Printable View
This is to continue a conversation about compression driver design and the effects material, surround type and geometry have on the HF time domain of the secondary resonance of the drivers.
Typically 10-20Khz.
Take a look at the attached video talking about material differences, surround differences and the effects. It's informative and we can discuss in the thread.
Rob :)
https://www.youtube.com/watch?v=x9NAVlwvLro
Hey Rob,
This has video does a good job of expanding all the various aspects of diaphragm materials and behaviours.
The 18 Sound presenter at the end has offered the pdfs of the video. Why not see if you can post them for awareness and educational purposes.
God idea see if I can find them.Quote:
Originally Posted by Ian Mackenzie
Rob :)
I have contact details of a Tech Lead at B & C so leave it with me.
With the exception of bizarre behavior like licking Beryllium diaphragms, that's a great video. Frequency response alone doesn't tell all you need to know.
Loved the anecdote of the pro audio guys who wanted to return to using the drivers with the Godawful ringing because they preferred the 'sizzle'. :-D
Damn... licking Be, now I must find the time to watch the video! :applaud:Quote:
Originally Posted by rusty jefferson
Widget
Better make yourself comfortable, the damn thing goes on for over an hour. I gave up trying to find the Be licking and the "sizzle" comment :snore:
1:18Quote:
Originally Posted by DerekTheGreat
What surprises me is how clean the JBL Ti SL and Magnesium drivers are compared to many of the driver CSD's in the presentation.
Rob :)
Thanks Rob! That was funny, the one guy goes, "Don't--" to which the Be licker guy goes, "Nah nah, that's if you grind it up into dust and snort it. Which this is rock & roll, so somebody might.. Don't do that." hahaQuote:
Originally Posted by Robh3606
...Matteo said they're going to release a 15" Coax driver??? Any details on this?
Just a reference here a CSD plot of an equalized 476 Mg. It's very clean and doesn't show as much stored energy say compared to the 2453 SL CSD. That said the 2453 is also very clean.
Rob :)
That's super clean and way nicer than the 3" MG diaphragm they showed in that video. Is that the mid driver in the 1400?Quote:
Originally Posted by Robh3606
HelloQuote:
Originally Posted by rusty jefferson
That's on an M2 waveguide, it's a 4" Compression driver used in 9900, 4365 and a lower cost version of Everest.
Rob :)
It's a 4" compression driver that JBL used to use in the 9900, 4365, and lower cost DD65000 Everests. Now JBL is using a Radian built 4" compression driver with presumably the same diaphragm. That is assuming anyone is still buying the K2-S9900 or the DD65000s are even still being offered.Quote:
Originally Posted by Robh3606
Unfortunately the rather excellent JBL 476Be and 476Mg drivers are NLA and just a couple of additional legacy pieces of excellent engineering and manufacturing.
Widget
Licking Be, I'll bet it tastes like chicken.Quote:
Originally Posted by Mr. Widget
You know the big step up in JBL compression drivers used to be the 375 not only physically but in price.Quote:
Originally Posted by jpw retired
On a lighter note l am wondering if we should come up with a Tech sheet that says use only the Radian aluminium- magnesium- copper alloy diaphragms in these monsters. ( a proprietary hardened alloy).
If you’re under 50 chop them with a steep slope @ 10,000 hertz and add the tweeter of your choice. (Radian have some excellent planar ribbon which Vance Dickason comments are worthy of studio monitors)
The vertical polar is lousy but let’s not go there.
https://audioxpress.com/article/test...io-engineering
From time domain data from European tests on a DD66000
Hey Rob,
Given the interest and intrigue on the Urei monitors it would be interesting to examine the time domain measurements of your Urei centre channel and compare to the M2 as a baseline. It would also be interesting to compare the impulse of the woofer and the horn to see if in fact they are time aligned at the crossover point?
I would love to see the impulse response of a UREI coax... I'll bet it is not truly time aligned. I would expect that it is improved over the original Altec/JBL, but I doubt they were able to nail the impulse response with simple passive components.
Widget
Hey WidgetQuote:
Originally Posted by Mr. Widget
Hope you are high and dry! I have the driver and crossover but no cabinet and don't have one that would accommodate and E-145 with a 2425S hanging off the back. So lets hope someone else can post it. Curious as well !!
