ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  
Pages: 1 2 3 [4] 5 6 ... 16   Go Down

Author Topic: Prosound Shootout 2007  (Read 49337 times)

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 9010
Re: Prosound Shootout 2007-Test Procedures-Impedance questions
« Reply #30 on: August 02, 2007, 01:15:03 PM »

Herein lies the problem with you collecting usable data.

I have NO problem with measuring impedance-assuming that the noise level aroudn the cabinets is quite-that will affect the impedance reading-sometimes quite a bit.  But it is what you are trying to do with that data that I have a problem with.

The way you want to do it (using the minimum impedance) is out of industry standards, the ONLY thing you could even hope to compare would be between boxes measure at your shootout.  Other boxes measured in different ways have no corrilation to your results-except maybe actual freq response, but with no absolute value assigned regarding sensitivity, becuase of the drive level you have choosen.

What if a particular box has one little spot that it is a good bit lower than the rest of the box-jsut for argument sake.  You choose that value as the impedance and figure the voltage from that.  But that is only a very small area of the freq response and not representative of the overall impedance of the cabinet.  So therefore you have crippled that box in it's sensitivity.

Yes I totally agree that the rated impedances are just arbritrary numbers, but at least there is some coorilation to the actual impedance graph, but you have to have some standards.  Just ask how people feel about a 6 ohm loudspeaker and they will usually ask what is wrong with it so that it doesn't have a "standard" value.  It is usually so ocnfusing for them to think outside of their little boxes of 4-8 16 ohms.

Yes as a general rule lower impedance boxes have the "power" advantage, but that is easy enough figured out after the measurement by adding or subtracting  a couple of dB.

Your way, you have no standard at all-every box will probably be driven with a different voltage-so trying to make any real comparisoms is going to be quite difficult.  Lets see- box A was driven with 1.2dB higher voltage based on it's higher impedance and box C is driven with .6dB less because of it's lower impedance-so which box has the highest sensitivity?

And add to that the real need to do all the back calculating- Wayne says the boxes lowest impedance is 6.8 ohms, but over most of the freq range it is around 8.4 ohms.  So therefore if I want to figure out what the rest of the world would have done is now I figure out what the drive voltage difference between the different impedances  (or using the manufacturers rating of 8 ohms) actually is and then convert that to dB and add that to the measurements taken and now I have a cure that would respond to 2.83V drive.

That seems like a lot more extra work to me.

What I really wonder is why are you doing it this way, when the rest of the world is not.  Are you really getting "better" data?  Why not use industry standards.  Anyway you look at it, you are going to have to convert something to something else to get some particular data that you are looking for, so keep it standard. 2.83V.

The real problem is that most of the people in our industry have no idea what we are even talking about and they just want a single number to give them all the answers, and it can't be done.  So why try a new and different way to get the data?

I just wonder how many other people (especially manufacturers)feel that using the lowest point on an imedance curve is the proper way to figure out the drive voltage to arrive at the sensitivity?  Yes it is A way, but I don't think it is the best or easiest way to understand what is really going on.  Nobody (to my knowledge) at the NY shootout had a problem with the way Mark (and myself) were going about doing it.

Heck, most of the industry can't even agree on where the -3dB point is on a response graph.  Just go look at their measured responses and the printed numbers on the specs and you will see that many manufacutrers claim a -3dB point (freq wise) that is really 9dB down from the rated sensitivity number.   Where did that come from-I know exactly where it comes from-but that is a different rant.

If you DO NOT tie your sensitivity number to the -3dB FROM THAT NUMBER, the specs are pretty much worthless-but don't tell that to the buying public who believe everything they read! Laughing


Logged
For every complicated question-there is a simple- easy to understand WRONG answer.

Can I have some more talent in the monitors--PLEASE?

Ivan Beaver
dB Audio & Video Inc.
Danley Sound Labs

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 9010
Re: Prosound Shootout 2007-Testing ideas-wasting time
« Reply #31 on: August 02, 2007, 01:20:13 PM »

Power compression is set by the length (and level) of the signal, while max peak output is set by amplifier max power.
See this for some more info.

http://srforums.prosoundweb.com/index.php/m/231622/6883/#msg _231622
Logged
For every complicated question-there is a simple- easy to understand WRONG answer.

Can I have some more talent in the monitors--PLEASE?

