Beyond the Ariel - Page 542 - diyAudio
Go Back   Home > Forums > Loudspeakers > Multi-Way

Multi-Way Conventional loudspeakers with crossovers

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 7th May 2009, 05:20 PM   #5411
diyAudio Member
 
auplater's Avatar
 
Join Date: Apr 2005
Location: KyOhWVa tristate
Quote:
Originally posted by soongsc

This is just looking at data from different viewpoints. I'm with ScottG on his question.
CSD is not "basically a series of SPL curves at different times"...

this is an erroneous interpretation in my opinion.

What it does seem to show is how an original impulse at time zero decays over a set time interval over the bandwidth allowable under testing constraints.

To state that it's a bunch of spl curves at different times implies that the time > 0 curves can be generated independent of the time = 0 curve, which is not the case.

John L
__________________
"...His brain is squirming like a toad..." Jim Morrison
  Reply With Quote
Old 7th May 2009, 05:31 PM   #5412
Retsel is offline Retsel  United States
diyAudio Member
 
Join Date: Jan 2005
Location: Ann Arbor, Michigan
"CSDs can be compared to each other. But what does a difference tell about sound quality?"

This is exactly right. As an engineer, the CSD plots are great summaries of data, but what do they mean?

I recall that Jon Ver Halen posted CSD plots of the Lowther DX4 drivers with and without Bud Purvine's enABL treatment. The CSD plots showed more stored energy in the enABL-treated driver which suggests poorer audio behavior and sound. But those who have heard the enABL treated DX4 drivers prefer them to those who have not heard them - and not by a little bit.

Perhaps there is information that speaker designers can glean from such plots, but I am in the same camp as yourself. I look at them but I am not willing to make any conclusions about sound quality based on those plots.

Retsel
  Reply With Quote
Old 7th May 2009, 06:00 PM   #5413
pjpoes is offline pjpoes  United States
diyAudio Member
 
Join Date: Jan 2005
Location: New York
Send a message via AIM to pjpoes
Quote:
Originally posted by Retsel
"CSDs can be compared to each other. But what does a difference tell about sound quality?"

This is exactly right. As an engineer, the CSD plots are great summaries of data, but what do they mean?

I recall that Jon Ver Halen posted CSD plots of the Lowther DX4 drivers with and without Bud Purvine's enABL treatment. The CSD plots showed more stored energy in the enABL-treated driver which suggests poorer audio behavior and sound. But those who have heard the enABL treated DX4 drivers prefer them to those who have not heard them - and not by a little bit.

Perhaps there is information that speaker designers can glean from such plots, but I am in the same camp as yourself. I look at them but I am not willing to make any conclusions about sound quality based on those plots.

Retsel
These issues are common in audiophilia, and I would argue that it represents a disconnect between what we measure and what we hear. However, do need to assume that what people heard was real. It's possible that the claimed superiority is nothing more than, I spent all that time or and/or money, I want it to sound better, so it does. The other issue, the disconnect issue, has to do with interpretation of measurements. The assumption is that lower stored energy, as shown in a CSD, will equate to better sound. However, is this true? It completely possible that, as has been shown in numerous distortion audibility tests, the ENABL process creates a sort of distortion we find desirable.

An additional and related issue is, what sounds better? We can argue that better is some concrete measurable variable. We can even argue that "sounds better" is something concrete and measurable, this might be something like the results for the average person (DBT's with a statistically diverse and representative sample), but none of this means that for every individual, "sounds better" is always the same thing. I might feel that large amounts of odd order distortion is desirable, the average person may disagree, but if I prefer it, then it sounds better to me. This makes equating something like a CSD plot against qualitative terms like, "Sounds better" very difficult, if not impossible, for all individuals. I like the way Dr. Geddes has described this issue before. We can't talk about all individuals, but we can talk about the average person (this is essentially the definition of the statistical models we use, we talk in terms of means comparisons). I think that those who like Lowther drivers and/or techniques like the ENABL treatment, are not the "average" person discussed. They are a very unique sample of people, who often also like things like singled ended triode tube amps, again, not average and not mainstream. I would guess that a large scale blind listening test session with a representative sample would indicate that the average person doesn't find those better, but that a select sample does.
  Reply With Quote
Old 7th May 2009, 06:10 PM   #5414
mige0 is offline mige0  Austria
diyAudio Member
 
mige0's Avatar
 
Join Date: Jan 2005
Location: Austria, at a beautiful place right in the heart of the Alps.
Uhh ... so many comments - maybe next time I shouldn't complain (hugh)

Quote:
Originally posted by xpert



CSD is quite useless. ...It's apperance leaves room for some free style interpretations. But that means nothing but confusion. ....

have fun

Quote:
Originally posted by xpert


... CSD is about NUMBERS. ...In short, what is the meaning of that numbers the CSD presents to the DIYer?

