[pmel]

Thats about correct.  The good thing though is that they are actually very well behaved as frequency goes down, so useful data interpretation is possible, but requires certain specified set-ups.

As in some kinda boundy effect calcution?  Make several measurements from different parts of the room , and use alot of math to calcuate known reflections outa the equation?  Withen reason all the reflections, and phase nulls are going to happen at the same place everytime on a linear curve with CW correct?

How far could you shink the chamber size at 100 Hz with this kinda methodolgy without running into problems for simple audio measurements you figure?

[Danny Richie]

I spent some time with John Dunlay at his facility. He had two chambers. Both were 25 feet long and foam wedges were 12" or more thick. He said they were accurate to about 40Hz (how accurate they really were, I don't know) and below that was little variance. He said that he could accurately calculate the response in the last octave.

Most reasonable sized chambers aren't going to give you 100% absorption in the bottom end (at all), but most can give you a good percentage.

As speaker designers we rarely work with speakers that reach -3db levels in the 20's. Plus we know that they will be used in a room and there will be room gain. All rooms are different and the amount of low end gain will be different. With the vast amount of variance possible with in room responses for any of us to know an absolute accuracy in those low ranges really isn't of great value. 

I can give you a speaker that is flat into the 20's and then you put it in your room and have 8 or 10db peaks in the bottom end.

Honestly the more end user adjustability you have on the first octave and a half to two octaves the better off you are.

Mine was 21 feet long and the thicker stuff was the Venus bass traps from Auralex: http://www.auralex.com/category_bass_traps/category_bass_traps.asp

I could get good un-gated responses with no reflection effects from about 100Hz and up. The lower I went the more low end gain I could see. It is pretty predictable though. It's not like there was a big reflection in the 80 or 90Hz range. I was still absorbing the vast majority of it. The real advantage was greater measurement distances.

My old chamber...

[Danny Richie]

Paul, you can post here anytime.

I am still not seeing too many comments on the measured distortion figures I posted. Just two guys? No one wants to go out on a limb and tell us which woofer will sound better?

[konut]

My 1st impulse would be to say woofer 2 would be a  yield a better sound if the xo was much above 200hz. . From 200hz on down it would be hard to say without some other info, so it would depend on where the xo was. 

[PaulHilgeman]

And likewise Danny,

The thing with a chamber is that you want to be able to get decent distances (6 - 9 feet), and that is what is the hardest part about measuring in any room, but less important is the LF data.  Unless you are designing a crossover in the 200-400Hz range, but even then an anechoic chamber may not be the best thing at all.  Crossovers like that are so dependent on floor reflections, and baffle step that so much of it has to be done by ear anyways, measurements are just sortof along for the ride, and even if you could have perfect anechoic data, it really wouldn't make much of a difference to me.

As to the question about figuring out the LF curve of the chamber... not all that hard, the total output of the speaker in question will be nearly the same as a nearfield measurement, just lower in level.  Basically, take a nearfield, then a farfield, and compare the two.  THe difference is how the room affects the FR curve with those two specific locations (speaker and mic).

-Paul

[PaulHilgeman]

Paul, you can post here anytime.

I am still not seeing too many comments on the measured distortion figures I posted. Just two guys? No one wants to go out on a limb and tell us which woofer will sound better?

What page are these on?  I glanced and couldn't find them... or are they on a web page?

[eric the red]

So after 23 pages of discussion, some of us are curious as to how does that infamous driver sound when properly implemented in a speaker?

[Grumpy_Git]

Danny

Take a M130, put a filter on the Hump, create a plot in a similar manner to Zaphs setup and publish, then let more arguments happen.
we will then all be able to observe amplitude peaks/ringing/resonance and any other nasties/compromises, and decide much notice the raw measurement should be given.

Do you make any big polymer baskets? I need to jack up my car

which driver would I choose? Neither, they aren't pretty enough, or expensive enough! 

cheers all

Nick.

Qualifiers: I claim to know nothing.
              This thread is taking way too much time from my life.
              My Speaker building skills are POOOOOORRRR

[JAD2]

[/quote]

Big problem here.  When you play your guitar, you're trying to create sound, not recreate.  Try this one on for size:

1) Plug your fender into your marshall/peavey stack
2) mic your stack and record it
3) rebroadcast the recording from your marshall/peavey stack and compare that to the sound being rebroadcast from one of krutke's or Danny's speakers or whatever.

The point is that you're not supposed to play your fender through gear designed for reproduction vs. production.  They have completely different design goals.  This says nothing about your preference for metal or paper drivers.
[/quote]

Ah, the normal type forum reply and EXCUSE.
So I guess there is a wild difference in my small marshal amp versus the amp I use on my stereo. So and Amps job has now changed from taking a sound and just amplifying a sound, it adds something other than power to drive the speakers.
Now speaker drivers goals have also change, instead of playing what the amp sent it, it also adds something that doesnt belong.