Ivan Beaver
dB Audio & Video Inc.
Danley Sound Labs

Tom Danley

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 500
Re: Prosound Shootout 2007-Test Procedures
« Reply #32 on: August 02, 2007, 01:49:11 PM »

“You and Mark Seaton are both personal friends and business associates with Tom Danley. The fact that the data was gathered, organized, processed and published by you and Tom made some people pretty uncomfortable. They told me so.”

Well Wayne, insinuation’s eh?

Ok so after quite some time had passed, the NY shoot out long over, I got a copy of the raw Data from the shootout, printed (raw) images of the curves and produced impedance compensated curves with a detailed explanation.  
After the LAB didn’t put it up, we put them and the raw data on our web site.
Well (thinking of Steve Martin) EXCUUUUSE ME.

If you were more familiar with the TEF or how it worked, what it did, maybe you wouldn’t feel like that was cheating.   Maybe your right, continuing on with no information would be better in your case, or curves like yours last year where you have to figure out what it is and re-plot it on a normal scale.
With the raw TEF data provided too, where exactly do you see a place to fudge anything anyway?

Regarding “who measures”;
For what its worth too, the only people your likely to find that really know how to do this stuff (like be proficient on a TEF machine), know how exactly because its been there job for some time and are usually associated with a company.
For example Ivan has probably set up and aligned more than a thousand auditoriums and stages with a TEF.
If a person from a participating speaker company isn’t acceptable to you or even a friend of a friend of one, you could obviously hire someone “un-involved” from the LAB.

You are the only person I can think of that has suggested that I fudge measurements too and insinuate that again here.
If you understood the big picture, with you again inferring that, you would also understand why I probably won’t recommend we send any speakers or equipment to your event.

I will leave you with some free advice, look at it as someone trying to trick you if you wish.  
Use a TEF machine, for measuring a loudspeaker, like you are doing, like I do, it is I think the best tool available.  
Heck, most (I won’t say all because I have not tried all of them) of the popular measurements systems don’t actually measure acoustic phase (while they do have a plot for it).
The erroneous phase graph has lead an entire movement of “fixing phase with all pass filters” based on bad (fictional) curves.
The TEF on the other hand passes the reality test I devised for phase.
Cheers,

Tom
Logged

Wayne Parham

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 302
Re: Prosound Shootout 2007-Test Procedures-Impedance questions
« Reply #33 on: August 02, 2007, 02:43:56 PM »


Good dialog, Ivan, thanks.  Let's continue to look at this.

Ivan Beaver wrote on Thu, 02 August 2007 12:15

The way you want to do it (using the minimum impedance) is out of industry standards, the ONLY thing you could even hope to compare would be between boxes measure at your shootout.  Other boxes measured in different ways have no corrilation to your results-except maybe actual freq response, but with no absolute value assigned regarding sensitivity, becuase of the drive level you have choosen.

...

Yes I totally agree that the rated impedances are just arbritrary numbers, but at least there is some coorilation to the actual impedance graph, but you have to have some standards.  Just ask how people feel about a 6 ohm loudspeaker and they will usually ask what is wrong with it so that it doesn't have a "standard" value.  It is usually so ocnfusing for them to think outside of their little boxes of 4-8 16 ohms.



You have suggested that there is a "standard" for published advertised impedance.  I do not believe there is one.  Advertised impedance appears to be an arbitrary number to me.

Minimum impedance, on the other hand is not arbitrary.  Minimum impedance is easy to measure, definite and useful for making these kinds of comparisons.

Let's look at some actual impedance charts, shall we?  Then we can run the numbers and see what we come up with.

http://www.prosoundshootout.com/Measurements/2006/Tuba36_impedance.gif

This is the impedance chart of a Fitzmaurice Tuba 36.  Looks to me like Zmin is 9Ω.  From the forumula E = √PZ, we find the drive voltage required to dissipate 100 watts in a 9Ω load is 30 volts.

What would you say average impedance is?  Should it be 8Ω, because that's the closest multiple of 8?  Should it be 12Ω, since that's the next higher multiple of 4?  Or should it be closer to 10Ω, since averaging the area under the curve up to 150Hz puts it closer to that?

Let's run the numbers and see what each of the drive voltages would work out to:

28.3v into 8Ω dissipates 100 watts
30.0v into 9Ω dissipates 100 watts
31.6v into 10Ω dissipates 100 watts
34.6v into 12Ω dissipates 100 watts

Now let's calculate the difference between each of these in decibels:

The 28.3v (8Ω) level is 0.94x the 30.0v level, which is -0.25dB.
The 31.6v (10Ω) level is 1.05x the 30.0v level, which is +0.23dB.
The 34.6v (12Ω) level is 1.15x the 30.0v level, which is +0.62dB.