... the representation lacks info due to the falling characteristics towards the high end, especially that data is of less use than one might project into it ... pps: and so it is for nearly any CSD I've seen yet, big sunny picture of AYERS ROCK, myths, no essence ...

Quote:
Originally posted by xpert


I won't discuss that to the end...

A CSD is in the first place some data expressed in numbers, generated by sophisticated mathematics and in the end represented as some kind of fancy graphics.
... To have a grip on a problem, I would of course prefer the data that I can know instead of looks, feels and the like.


Below the data - consisting of mere numbers - you'r asking for.
Will be looking forward to your interpretation - Mr. Xpert - *without* the use of any processing or visualisation of course.


Have fun!


Impulse reponse
Numpoints = 65535
SamplingRate = 44100 Hz
Time(s) Amplitude(V)
0.000000E+000 4.978819E-005
2.267574E-005 8.200338E-005
4.535147E-005 1.155576E-004
6.802721E-005 1.411281E-004
9.070295E-005 1.541958E-004
1.133787E-004 1.608117E-004
1.360544E-004 1.666615E-004
1.587302E-004 1.560761E-004
1.814059E-004 1.415217E-004
2.040816E-004 1.312348E-004
2.267574E-004 1.224472E-004
2.494331E-004 1.151417E-004
2.721088E-004 1.091725E-004
2.947846E-004 1.108898E-004
3.174603E-004 1.055525E-004
3.401361E-004 9.352512E-005
3.628118E-004 7.463079E-005
3.854875E-004 5.536318E-005
4.081633E-004 2.999350E-005
4.308390E-004 3.761063E-006
4.535147E-004 -1.973561E-005
4.761905E-004 -3.686181E-005
4.988662E-004 -5.279817E-005
5.215420E-004 -5.776405E-005
5.442177E-004 -6.138336E-005
5.668934E-004 -5.032311E-005
5.895692E-004 -2.638847E-005
6.122449E-004 -1.251786E-005
6.349206E-004 8.793495E-006
6.575964E-004 1.746080E-005
6.802721E-004 2.808169E-005
7.029478E-004 2.962167E-005
7.256236E-004 3.607315E-005
7.482993E-004 3.958962E-005
7.709751E-004 4.673347E-005
7.936508E-004 5.603916E-005
8.163265E-004 7.176112E-005
8.390023E-004 8.202020E-005
8.616780E-004 8.945186E-005
8.843537E-004 9.264094E-005
9.070295E-004 9.349603E-005
9.297052E-004 8.714029E-005
9.523810E-004 6.733047E-005
9.750567E-004 5.382451E-005
9.977324E-004 2.838038E-005
1.020408E-003 -1.612971E-006
1.043084E-003 -2.545841E-005
1.065760E-003 -5.232329E-005
1.088435E-003 -7.512701E-005
1.111111E-003 -8.971974E-005
1.133787E-003 -1.150971E-004
1.156463E-003 -1.343715E-004
1.179138E-003 -1.644310E-004
1.201814E-003 -1.827846E-004
1.224490E-003 -1.854525E-004
1.247166E-003 -1.770450E-004
1.269841E-003 -1.694493E-004
1.292517E-003 -1.565076E-004
1.315193E-003 -1.550580E-004
1.337868E-003 -1.566784E-004
1.360544E-003 -1.641269E-004
1.383220E-003 -1.675687E-004
1.405896E-003 -1.716496E-004
1.428571E-003 -1.902323E-004
1.451247E-003 -1.982230E-004
1.473923E-003 -2.030993E-004
1.496599E-003 -2.062584E-004
1.519274E-003 -1.999993E-004
1.541950E-003 -1.903806E-004
1.564626E-003 -1.857201E-004
1.587302E-003 -1.628930E-004
1.609977E-003 -1.546867E-004
1.632653E-003 -1.406611E-004
1.655329E-003 -1.287366E-004
1.678005E-003 -1.277409E-004
1.700680E-003 -1.166425E-004
1.723356E-003 -1.035604E-004
1.746032E-003 -8.462804E-005
1.768707E-003 -7.016260E-005
1.791383E-003 -5.838831E-005
1.814059E-003 -4.670287E-005
1.836735E-003 -4.946134E-005
1.859410E-003 -5.204065E-005
1.882086E-003 -5.341549E-005
1.904762E-003 -5.679822E-005
1.927438E-003 -5.736984E-005