Well when I played through the wrong type amp and the wrong type speaker, which was the paper driver/soft dome set I built, my guitar sounded dam close through those as it did my marshal peavey setup. But through the metal drivers it had a total completely different sound.
Then if I take my classical guitar, no amp or speakers,play it, then play a demo tape I had made through both the metal driving ones and the other and the metal drivers add something ringy to the sound the others dont, whats the excuse there????

Preference for something??? Does that mean, or implying some form of subjectivity. Now according to Zaph's Audio, subjectivity is and excuse showing lack of knowledge. I just thought some preferred realism to their sound, while others like adding something that doesnt belong for whatever reason.

[tberd]

 I'm still waiting for the "experts" to chime in also . I thought for sure that John would set us all straight . And his posse got pretty darn quiet too .

 And way to shut'em up Danny with that car jacking trick . Next page the "experts" will scream Photoshop for sure .

 All kidding aside , it's funny how the topics in this 23+ page comedy show wander .

 And what's up with that orange peel measuring kit John was using on the drivers ? Is that an audio standard somehow ?

 Ok sarcasm over . This is better than the OJ trial .


tberd

[Danny Richie]

What page are these on?  I glanced and couldn't find them... or are they on a web page?

First post page 20.

Then if I take my classical guitar, no amp or speakers,play it, then play a demo tape I had made through both the metal driving ones and the other and the metal drivers add something ringy to the sound the others dont, whats the excuse there????

Your answer can be seen if you check what I feel are Zaph's most useful and telling measurements. See the spectral decays of the metal cone woofers. Despite the fact that most crossovers will push the output level of the resonance down about 20 to 25db, it is still very audible and the long decay time of the resonance can easily be heard as a "ringing" sound.

[WOR Radio]

After 35 years of loudspeaker building:

I say..

"measurements are the start, listening is the goal"

[JoshK]

Despite the fact that most crossovers will push the output level of the resonance down about 20 to 25db, it is still very audible and the long decay time of the resonance can easily be heard as a "ringing" sound.

Jon's xo for the Modula MTM pushes the level down 50 or 60 dbs, can't remember which.  He has lot of data to show this too, look for his CE filter threads.  By your own argument, if -50db is unaudible, then so is the resonance in this out of band topology.

[Ronno6]

Danny, in regards to the graphs of the two woofers, you asked for impressions, even from those with no idea. That would be me. I am going to cast the dissenting vote and go with woofer #1, as the plots indicate a more level average than that of the plots of woofer #2, which shows a steady decline. I don't know how much of this is audible. Maybe the steady decline is better, but as I said, I am he who hasn't a clue.That's my two cants.

Ron

[dlr]

Your answer can be seen if you check what I feel are Zaph's most useful and telling measurements. See the spectral decays of the metal cone woofers. Despite the fact that most crossovers will push the output level of the resonance down about 20 to 25db, it is still very audible and the long decay time of the resonance can easily be heard as a "ringing" sound.

All part of any crossover design. You have to account for everything. It would be interesting to see a debate between you and Siegfried Linkwitz on hard cones, distortion and audibility. Is it your take that his designs that use those nasty hard cones are by default going to sound worse than yours due to the use of hard cones? Or does he just not know what he's talking about?

I'm also interested in hearing a response for your take on motor generated distortion and the impact of the raw driver response. You were adamant on the Mad posting that when driven hard, other drivers such as Scan-Speak might not hold up in some ways as might be expected.

From the Madisound board: "Push those woofers to levels that take the exertion to ranges near their X-max and them compare the M-165X to the Usher or Scan Speaks. You may find the Usher and Scan Speak woofers to not be measuring up."

When the M130 is driven hard (near Xmax per the above quote), what is the impact of the distortion? This is primarily a focus on the motor's non-linear distortion. What is the distortion of your drivers when driven to Xmax? We all know it is present in all drivers. You've not responded to any of these points discussed previously.

What's curious about all of this is that I thought that you went to the added expense of incorporating Faraday rings on the drivers. Is that correct? What justifies this expense? I'm sure the readers here would like to know specifics of why you use them if you do. For some reason you seem to be belittling distortion measurements. Since you are the manufacturer (even if by proxy), can you provide a full set of what you would deem to be accurate measurements? As a manufacturer, certainly you must be aware of what the distortion characteristics are for your own product.

If you don't include Faraday rings, why not? The added expense? That would be an acceptable response.