Rob :)
I've got a spare 811C you guys can test, you just have to pay & arrange for shipping or come to Michigan. ;)
Quote:
Originally Posted by DerekTheGreat
If you have a Laptop you can do with some assistance.
:uhmmmm:
Well, it's not likely to be good by today's standards but I'd be curious to see how different the UREI is from the Altec 604-8g, which I think is the model UREI used when they started. I know eso has some 811b(?) but don't know if he has measurement tools, and may have 604-8g? Toddalin has Big Reds (604-8g?) but they likely have the Soundlabs(?) crossovers which might have time alignment(?) not Altec crossovers.
And I'd really like to see a waterfall of the 2425h with an original aluminum diaphragm and the later titanium replacement. Have to believe the ringing in the titanium diaphragms is wicked.
Maybe. But it’s probably appropriate to keep opinions out of this thread. It’s about the facts.
Hello RustyQuote:
Originally Posted by rusty jefferson
I have been going back to look at older measurements. I have darn near all of them saved since I first started using Clio. So I looked around and found a couple of 2425/26 measurements on the 2344 and 2370 Biradial. They look reasonably clean last octave compared to what some of the drivers in the presentation look like.
These are Ti never had a pair of the original aluminum ones.
Take a look
Rob :)
:-D Hey, thanks Rob. You are the keeper of the technical files. I'm surprised both are so clean at high frequency as you mentioned, but that first one (2344?) looks pretty rough in the 1khz-2khz range. I should have added a quote to that last post as my curiosity about the 2425 is time alignment and waterfall mounted as the 801. Now I wonder if it's not high frequency ringing but rather closer to the crossover point?Quote:
Originally Posted by Robh3606
I have a pair of 2425 here and Dhar has a 3d printed Joseph Crowe 1" exit horn we should try getting some measurements on. Might be surprised. If anybody in the DMV has an 811 to measure I'm sure he'd be happy to do it.
Hi Rob,
Under what conditions did you test those drivers ?
Were those measurements gated and what distance?
Looks like you smoothed them 1/12 octave.
I have been trying to obtain some detailed information from the presentation. Attached are some files comparing the titanium and beryllium diaphragms. (18Sound)
* Most of the B&C HF compression drivers are for OEM customers using line arrays.
* B&C HF drivers are designed for maximum efficiency and output in pro applications
* Specific B&C HF drivers are designed for studio monitors
* ln some instances in the B&C drivers the hf resonance is designed into the driver at the request of the OEM client @ 15 khertz ( see the presentation).
* While not scientifically validated it’s the longer time domain resonances that can be detected by the ear/ brain above 10,000 hertz. For example a 15000 hertz frequency has a time domain duration of 0.15 of a ms per wavelength. If this takes 4 ms to decay then there are 26 cycles during that time period. That’s what’s detected as a spatial information. Four ms is the equivalent of a 576 hertz wave length.
18Sound are a premium brand now owned by B&C loudspeakers
* aluminium, titanium and beryllium diaphragms are available across the range
A few general trends
* An HF compression driver that has an extended non normalised above 10,000 hertz on a snap shot FR curve will have resonances with longer time domain.
* An HF driver (including beryllium) that doesn’t have resonances will require greater EQ to obtain a normalised response. (Annular diaphragm drivers are a different case)
*Beryllium diaphragms are very costly.
Attached below is a screenshot of the 18Sound HF driver range.
Hello IanQuote:
Originally Posted by Ian Mackenzie
The gate times are right on the graphs. Smoothed as little as possible in Clio 7 and also what makes sense. Distances?? Not sure those are at least 12-15 years old.
Rob :)
If you want that as a reference for comparison then it needs to be validated. Otherwise it’s misleading.
On the basis of you’ve posted if it was true then why bother with beryllium??
I think there needs standard for the rise time and the other settings
Quote:
Originally Posted by Ian Mackenzie
The rise time is on every measurement, the smoothing and gate time. I don't change the rise time so same conditions. So what's the beef?
That is the default in Clio 7.11 and Clio Pocket 3.0 . So even between the 2 platforms same conditions. Both of which are calibrated for accuracy.
