Why Do TEF Systems read 3-5 dB high ?

Re: Why Do TEF Systems read 3-5 dB high ?

Brad Weber said:
Unless they're skydiving. Maybe that's how you could get valid whole space measurements! Ivan, you need to take up skydiving and get Mike to buy a plane.
Click to expand...

OK so let's assume we get "proper" half space measurements. So what?

If nobody else has them-then there is nothing to compare to.

That is part of the problem with providing more accurate data-unless others are doing the same-it is hard to tell if one product is actually better or worse than another.

I have noticed that some people love to critique a measurement that offers detail, but just give other products a "free pass" because they do not offer any data. So somehow that means that the other products simple single numbers give all the information needed-NOT.
 
Re: Why Do TEF Systems read 3-5 dB high ?

Ivan Beaver said:
OK so let's assume we get "proper" half space measurements. So what?

If nobody else has them-then there is nothing to compare to.

That is part of the problem with providing more accurate data-unless others are doing the same-it is hard to tell if one product is actually better or worse than another.

I have noticed that some people love to critique a measurement that offers detail, but just give other products a "free pass" because they do not offer any data. So somehow that means that the other products simple single numbers give all the information needed-NOT.
Click to expand...
Ivan,

I have assumed the old standards I grew up with, JBL and Electro-Voice to have “proper” half space measurements, as their charts agree well to real world testing.

I applaud DSL for giving phase and impedance charts in addition to frequency response charts. I had also assumed, until recent close inspection and comparison, that DSL would also be among the “old standards” where what you see is what you get (if you can see beyond marketing hype).

That said, the fact that a DSL cabinet when measured against an EV cabinet using TEF shows the DSL relative reading 3 dB lower than the published chart spec makes me wonder why that should be so.

The fact that B&C publishes an Ap graph showing a sensitivity of around 95 dB for their BC18SW115, which agrees with their cabinet simulation, then posts a TEF reading that is about 3 dB higher also makes me wonder why.
The fact they rate the same speaker as 97 dB sensitivity when it only reaches that level at about 1200 Hz simply annoys me, but such is the world of marketing.

Everyone knows if you are worried about 3 dB, you don’t have enough “rig for the gig”, but a 3 dB rated difference in sensitivity turning out to be the same is the same as if you looked at a car’s specifications and one is rated for 15 MPG (miles per gallon, your milage may vary..), another 30 MPG, then you find they actually get the same MPG.

The fact my Mustang gets about 25 MPG instead of it's 28 MPG rating can easily be explained by my lead foot and local mountainous conditions.

If it only got 14 MPG, I would wonder why .

Still hoping you (or anyone else reading this) will do a side by side half space comparison of a DSL sub to some other “old standard” to double check my observations.

Art Welter
 
Re: Why Do TEF Systems read 3-5 dB high ?

Jeff Permanian said:
It's amazing how inexpensive technology has become:

http://www.bhphotovideo.com/c/produ...Audio_CM_C200_CM_C200_Calibrator_for_SPL.html
Click to expand...
Jeff,
Phil Lewandowski’s measurements of his Growler and his JBL SRX 718 were remarkably similar to each other, showing the Growler is within a dB of the same sensitivity as the 718, which has a one watt one meter sensitivity of 95 dB.
The 2007 Prosound Shootout also shows the Growler to be about 95 dB average sensitivity 45 to 100 Hz.

Yet you rate the Growler at 100 dB sensitivity.

Amazing.
 
Re: Why Do TEF Systems read 3-5 dB high ?

Art Welter said:
Jeff,
Phil Lewandowski’s measurements of his Growler and his JBL SRX 718 were remarkably similar to each other, showing the Growler is within a dB of the same sensitivity as the 718, which has a one watt one meter sensitivity of 95 dB.
The 2007 Prosound Shootout also shows the Growler to be about 95 dB average sensitivity 45 to 100 Hz.

Yet you rate the Growler at 100 dB sensitivity.

Amazing.
Click to expand...

I have still not had a chance to do the measuremets-I have been working 14-16 hr days and there simply is not time.

HOWEVER-here are some thoughts. It all of a sudden hit me-while working on another project. If the 718 is measured in whole space (as I understand it the jury is still "out" on that one-then you add 6dB to 95 and you are within a dB of the Growler.