Either distortion is audible or it is not. What is your position on non-linear distortion and the impact of the raw driver's frequency response? Do you believe that there is any correlation or not and on what basis do make your claim?

I get the impression that your position is focused only on the linear distortion of the FR. Is that all that you care about?

As to the posting on distortion measurements and sound, here's what I'm sure will be the eventual follow-up, something like another quote from one or your posts at the Madisound board:

The Scan Speak and Usher drivers are great drivers, but you can't just look at one aspect of something and get a full picture or draw conclusions.

Given your own words, it's obvious that the post with a single (exceedingly incomplete) set of distortion measurements is nothing but a straw-man setup. Danny, I really expect better from you.

For those curious, the link is here:
http://www.madisound.com/cgi-bin/audio_forum/index.pl?read=392491

Just so others here don't make baseless charges, my preference is for high-quality doped paper such as the SS 12m/4631 and previously the 13m/8640. For woofers I like the SS 25W/8565-01. However, I have also heard excellent sounding hard cone drivers, those with a properly designed crossover, that is.

[PaulHilgeman]

Despite the fact that most crossovers will push the output level of the resonance down about 20 to 25db, it is still very audible and the long decay time of the resonance can easily be heard as a "ringing" sound.

Jon's xo for the Modula MTM pushes the level down 50 or 60 dbs, can't remember which.  He has lot of data to show this too, look for his CE filter threads.  By your own argument, if -50db is unaudible, then so is the resonance in this out of band topology.

I hit -55dB in the Ronin, and yes, I feel that pushing the resonance down as far as possible along with crossing over below the 1/3 point of the resonance is essential to making metal cone midwoofers sound good, and when this is done properly, they can sound great.  The peak in 3rd Order distortion at 1/3rd the resonance point needs to be outside of the passband of the driver as well.  So, with the W22 used in the Ronin, Crossing at 1350Hz, keeps the 1/3rd resonance point at 1600Hz down by about 10dB, enough to reduce the 3rd order distortion peak by almost 20dB relative to the output of the system at that point.  It also puts the real resonance down by 55dB.

In order to achieve a neutral sound with a metal driver, practices like this are absolutely necessary in my book. 

In is unfortunate that metal drivers have gotten a bad rap because people don't follow these practices.

And yes, the measurements count... but it is the listening that should 'sign off' on any design. 

-Paul

[Danny Richie]

"measurements are the start, listening is the goal"

And yes, the measurements count... but it is the listening that should 'sign off' on any design. 

Good, good, very true.

Jon's xo for the Modula MTM pushes the level down 50 or 60 dbs, can't remember which.  He has lot of data to show this too, look for his CE filter threads.  By your own argument, if -50db is unaudible, then so is the resonance in this out of band topology.

That's good, that's real good. Now you are just left to contend with the unnatural sound of a metal cone driver, but that's a subjective opinion and a different argument.

It would be interesting to see a debate between you and Siegfried Linkwitz on hard cones, distortion and audibility. Is it your take that his designs that use those nasty hard cones are by default going to sound worse than yours due to the use of hard cones? Or does he just not know what he's talking about?

Hey that would be great! So long as it is a face to face discussion and involved listening as well. I'd love to get to spend that kind of time with one of the brilliant minds and pioneers of our industry.

When the M130 is driven hard (near Xmax per the above quote), what is the impact of the distortion? This is primarily a focus on the motor's non-linear distortion. What is the distortion of your drivers when driven to Xmax? We all know it is present in all drivers. You've not responded to any of these points discussed previously.

The M-130 does not use an XBL^ motor and responds to being driven hard much like any other woofer. I do however have some on order with the XBL^ motor. I have received 4 samples. Two of them were used in a prototype speaker for RAW Acoustics. While the two woofers (M-130 and M-130X) sounded very much the same, when driven hard it was clear to hear that the one with the XBL^ motor seemed to handle the higher output levels with more ease, less strain, and remained very clean. Clearly the more linear Bl does make some audible differences not hard to notice.

The M-165X is the woofer referred to in the post on the Mad forum and it does use the XBL^ motor.

What's curious about all of this is that I thought that you went to the added expense of incorporating Faraday rings on the drivers. Is that correct? What justifies this expense? I'm sure the readers here would like to know specifics of why you use them if you do. For some reason you seem to be belittling distortion measurements. Since you are the manufacturer (even if by proxy), can you provide a full set of what you would deem to be accurate measurements? As a manufacturer, certainly you must be aware of what the distortion characteristics are for your own product.

If you don't include Faraday rings, why not? The added expense? That would be an acceptable response.

Either distortion is audible or it is not. What is your position on non-linear distortion and the impact of the raw driver's frequency response? Do you believe that there is any correlation or not and on what basis do make your claim?