I don't see a reference on the plot for rise time on the 66000? So it's not valid.
I posted those in response to some of the drivers in the presentation. It's obvious to me that JBL did a good job with the diamond surround.
Let's see what you get. You still have your 2344/2425-26 combo? Set your conditions and we can compare. A 2370 or a 2307?
Did this with Zilch years ago and they were damn near identical.
Rob :)
https://www.audioheritage.org/vbulle...highlight=clio
Hello Ian
435Al/2431 measurement standard .580 and .15 rise time un-smoothed same gating .
Still "clean" either way. Little more detail faster rise time. Sticking with default standard.
Rob :)
"Your waterfalls have a different windows.
That completely changes the perspective of a waterfall plot. They all have to be like for like or its a waste of time comparing them because it's a visual representation. In REW I can make any driver look good, bad or indifferent by changing the perspective.
The real point is that the diaphragm output decays smoothly for a good 2ms then here are resonances ringing for 2.5ms or so out 5.1 ms. The thing is this diaphragm isn't well damped and that is why it sound's live."
Hello Ian
Changing the window length does not change the result.
You have choices for the number of slices I use 60 typically and if you increase the window the time between each slice increases.
This slightly changes the CSD appearance but but does not change the results
Do you or do you not have substantial ringing? That doesn't change.
Also looking down 40db you are in the mud. Go back and look at the CSD range used in the Beryllium presentation you posted it's 25db.
Also if you go back to the video 1st post you will see they are comparing CSD plots with different time windows and Db scales yet they seem to have no issue with it?
So I guess their presentation is wrong?
Change your view to 25db it will be closer to what I typically post. Also try a .58 msec rise time vs the 2.9 msec you have now.
Using the same conditions they should be much closer than your current settings.
Posted are window length changed CSD plots first and second posted at 25db.
No significant change.
Changes from 25db range to 40db last 2
Lots of garbage down there.
Now last one I adjusted to your parameters 40db scale 2.9m sec rise time. We are trying to see what's happening say 2 msec later to see the decay. With a rise time that wide you are actually obscuring what you are trying to see.
The actual decay is faster than you think and your rise time is effecting what you are seeing.
Probably that is why Clio default is .58 Msec risetime.
If your are going to post the 3 stooges please use the real ones and not the remake guys!
Rob :)
Ian,Quote:
Originally Posted by Ian Mackenzie
Barry was being diplomatic, you are being a real jerk and your behavior is totally unacceptable. You have been attacking anyone you think disagrees with you. Disagreement is fine, but personal attacks are absolutely not.
Put a full stop to this personal attack BS or you will be blocked from posting.
Widget
Enough of this angry BS, Ian - no one wants to read it and it gains no one any ground.
I've just deleted a ton of Ian's posts and other posts that quoted and replied to it.
I sit quiet most of the time, but I have no problem swinging a permanent BAN on such abusive attacks.
I've been here for near 20 years and its always been a friendly place to learn and grow,
but this "any disagreement is an attack on me" crap has got to stop.
I'll ban you myself if you keep up with this!
And "blaming americans" for your disagrements is too ridiculous to even address in any detail.
Just chill - and if you can't chill - LEAVE (or be cast out).
Hey! I was in the middle of reading those angry, drunken posts! :lol_fit:Quote:
Originally Posted by hjames
Its an Attitude thing. If you can discuss technical issues without insulting folks that differ from you, feel free to stay.
The site has existed for a long time while people treated each other as peers.
I am not telling you to fuck off, I am telling you to be FRIENDLY and respectful to each other.
If you gotta push people down to get your opinion across, head somewhere else.
Accept it or not, that's just how it is.
Quote:
Originally Posted by Ian Mackenzie
Hello Ian
Let's set the record straight.
Go look at post 31 you had concerns about the CSD's and thought they should be validated.
So in the same post I referenced the Clio Clinic we set up years ago. I corroborated with Zilch by verifying and changing test conditions.
I asked you to post what you have, all of my relevant test conditions were on the graphs.
Post 33
You opened with my windows were different and that made comparisons between them invalid for comparison purposes.