However that is not your initial compairison. As I understand it you are basing this mostly on the EV218 and the TH115's that were at the NY sub shootout.

Here is a problem with measuring horns or cabinets that are physically deep. If the measurement is done @1m from the front grill (as was done in NY so that the room would have less of an influence on the overall sound/response of the cabinets) Inverse square law will greatly affect the measurements.

Danley does the sub measurements at 10M to minimize this effect. Basically what happens is that if you start out at a large distance (let's say 100') away from the cabinet and then start walking towards them (lets say a front loaded cabinet and a horn), you will notice something interesting.

Initally the level will drop the same for both cabinets at the same distance away. However as you get closer the level will continue to rise as you get closer to the font loaded box. However as you get closer to the horn, the level will not rise as much.

The reason being that the length of the horn puts the acoustic origion much further back, behind the cabinet-so inversquare is not having as much of an effect.

So if you had cabinets that were producing the same SPL at say 10M, and then you measured them at 1M, the front loaded cabinet would measure higher than the horn-because you are much closer to the acoustical origion of that cabinet..

I am suspecting that is what is causing at least some (if not all) of the error that you noticed.

Part of the "work" in trying to measure loudspeakers is not having anything influence the measurement. That is why they are measured at a distance and calculated to the 1m measurement.

Another good example is our SH25. If you measure a 1M it is not as loud as our specs say. However a independant lab measures the sensitivity the same as we do-when measured at a distance and then back calculated to 1M. This is due to the depth of the cabinet-which places the acoustic origion a good bit further away than the 1M meter measurement would suggest. But at a distance (and that cabinet is designed for long throw applications) the numbers work out just fine.

The whole idea of providing specs is to try and give the customer an idea of how well the loudspeaker will work in a normal room. And in most cases (short of the home) people are not listening to loudspeakers at a meter or two.

And yes I will measure the EAW and something comparable in SPL ratings to see the difference. But I have to clear some time out first.
 
Re: Why Do TEF Systems read 3-5 dB high ?

Ivan Beaver said:
I have still not had a chance to do the measuremets-I have been working 14-16 hr days and there simply is not time.

HOWEVER-here are some thoughts. It all of a sudden hit me-while working on another project. If the 718 is measured in whole space (as I understand it the jury is still "out" on that one-then you add 6dB to 95 and you are within a dB of the Growler.

However that is not your initial compairison. As I understand it you are basing this mostly on the EV218 and the TH115's that were at the NY sub shootout.

Here is a problem with measuring horns or cabinets that are physically deep. If the measurement is done @1m from the front grill (as was done in NY so that the room would have less of an influence on the overall sound/response of the cabinets) Inverse square law will greatly affect the measurements.

Danley does the sub measurements at 10M to minimize this effect. Basically what happens is that if you start out at a large distance (let's say 100') away from the cabinet and then start walking towards them (lets say a front loaded cabinet and a horn), you will notice something interesting.

Initally the level will drop the same for both cabinets at the same distance away. However as you get closer the level will continue to rise as you get closer to the font loaded box. However as you get closer to the horn, the level will not rise as much.

The reason being that the length of the horn puts the acoustic origion much further back, behind the cabinet-so inversquare is not having as much of an effect.

So if you had cabinets that were producing the same SPL at say 10M, and then you measured them at 1M, the front loaded cabinet would measure higher than the horn-because you are much closer to the acoustical origion of that cabinet..

I am suspecting that is what is causing at least some (if not all) of the error that you noticed.

Part of the "work" in trying to measure loudspeakers is not having anything influence the measurement. That is why they are measured at a distance and calculated to the 1m measurement.

Another good example is our SH25. If you measure a 1M it is not as loud as our specs say. However a independant lab measures the sensitivity the same as we do-when measured at a distance and then back calculated to 1M. This is due to the depth of the cabinet-which places the acoustic origion a good bit further away than the 1M meter measurement would suggest. But at a distance (and that cabinet is designed for long throw applications) the numbers work out just fine.

The whole idea of providing specs is to try and give the customer an idea of how well the loudspeaker will work in a normal room. And in most cases (short of the home) people are not listening to loudspeakers at a meter or two.

And yes I will measure the EAW and something comparable in SPL ratings to see the difference. But I have to clear some time out first.
Click to expand...

