense to me, much more than simple pink noise. 

I am glad that Meyer is releasing the wav file and encouraging others to use it.

But putting a full range speaker on the ground is not generally what I consider the best way to measure them, it gives a false sense of the low freq output, vs up in the air where they would normally be located.

I would "argue" that different types of measurements should/could be done different ways, in order to get a real world usable measurement.

Some cabinets should be measured MUCH further away than 1m to get a usable number that will translate to a modeling program.  While other cabinets are fine at 1m.

Some cabinets will give greater than usable numbers at 1m, and other cabinets will give lower than usable numbers.  It depends on a number of different factors involved.

BTW, I did a quick measurement using the M noise and the same NTI XL2 meter on a non Meyer loudspeaker (that is suitable for being measured at 1m).

I got 9.5dB higher peak output than the normal loudspeaker specs, so they are correct, maybe the spec sheet needs to change.

In complete agreement with different measurement needs....along with exaggerated low end from mains on ground.
I guess I've just found as a simple DIY tuner type, that the both on the ground technique gives the best consistency I can achieve.
And then I can pretty simply adjust low end when on a stand or subs..
Wish I had one of your guys cranes  :)

Yep, I like the idea of M-Noise a lot too.  Makes intuitive sense. 
First thing I did was go bang some spoons taking a Smaart transfer, to see if Smaart is as fast as Meyer's SIM3 (since Meyer said SIM2 wasn't fast enough).

Title: Re: How cabinets are measured
Post by: Steve-White on March 15, 2020, 05:14:34 PM
Ok, many manufacturers specs based upon half-space.  My understanding is this is with a non-reflective baffle surrounding unit under test to create the half-space plane.

Do they operate the in true half-space or calculate it?

Can someone enlighten me on this?
Title: Re: How cabinets are measured
Post by: Luke Geis on March 15, 2020, 11:16:55 PM
I think most compute it. They take a measurement at a distance like 2 meters away and then add +6db for a 1-meter measurement, and then if calculating half-space, add yet another +6db if it isn't already actually half-space loaded. This assumes an anechoic chamber for all the measurements.

Half space calculating is a challenging question. Technically if the speaker is 50' in the sky and you take the measurement from the ground, you are getting a half-space measurement. If the speaker is on the ground and you take the measurement 50' above the speaker, again it is a half-space measurement. The ground plane doubles the energy either way. I think with most subs, a half-space measurement is calculated to better the result?

I don't agree so much with the measure at 2 meters and then scaling it back to 1 meter bit. I think it gives the measurement a +3db boost than is real. Most speakers don't start abiding by the inverse square law until around 10'-12' where they will then actually drop off per the rule. Within that 10'-12' the speaker is only losing -3db per doubling of distance ( line array type crap, yadda yadda ). So if you measure at 2 meters or roughly 6.5' away, and then add +6db, chances are good that you are adding back +3db more than is likely. I think measurements are taken at 2 meters to help acquire a more even response from the speaker measurement. At 1 meter you are pretty close to the speaker and the beaming and coverage of the woofer and horn may not have propagated or evened out much yet but at 2 meters ( a more realistic listening distance ) at least the measurement will be a little more telling of what you would actually hear. The likely +3db boost the stats get is just a plus.
Title: Re: How cabinets are measured
Post by: Steve-White on March 15, 2020, 11:52:08 PM
^^^ Thanks Luke, that's helpful.

FWIW:  My employer has a chamber - I figured it was electrically "quiet", isolated and such.  Little did I know, it's all that and is also acoustically "quiet".  Large enough to hang an aircraft on straps to emulate flying and run the avionics for testing.  Real strange experience being inside there - real quiet, finger foam all around.  It was great - wish I had access to it for some testing.

Anyway, I have some testing to do and it will probably be above 250hz on out.

Considering setup.  I'll have to play around - will be outside in backyard.  Considering throwing a moving quilt on the lawn and laying cabinets on their backs, measuring from above.  Should be fun, neighbors outta love it.
Title: Re: How cabinets are measured
Post by: Luis_Marquez on March 16, 2020, 01:57:47 AM
I think most compute it. They take a measurement at a distance like 2 meters away and then add +6db for a 1-meter measurement, and then if calculating half-space, add yet another +6db if it isn't already actually half-space loaded. This assumes an anechoic chamber for all the measurements.

Half space calculating is a challenging question. Technically if the speaker is 50' in the sky and you take the measurement from the ground, you are getting a half-space measurement. If the speaker is on the ground and you take the measurement 50' above the speaker, again it is a half-space measurement. The ground plane doubles the energy either way. I think with most subs, a half-space measurement is calculated to better the result?

I don't agree so much with the measure at 2 meters and then scaling it back to 1 meter bit. I think it gives the measurement a +3db boost than is real. Most speakers don't start abiding by the inverse square law until around 10'-12' where they will then actually drop off per the rule. Within that 10'-12' the speaker is only losing -3db per doubling of distance ( line array type crap, yadda yadda ). So if you measure at 2 meters or roughly 6.5' away, and then add +6db, chances are good that you are adding back +3db more than is likely. I think measurements are taken at 2 meters to help acquire a more even response from the speaker measurement. At 1 meter you are pretty close to the speaker and the beaming and coverage of the woofer and horn may not have propagated or evened out much yet but at 2 meters ( a more realistic listening distance ) at least the measurement will be a little more telling of what you would actually hear. The likely +3db boost the stats get is just a plus.

Luke. Thanks and to others sharing. I have a spot made out for testing in my dirt back yard. Their is a minimum of 30 ft away from any object. So now gong to try “measurements done in different ways”. Speaker on 4ft scaffold mic on ground @ 2m, speaker on ground sideways mic @2m, speaker on ground cones pointing to sky mic on 50ft Pvc pipes pointing down to speaker. REW software and waiting for a umc202hd interface.
Do I run on default REW settings or can someone share some info? Can m-noise be set for use on REW?

I am very green at this measurement process. So my goal is to get comfortable with the process and understand traces. I have several passive speakers in figuring out some odd beaming or harshness. So far unsuccessful in pinpoint where the problem lies. Smaart is not in my budget and don’t want to burn the free trial with in unpredictable weather and limited free time.
Title: Re: How cabinets are measured
Post by: Chris Grimshaw on March 16, 2020, 04:50:36 AM
REW is fine. I've been using it to set up my systems for years.

REW uses a swept sine sweep. There are lots of good reasons for that - check out the Help file for a lot of good info.


You'll need drivers for the UMC202HD - download from the Behringer website.
You can see the config page in REW here: https://youtu.be/uD9HBQ6lSEo?t=43

If your speakers are passive and have uneven directivity, you might find you need to alter the crossover. Not too difficult for boxes with active crossovers, but if there are passive crossovers inside the cabinets, it'll be more tricky.

Chris
Title: Re: How cabinets are measured
Post by: Tim McCulloch on April 26, 2020, 11:55:13 AM

Considering setup.  I'll have to play around - will be outside in backyard.  Considering throwing a moving quilt on the lawn and laying cabinets on their backs, measuring from above.  Should be fun, neighbors outta love it.
Thirty five years ago, that's how Community measured their horns and loudspeakers.  The had the measurement mic on a 'fish pole' attached to a TV antenna rotater, all mounted to a 20 ft tower, IIRC.  They gathered the data on a 1/3 octave RTA and took photos of each step.  The data was manually plotted from the screen shots.