I changed the windows and verified the window differences were not significant.
You also said that the diaphragm decays smoothly for 2 msecs
I asked you to make changes so we could compare what I had posted
My last CSD was posted using your conditions and it looked similar to REV's
If you would have posted your CSD's with my settings we would have been able to confirm the CSD in both REV and Clio Pocket 3.0 were essentially interchangeable.
You chose not to and instead you got down right nasty.:
"I didn't ask to be told how think, to be interrogated. Why this why that, you should do this and do that? Really??
His job is to listen and learn. Above all listen and learn from others. Earl recently insulted me without thinking. It was a fuck off statement. Re educate yourself. FFS. That's unacceptable. If they can get way with that they will keep doing it because they have no cause. I will drop them point blank hard. No excuses. No prisoners."
That say's it all.
You missed an opportunity and the point.
Rob
Hello Rob and All;
I for one am keenly interested in this topic. I have for the last decade generally used my plane wave tube for my compression driver testing. I use the phase response to see where the driver becomes non linear. The visual plots you are showing are very easy to see where the ringing occurs and the extent of it.
My go to platform is TEF. That is what JBL used in the day for drivers like the 435’s. What is nice is that I can see exactly how they set up the test parameters and duplicate them for some continuity. TEF has the world’s worst waterfall graphics, it works well though.
I have just scratched the surface of REW and am quite surprised at how much functionality is buried in there.
I am also interested in applying the GedLee metric to my measurement routines. I don’t really want to buy and learn another platform and I think that Virtins virtual is the only platform that offers a somewhat turn key software pack to use GedLee? If anyone knows otherwise please share.
Responding to an Ian critique, it is true that I rarely have much to offer here. I am too busy to pursue this hobby anywhere near as much as I would like to and then to spend the time to document it and post it and defend it, just doesn’t fit into my life currently.
I don’t know if there is anything that I can do to advance this but I am in for what ever I can do.
Thanks Rob.
Barry.
Well Good luck with whatever it’s your about.
The problem is you’re now resentful and recalcitrant to anything you don’t already know. Giskard had the same problems although you tolerated it because there was something in it for you. He literally ripped a new hole every day of the week. You now gather in numbers to form a consensus as a decision making body. It’s like a Mr Wick movie. Oh John you’re out of favour with the society old buddy. The Adjudicator wants to see you. Mustn’t stir the pot…Lol. It’s amusing that you haven’t been able to sort out how to ban me……..muhhhhhhhh. OMG 😨.
Just look at yourselves suddenly waking up after you’ve been told. Bit late. The train has left the station. It’s a joke. I am not here for your entertainment. I’ve never been vilified ever and it’s not starting now.
The reality is these issues only exist statistically in this environment here. Robert has been trawling behind me for a number of years making really stupid assertions. In his own mind he needs control. He hides in the shadows. In 2003 he asserted that l shouldn’t start a 4343-4344 upgrade thread because how do l know it will make a difference? He was very defensive about. He’s been resentful ever since. I do nt want that so you can’t have it mentality. It’s called controlling behaviour and it’s been outlawed on social media and public forums. People with these personalities cling to the nearest risk because they fear the unknown. They don’t like change or unexpected events.
The world embraced the 4343-4344 thread that Ed and l wrote and thousands came to become involved. I find it very disturbing that someone in a position of authority holds that kind of dark attitude and behaviour. It’s actually scary.
I’ve compared posts and interactions from 10-15 years ago and there is a world of difference in the attitudes of the players involved. The LHS has become a very niche non technical jabber fest. In comparison to other environments this here is a totally un moderated church (sect) type gathering that now caters for a locked in residence. It’s like a small group who doesn’t like being interrupted.
“Oh what’s he talking about??? No the world is flat. Can’t you see.” There are no rules except the rules that suit on the day. A nursing home for some frail old folks.
Others are well aware and that is why they now post elsewhere. Unfortunately is too late to re build. They have moved on and they repeatedly have commented on behaviours here. The words strange come up.
You can’t go singling someone out and passing the blame. It’s doesn’t work. It’s what you’ve become. Obviously l no longer fit and l accept that. A whole new world awaits.
I wish you all fun with your JBL systems.