Clearly I know that low distortion is important, isn't it? Why else would I be using XBL^ motors with Copper rings with these new models?

My issue is and has been the value of, inconsistency, and credibility of distortion measurements made outside of the controlled environment of an anechoic chamber. Both near field and far field measurements have their own set of problems as measurements posted here by myself and Dan Wiggins clearly illustrate.

Given your own words, it's obvious that the post with a single (exceedingly incomplete) set of distortion measurements is nothing but a straw-man setup. Danny, I really expect better from you.

No, it very accurately illustrates a point, and that was the point (you'll see). People view a single measurement, like posted distortion measurements, and draw opinions without looking at any other forms of data to get a complete picture. They do it all the time.

Just so others here don't make baseless charges, my preference is for high-quality doped paper such as the SS 12m/4631 and previously the 13m/8640. For woofers I like the SS 25W/8565-01. However, I have also heard excellent sounding hard cone drivers, those with a properly designed crossover, that is.

I tend to agree with your impressions and observations.

[PaulHilgeman]

Looks to me like Woofer 2 would probably be a better choice.

Lower 2nd order across the board

Lower in the midrange, except the peak at 1.6Khz, however the peak is not as bad as it looks, it is above the level of distortion in the driver, but not all that much above the level of Woofer 1 in the frequency range in question.  Woofer 1's motor looks a bit better as it has better distortion down low, but not by too much.

4th order is a touch worse at 100Hz, but significantly better everywhere else.

5th is hard to tell, but woofer 2 has more points that dip below 50dB.

I would choose Woofer 2 and do further evaluation on what is going on at 1600Hz. 

That said, I would really want to take a look at the frequency response on both an infinite baffle and the baffle in question, from both near and far field.  Plus I would do my own distortion measurements 

All that said, these are very bizzare looking distortion measurements.... why do they have so many peaks and dips?  I do my distortion sweeps unsmoothed and they never look like this.  Even when I run spot-checks indoors....

-Paul

[JAD2]

Then if I take my classical guitar, no amp or speakers,play it, then play a demo tape I had made through both the metal driving ones and the other and the metal drivers add something ringy to the sound the others dont, whats the excuse there????

Your answer can be seen if you check what I feel are Zaph's most useful and telling measurements. See the spectral decays of the metal cone woofers. Despite the fact that most crossovers will push the output level of the resonance down about 20 to 25db, it is still very audible and the long decay time of the resonance can easily be heard as a "ringing" sound.
[/quote]

OK,I have finally learned something again.

Then DLR gets going again playing out the part of the crossover being wrong.

Now I act as a referee and blow the play dead!!!!!!!!


YOU WHO, remember me, Mr. Average Joe!!!!

I am the majority purchaser in the DIY arena. There are many of us, many, many many more so than what I will now add as the advanced DIYer's. We are the ones whom Danny stands to lose with falsified information at Zaph's site if he with poor measurment ways, reviews negatively a driver.

Got it!!!!
We have at best a table saw, router, soldering iron and the nerve to proceed further in DIYing our own.

WE CANT MEASURE, most dont have programs to do anything other than a simple 2nd order type crossover.

If what I read is correct the way I took it. Then metal drivers shouldnt be considered, because with the simple tools I have, I cant correct the natures of the beasts.

And as other have stated, myself included, the final sound of the product is the most important part. Yeah the other stuff plays a part, but based upon the original premise of this thread of someone downplaying another product, this measurment game your playing, CONFUSES the sh*t outta us average Joes.

Boils down to this.

How does the dam thing sound!
How realistic does it replicate the original sound.
If you want to get back into this measurement means all crap, then:
Which ones are truly audible, I mean make enough difference, me Average Joe could detect.
Which ones dont.
Are all measurements picked up by all this stuff I dont own, truly that important, or just something the program and mic can detect. Again can we hear it or not.
I find alot of this measurment stuff just being politically correct so to say, then having much bearing on the final product at my ear.

See Mr. Zaph's Audio's review of the GR research speaker kept me from purchasing one because for cheaper I got a much better reviewed one, actually one of his top recommendations, which was a metal ringing nag to my ears. Thats what this is all abput, isnt it????

Dont forget, Mr. Average Joe is watching, we are the ones being swayed by all this!!

[Danny Richie]

That said, I would really want to take a look at the frequency response on both an infinite baffle and the baffle in question, from both near and far field.  Plus I would do my own distortion measurements

You are correct. More data is needed. Thanks for your input.

All that said, these are very bizzare looking distortion measurements.... why do they have so many peaks and dips?  I do my distortion sweeps unsmoothed and they never look like this.  Even when I run spot-checks indoors....

They are a meter away from the source so room noise and its effects are more prominent in the measurements.