Ivan,

As far as I’m concerned, the jury is not out regarding whether JBL’s SRX 718 or 728 are half space measurements, it is confirmed with JBL’s measurements of the same speaker used in the SRX cabinets measured in half space on the roof of one of their buildings.
I have also measured the 728 side by side with my ported dual Lab 12, the relative measurements confirm that the JBL measurements are half space, or that my 2x12 cabinet has magically gained 6 dB over what Eminence sensitivity specifications show it to be.

Regarding near field vs. far field differences in measurement, this effect may be noticeable at one meter if the horn mouth was significantly larger than the front loaded speaker’s cone area, say a 4 x4 foot horn mouth compared to a single front load 18.

But the TH-115 has a similar mouth area to the EV2118 cone area.

Having done a quick test with horn loaded vs. BR cabinets to see if there was a change at distance relative to near feild results, using relatively small cabinets, the one meter results held up to far field results. Phil Lewandowski repeated my experiment in better conditions (no nearby buildings), comparing a Growler to a JBL SRX 718, his test confirmed that both cabinets followed the inverse square law near perfectly from one meter to 30 meters.

I do agree that a comparative testing of subs (especially large units) at 10 meters is better than one meter, looking forward to those tests when you get a chance.

Art Welter
 
Re: Why Do TEF Systems read 3-5 dB high ?

This thread is a little old now and my comment is a little bit ancillary but I still want to bring it up while thinking about it. Please be aware that the 8 ohm and 4 ohm versions of the B&C woofers are not necessarily identical except for the "ohmage". I've done some modeling with the B&C specs for the 4 ohm and 8 ohm versions of the 12TBX100 driver. The 4 ohm is actually somewhere between 4 and 8 ohms based on dcr alone and although very similar in it's overall behavior I don't get 3 db more sensitivity out of it compared to the 8 ohm version. A similar thing exists with the 8 and 16 ohm versions of some of the compression drivers. What I've heard is that instead of changing the total number of windings (which would change the moving mass among other things) they actually use a different kind of wire that has a different resistance.
Besides this, I too have been baffled at the high sensitivity of the DSL graphs compared to rough models I've worked on. I have mostly assumed that it's my models that are inaccurate. Although, Tom measured a box of mine once with his TEF and it seemed to match up pretty well with my own measures and models of the box. Maybe I need to dig that graph out again and make sure I'm not imagining this.

-drew
 
Re: Why Do TEF Systems read 3-5 dB high ?

drew gandy said:
This thread is a little old now and my comment is a little bit ancillary but I still want to bring it up while thinking about it. Please be aware that the 8 ohm and 4 ohm versions of the B&C woofers are not necessarily identical except for the "ohmage". I've done some modeling with the B&C specs for the 4 ohm and 8 ohm versions of the 12TBX100 driver. The 4 ohm is actually somewhere between 4 and 8 ohms based on dcr alone and although very similar in it's overall behavior I don't get 3 db more sensitivity out of it compared to the 8 ohm version. A similar thing exists with the 8 and 16 ohm versions of some of the compression drivers. What I've heard is that instead of changing the total number of windings (which would change the moving mass among other things) they actually use a different kind of wire that has a different resistance.
Besides this, I too have been baffled at the high sensitivity of the DSL graphs compared to rough models I've worked on. I have mostly assumed that it's my models that are inaccurate. Although, Tom measured a box of mine once with his TEF and it seemed to match up pretty well with my own measures and models of the box. Maybe I need to dig that graph out again and make sure I'm not imagining this.

-drew
Click to expand...
Yes the numbers are not exact 4 or 8 or 16 ohms. Like the rest of us they are choosing a number that is close to the "standards".

That is why measurements need to be expressed in terms of VOLTAGE-no wattage and an impedance CURVE given-not just a single number.

When the standard TH115 went from 8 to 4 ohms, the actual impedance of the cabinet changed to 5 ohms. But it is called a 4 ohm box, because (believe it or not) if it was called a 5 ohm box, we would get lots of calls from people who could not find a spec on any amp to drive a 5 ohm box. YES- this actually happened on SH50 when we used to call it a 6 ohm box. I got tired of the calls and we now call it a 4 ohm box and nobody calls about that anymore.

So the sensitivity did not go up 3dB when going from the 8 to the "4" ohm driver. But it is close.

AGAIN, one of the reasons you have to look at bit closer at the specs-and the manufacturer has to provide enough of those real specs for a person to figure out if the product really does what it says or if it is right for an intended usage in a project.