Hi BarryQuote:
Originally Posted by 1audiohack
What metric are you referring too? Is it related to his polar map tool?
IIRC I once read Earl Geddes state that he preferred to rely on magnitude/phase measurements rather than waterfalls and such, as he found they provided just as much info when properly analyzed.
How can you go about practically insulting people and then wishing them well? Why bother? The posts are gone now, but from what I remember reading, I recall being downright flabbergasted at your responses to Rob, you insulted the fella. Comments like: "..I don't need a reeducation.." are about as condescending as it gets. And they say Millennials, gen x and gen z's have a problem with constructive criticism... Everybody here has helped me in some way, including yourself. In the words of Rodney King (Wasn't it him?), "Can't we all just get along?" You know, I don't make sense of ~90% of what I come across here. For example: Ringing? Where/how do you see that on those graphs? And all those stats and parameters that are posted? WHOOSH! Right over my ignorant head. Even so, I still bothered to join this forum, well aware that I'd add no value. Glad I did, for if I hadn't, I wouldn't have the connections I have now, nor would I be having "fun with my JBL systems" as I am now. But I feel this forum still thrives on the technical side of things. Most of the non-technical conversations don't seem to inspire half as many responses as those which are asking for advice or guidance.Quote:
Originally Posted by Ian Mackenzie
Quote:
Originally Posted by pos
Hello pos
Do you know what software or tools he was talking about? I have always been curious why HOM's are not lets say obvious in Group Delay and CSD plots.
Unless of course they are there but are difficult to see. Like the Step vs Impulse where the Step is easier to see actual delay between the drivers.
If I remember he uses Mathlab? And he has several applications he developed like his polar maps as an example.
Hello Barry
Let's see where the thread leads. I have been looking at CSD vs Wavelet. Admittedly I find the CSD's easier to understand vs Wavelets especially the Time Preferred Plot.
Rob :)
Hello Rob and Thomas;
My limited understanding of the GedLee metric is that it measures for many? types of distortions and applies a metric to the level of each of what Dr. Geddes believes are objectionable. Some or all of these distortion objections vary with level.
I have had a couple of disagreements with Dr Geddes on the audibility and objectionability of some types of audio chain distortions. For instance, I still don't believe that his choice to use the Talking Heads song Burning Down the House was the best audio track one could use to test people's ability to discern different levels of harmonic distortion in a sound system. Be that as it may, I do believe that he is correct on many fronts and his work to create a metric to actually show, through measurement, possibly the why this one, sounds better than that one, despite what some of the measurements that we are used to using might show us, but don't or can't answer this question. I believe this has real value and I would like to explore it much further.
The only software that I am aware of that measures for GedLee Metric distortions is available from Virtins Multi Instrument. See Virtins.com. Apparently you need the Pro version that currently costs $229.95 USD and the Device Test Plan that costs an additional $100. USD.
Thet price of the Virtins software is a total non issue, it is just one more thing I will have to learn. I wish GedLee metric was available in REW. Maybe it soon will be.. I should probably just ask.
All the best.
Barry.
Hello Thomas;Quote:
Originally Posted by pos
If recall correctly Dr Geddes has interest in the phase relationship of the higher harmonics with the fundamentals? I do. This is where for instance if a diaphragm is ringing, the sound is no longer correlated with the fundamental and is just noise. This is what we see with the diamond surround drivers. The extra HF is not signal correlated, its energy seen in many frequency domain measurements is energy dispersed in the time domain.
Richard Heyser spoke about this and gave is the tools to measure this in the late sixties I think or early seventies.
I hope we get somewhere on this.
All my best.
Barry.
Hi RobQuote:
Originally Posted by Robh3606
I wouldn't know! I have never seen those elusive HOMs myself.
The M2 waveguide is surely packed with those, so much so that they stop being a problem in the end :)
Funny to think that Earl Geddes took two opposed routes for dealing with HF and LF problems: trying to suppress any defect for the HF (and ending up with a big one on axis, that he deemed irrelevant in his preferred listening scenario) while trying to mix and compensate different modes together with his multi sub approach in LF.