We're only talking about a half decibel max difference between Zmin and any sort of realistic average impedance.  You actually could eye-ball an average value and be pretty close, like probably less than +/-0.25dB.  But the minimum impedance method uses a tangible number that nobody has to argue about.  It's what you see on the chart.

Honestly, Ivan, I think we could go with fixed voltage measurements and calculate impedance offsets, just like you've described.  I would be comfortable with that, and might do it myself if I were viewing my own data.

But what we're talking about here is confidence.  This is an event where a lot of exhibitors will be bringing their best products to be compared with others.  I think they would be more comfortable seeing what steps were taken to determine their SPL charts, right there in front of them.  They see the impedance chart.  They see the power formula run.  They see the voltage set on the amplifier.  And they then see the SPL chart made during the sweep, and that's what they see published later on the web.  No surprises.
Logged
Wayne Parham
π Speakers
PiSpeakers.com

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 9010
Re: Prosound Shootout 2007-Test Procedures-Other opinions?
« Reply #34 on: August 02, 2007, 03:53:47 PM »

My questions are at the end of this.

What I have seen by looking at MANY manufacturers specs (and trust me I look at  a lot) is when stating the sensitivity many say 2.83V 1M.  Yes there are those that still say 1W but more and more they are saying 2.83V-impedance be damed Shocked .

There are some people that do say 1W/2.0V 1M for a 4 ohmn cabinet-so they are sticking with that spec type and that is fine-they are telling you exactly what is going on.

The assumption by the general public is that this is 1W, so it works in favor of the lower impedance cabinets, but there is a saying "buyer beware".  They are not suggesting anything or saying anything false, but if the buyer misinterpets it or does not read the rest of the data sheet where it ways 4 ohms, then that is an issue that is up to them.

Most people are not concerned with how loud a cabinet is with a particular wattage input.  There are two concerns. 1: How loud is this cabinet when I hook it up to my so and so amp.  2: How many of these cabinets can I hook up to my so and so amp without damage?

When you really dig into the wattage issue it raises questions.  

Let's look at your example of Bills cabinet.  Where is the highest impedance point?  Right where a lot of kicks are tuned to-between 60-70 Hz.  PLEASE LABSTERS let's not turn this into a where is your kick drum tuned to thread Laughing .  This is just an example.  So at that freq (which is where many would require the greatest amount of output (talking rock type stuff here) the amp is seeing a very high impedance and actually putting out very little power.  Therefore if the kick is your primary source of volume in the sub, then you might could actually hook up more than would be normally suggested as safe-BUT you did not read me saying that Laughing .

I am siding with the rest of the industry @ 2.83V.

It is very much like back in the late 80's and the pin 2 pin 3 issue and what was standard.

It is hard for a manufacturer to change to the "new" standard when they have been doing so fo so long in a particular way.  It would confuse the general public.  Look at JBL and red negative for example.  They are different than the rest of the world (as far As I know), but still do it, and with good reason.  Most people understand this-but many still learn every day-It was quite eye opening when one of my competitors told me this oh so many years ago-I didn't belive him untill I took a battery to one myself.

I would love to here how others feel about what the input should be-a constant voltage or a voltage derived from the lowest impedance on the curve and not the manufacturers rating.

What would they like to see-not only at a shootout, but on spec sheets in general?
Logged
For every complicated question-there is a simple- easy to understand WRONG answer.

Can I have some more talent in the monitors--PLEASE?

Ivan Beaver
dB Audio & Video Inc.
Danley Sound Labs

Tim McCulloch

  • Newbie
  • *
  • Offline Offline
  • Posts: 0
Re: Prosound Shootout 2007-Test Procedures
« Reply #35 on: August 02, 2007, 04:06:22 PM »

Wayne Parham wrote on Wed, 01 August 2007 19:35

You and Mark Seaton are both personal friends and business associates with Tom Danley.  The fact that the data was gathered, organized, processed and published by you and Tom made some people pretty uncomfortable.  They told me so.

Again, please don't take that wrong.  I saw what you guys did and I could tell you were trying to be fair.  But please understand that it's the same thing as if Danley had to trust measurements I made, processed and published.  He has been pretty vocal that he would be uncomfortable with that.  I understand, and I don't expect him to be forced to accept a situation that makes him uncomfortable either.