You can't express complicated specs as simple numbers, so the closest simple number is usually choosen to keep it simple for the people who don't know enough to read the real specs.

Sad but true.
 
Re: Why Do TEF Systems read 3-5 dB high ?

drew gandy said:
I've done some modeling with the B&C specs for the 4 ohm and 8 ohm versions of the 12TBX100 driver. The 4 ohm is actually somewhere between 4 and 8 ohms based on dcr alone and although very similar in it's overall behavior I don't get 3 db more sensitivity out of it compared to the 8 ohm version. A similar thing exists with the 8 and 16 ohm versions of some of the compression drivers. What I've heard is that instead of changing the total number of windings (which would change the moving mass among other things) they actually use a different kind of wire that has a different resistance.

-drew
Click to expand...

Many times the rated impedance of a driver is modified by changing both the wire size (resistivity) and the number of windings (total length) for the voice coil. The total resistance is the resistivity, measured in ohms/meter, multiplied by the total length. An 8 ohm driver can be changed into a 16 ohm driver by using smaller wire and more windings. Both of these changes increase DCR. The small wire size has less mass per meter so that the longer total length for the 16 ohm VC can often come close to the same mass as the 8 ohm VC.
 
Re: Why Do TEF Systems read 3-5 dB high ?

Quote Originally Posted by drew gandy
This thread is a little old now and my comment is a little bit ancillary but I still want to bring it up while thinking about it. Please be aware that the 8 ohm and 4 ohm versions of the B&C woofers are not necessarily identical except for the "ohmage". I've done some modeling with the B&C specs for the 4 ohm and 8 ohm versions of the 12TBX100 driver. The 4 ohm is actually somewhere between 4 and 8 ohms based on dcr alone and although very similar in it's overall behavior I don't get 3 db more sensitivity out of it compared to the 8 ohm version. A similar thing exists with the 8 and 16 ohm versions of some of the compression drivers. What I've heard is that instead of changing the total number of windings (which would change the moving mass among other things) they actually use a different kind of wire that has a different resistance.
Besides this, I too have been baffled at the high sensitivity of the DSL graphs compared to rough models I've worked on. I have mostly assumed that it's my models that are inaccurate. Although, Tom measured a box of mine once with his TEF and it seemed to match up pretty well with my own measures and models of the box. Maybe I need to dig that graph out again and make sure I'm not imagining this.

-drew
Ivan Beaver said:
Yes the numbers are not exact 4 or 8 or 16 ohms. Like the rest of us they are choosing a number that is close to the "standards".

That is why measurements need to be expressed in terms of VOLTAGE-no wattage and an impedance CURVE given-not just a single number.

When the standard TH115 went from 8 to 4 ohms, the actual impedance of the cabinet changed to 5 ohms. But it is called a 4 ohm box, because (believe it or not) if it was called a 5 ohm box, we would get lots of calls from people who could not find a spec on any amp to drive a 5 ohm box. YES- this actually happened on SH50 when we used to call it a 6 ohm box. I got tired of the calls and we now call it a 4 ohm box and nobody calls about that anymore.

So the sensitivity did not go up 3dB when going from the 8 to the "4" ohm driver. But it is close.

AGAIN, one of the reasons you have to look at bit closer at the specs-and the manufacturer has to provide enough of those real specs for a person to figure out if the product really does what it says or if it is right for an intended usage in a project.

You can't express complicated specs as simple numbers, so the closest simple number is usually choosen to keep it simple for the people who don't know enough to read the real specs.

Sad but true.
Click to expand...
What is "sad but true" is most of the number chasing in this post stems from the fact that though the TH-18 using the 18Sound 18NLW9600-4 (2.5 ohm Re) was tested for impedance, sensitivity, and frequency response, when the higher impedance BC18SW115-4 (3.3 ohm Re) replaced the 18Sound three years ago DSL did not update the TH-118 specifications.

I had assumed DSL had published the specifications for the higher impedance, lower sensitivity B&C driver used in the TH-118, but recently discovered that was not the case.
http://soundforums.net/basement/7757-4-july-hi-fi.html#post61388
Same appears to apply to the DBH-218, assuming it does not currently use the 18NLW9600-4.

Mystery solved, only took two years...
 
Top Bottom