The M2/sub18 system is the exact opposite :D
I guess both approaches can work, as well as many others probably, as long as the system is coherent and well thought out.
I think the same goes for measurements: you can look at them in many different ways and representations, but this is still the same info. No need to fight over this.
Oh I see, interesting!Quote:
Originally Posted by 1audiohack
Is there any specific representation he is using for that?
Hello Pos BarryQuote:
Originally Posted by pos
I remember Wavelets used in the Truextent PDF. I was looking at CSDs and Wavelets to look for correlation's and differences.
https://www.audioheritage.org/vbulle...mat-Diaphragms
Here are the plots Ribbed Ti Smooth Ti Aluminum and Berylium
Also Joseph Crowe reviews
https://josephcrowe.com/blogs/news/1...er-test-review
Looking at Wavelets they seem to highlight the resonances better than the CSD's
Rob :)
Hi RobQuote:
Originally Posted by Robh3606
Oh ok, I read Barry's comment wrong and thought he was talking about harmonic distortion. It appears it was just the HF itself.
Sorry for the confusion.
Hello pos
I did as well he mentions Geede's Metric and assumed it was the distortion metric. I thought his second post was a new topic. I sure when he see's this he will clarify,
Rob :)
Hello guys;
I am re-reading a couple of Dr Geddes’ papers to refresh my memory as it has been a decade or more since I last read them. I am pretty fuzzy on exactly how his metric scores a device.
Very basically it says all distortions are not equally egregious to our perceived sound quality and some distortions become more unpleasant to our perception with increased SPL. An example (mine and not to be construed as certain) could be that second harmonic is not even noticeable until say 6% and carries a very low score with increased SPL but diffraction which may be inaudible at low SPL becomes very objectionable at high SPL would carry a much higher score in the metric.
I will come back with a better understanding of this metric after some reading. GedLee.com has all of his papers for reading if you are interested.
As I said before, a metric like this is a step forward in the attempt to measure the “why this one sounds better than that one despite what some of our normal measurements tell us”.
I think that’s the essence of this thread yeah?
Thanks all.
Barry.
Earl Geddes did bring a lot of interesting topics to the table for sure!
One I find very interesting is the instantaneous thermal compression that he classifies as a distortion, calling it "thermal distortion".
Although the 2.7sec he is using is way too long to fit Earl Geddes' definition, Erin's "Instantaneous Compression Test" is pretty discerning already:
M2: https://www.erinsaudiocorner.com/lou...mpression-test
708p: https://www.erinsaudiocorner.com/lou...mpression-test
305p: https://www.erinsaudiocorner.com/lou...mpression-test
That is an area where compression drivers and big woofers do shine :)
Hello pos
I found his thermal distortion interesting as well.
Those are interesting plots! I have a pair of 308 Mk 2 that I have next to my passive M2's and they definitely run out of gas. They don't obviously break-up they just compress like a limiter.
Agree lively is the way to go!
Rob :)
Indeed, on Erin's measurement you can see that the 305p has a limiter set somewhere between 96 and 102dB SPL at 1m, as the 102dB SPL curve is buried down low.Quote:
Originally Posted by Robh3606
I bet the 308p can go a few dB higher.
I'd be curious to see similar measurements for the LSR32 as its woofer has an integrated brake system.
But that is utterly off topic, sorry for the digression :D
Back on the topic of compression drivers, did you ever measure a difference between the different 1.5" cores?
When I compared a few of these 12 years ago I remember that the 2452 core sounded way worse to my hears compared to the 2450SL one (less "realistic" dynamic), both equipped with the exact same diaphragm.
I don't know if this was due to the magnetic field ("optimized" at half the weight, both neo), the smaller back cap or the somewhat shorter phasing plug.
That might also just be my personal bias (bigger=better)... or my failure to reliably/consistently install diaphragms :)
Quote:
Originally Posted by pos
Hello pos
No I have never compared the cores. Never had the opportunity. I only own 2 pairs of large format drivers the 476Mg's and a pair of 2453Sl's. I got them at different times and never compared them one on one. The way the 2453Sl's are built you can't swap as the the back cap and diaphragm are one piece. The ultimate cost savings I would guess. Use them from 1.5K and up and they sound great.
Rob :)