We'll need to work in a spirit of understanding and cooperation to find and agree on methods that everyone is comfortable with.


The data from the NYC Sub shootout wasn't altered in any material way I could determine.  I was there, I watched and listened to the measurement process.  While I'm not a TEF expert, I've measured a few speakers for acoustic performance, and have comprehension of what takes place.  Every speaker received identical treatment (as best we were able).  The results were obviously influenced by the small and severely reflective space, but what was released publically jived with my memory of what I saw on-screen in NYC.

The "Monday morning quarterbacking" after the results were released left a bad taste, and made the agendas of several individuals readily apparent.  I suspect that some of those folks would have been unhappy regardless of who's fingers were pushing the buttons.  Oh well, water under the bridge...

Good luck in October, if I'm not working I'll try to be there.

Tim Mc
Logged
"Will you stand by me against the cold night, or are you afraid of the ice?" Crack The Sky

John Roberts {JR}

  • Newbie
  • *
  • Offline Offline
  • Posts: 0
Re: Prosound Shootout 2007-Test Procedures-Other opinions?
« Reply #36 on: August 02, 2007, 04:09:30 PM »

Just to expand upon what Ivan said, in use we hook up amplifiers that put out volts not watts, so it's useful to see how the speakers respond to reference voltages. Simultaneously it is useful to know minimum impedances (in band) so we can know how many speakers in parallel a real amplifier can tolerate. A speaker designer can cheat somewhat by making a 8 ohm speaker 5 ohms, but that could come back to bite the poor customer who tries to parallel 2 of them, expecting a 4 ohm load.

Since these impedances are not flat lines or simple to evaluate I would be interested in minimums within the expected passband. A dip at 20kHz in a woof is not much concern. In full range cabinets another story.  

So SPL for a given voltage with an asterisk for impedance which is only important when unusual. I prefer to see plots but I am not the typical customer for these shoot outs. A minimum will tell the typical customer what (s)he needs to know.

Logged
 https://www.resotune.com/


Tune it, or don't play it...
-----

Wayne Parham

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 302
Re: Prosound Shootout 2007-Test Procedures-Other opinions?
« Reply #37 on: August 02, 2007, 09:26:28 PM »


Ivan Beaver wrote on Thu, 02 August 2007 14:53

What I have seen by looking at MANY manufacturers specs (and trust me I look at  a lot) is when stating the sensitivity many say 2.83V 1M.


We're measuring 2.83v/M, always have and always will.  Actually, we measure using 28.3v at 10 meters, which is the same SPL, as you know.  So you can use that figure and disregard the others, if you wish.
Logged
Wayne Parham
π Speakers
PiSpeakers.com

Wayne Parham

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 302
Re: Prosound Shootout 2007-Test Procedures
« Reply #38 on: August 02, 2007, 09:54:52 PM »


Tim McCulloch wrote on Thu, 02 August 2007 15:06

The data from the NYC Sub shootout wasn't altered in any material way I could determine.  I was there, I watched and listened to the measurement process.  While I'm not a TEF expert, I've measured a few speakers for acoustic performance, and have comprehension of what takes place.  Every speaker received identical treatment (as best we were able).  The results were obviously influenced by the small and severely reflective space, but what was released publically jived with my memory of what I saw on-screen in NYC.


I don't think anyone said the NYC data was altered.  Some were just uncomfortable with how it was handled.  So what we've learned from that is to be careful with what we do and how we do it.
Logged
Wayne Parham
π Speakers
PiSpeakers.com

John Roberts {JR}

  • Newbie
  • *
  • Offline Offline
  • Posts: 0
Re: Prosound Shootout 2007-Test Procedures
« Reply #39 on: August 03, 2007, 12:24:28 AM »

Wayne Parham wrote on Thu, 02 August 2007 20:54




I don't think anyone said the NYC data was altered.  Some were just uncomfortable with how it was handled.  So what we've learned from that is to be careful with what we do and how we do it.


Any names or specifics?

I wasn't uncomfortable, but I don't have an axe to grind.

JR

Logged
 https://www.resotune.com/


Tune it, or don't play it...
-----
Pages: 1 2 3 [4] 5 6 ... 16   Go Up
 

Site Hosted By Ashdown Technologies, Inc.

Page created in 0.029 seconds with 20 queries.