1.950113E-003 -6.384457E-005
1.972789E-003 -6.850538E-005
1.995465E-003 -7.404288E-005
2.018141E-003 -7.073495E-005
2.040816E-003 -5.690942E-005
2.063492E-003 -3.756748E-005
2.086168E-003 -1.702972E-005
2.108844E-003 4.014330E-006
2.131519E-003 1.728948E-005
2.154195E-003 2.356351E-005
2.176871E-003 3.670547E-005
2.199546E-003 4.248914E-005
2.222222E-003 4.094214E-005
2.244898E-003 4.277102E-005
2.267574E-003 3.770122E-005
2.290249E-003 3.295886E-005
2.312925E-003 1.529059E-005
2.335601E-003 1.077327E-006
2.358277E-003 -2.032949E-005
2.380952E-003 -3.987244E-005
2.403628E-003 -6.032026E-005
2.426304E-003 -8.210728E-005
2.448980E-003 -9.386011E-005
2.471655E-003 -1.058793E-004
2.494331E-003 -1.126294E-004
2.517007E-003 -1.185533E-004
2.539683E-003 -1.198617E-004
2.562358E-003 -1.133495E-004
2.585034E-003 -9.533159E-005
2.607710E-003 -6.579189E-005
2.630385E-003 -3.397748E-005
2.653061E-003 -9.395209E-006
2.675737E-003 2.186144E-005
2.698413E-003 5.342256E-005
2.721088E-003 7.023686E-005
2.743764E-003 7.888185E-005
2.766440E-003 8.789096E-005
2.789116E-003 8.680589E-005
2.811791E-003 8.909500E-005
2.834467E-003 7.853291E-005
2.857143E-003 7.327774E-005
2.879819E-003 7.086909E-005
2.902494E-003 5.427035E-005
2.925170E-003 4.465881E-005
2.947846E-003 3.963392E-005
2.970522E-003 3.213442E-005
2.993197E-003 2.684135E-005
3.015873E-003 1.636058E-005
3.038549E-003 1.059276E-005
3.061224E-003 1.597342E-005
3.083900E-003 3.706929E-005
3.106576E-003 6.968741E-005
3.129252E-003 1.036784E-004
3.151927E-003 1.394112E-004
3.174603E-003 1.721324E-004
3.197279E-003 2.059761E-004
3.219955E-003 2.394627E-004
3.242630E-003 2.724809E-004
3.265306E-003 3.096776E-004
3.287982E-003 3.481819E-004
3.310658E-003 3.853184E-004
3.333333E-003 4.158767E-004
3.356009E-003 4.376785E-004
3.378685E-003 4.548866E-004
3.401361E-003 4.732991E-004
3.424036E-003 4.882468E-004
3.446712E-003 4.847987E-004
3.469388E-003 4.808728E-004
3.492063E-003 4.711728E-004
3.514739E-003 4.554593E-004
3.537415E-003 4.363849E-004
3.560091E-003 4.344968E-004
3.582766E-003 4.343704E-004
3.605442E-003 4.342516E-004
3.628118E-003 4.346456E-004
3.650794E-003 4.426065E-004
3.673469E-003 4.421764E-004
3.696145E-003 4.535863E-004
3.718821E-003 4.616954E-004
3.741497E-003 4.722354E-004
3.764172E-003 4.730160E-004
3.786848E-003 4.732915E-004
3.809524E-003 4.628926E-004
3.832200E-003 4.566626E-004
3.854875E-003 4.621565E-004
3.877551E-003 4.614130E-004
3.900227E-003 4.653154E-004
3.922902E-003 4.602826E-004
3.945578E-003 4.647135E-004
3.968254E-003 4.608776E-004
3.990930E-003 4.713660E-004
4.013605E-003 4.858016E-004
4.036281E-003 4.967210E-004
4.058957E-003 5.140108E-004
4.081633E-003 5.314787E-004
4.104308E-003 5.396187E-004
4.126984E-003 5.455728E-004
4.149660E-003 5.469846E-004
4.172336E-003 5.567462E-004
4.195011E-003 5.656570E-004
4.217687E-003 5.713064E-004
4.240363E-003 5.705704E-004
4.263039E-003 5.726222E-004
4.285714E-003 5.666830E-004
4.308390E-003 5.644757E-004
4.331066E-003 5.591016E-004
4.353741E-003 5.523742E-004
4.376417E-003 5.381099E-004
4.399093E-003 5.316719E-004
4.421769E-003 5.157728E-004
4.444444E-003 5.039459E-004
4.467120E-003 4.841660E-004
4.489796E-003 4.623049E-004
4.512472E-003 4.379607E-004
4.535147E-003 4.264317E-004
4.557823E-003 4.102869E-004
4.580499E-003 3.901879E-004
4.603175E-003 3.648174E-004
4.625850E-003 3.203221E-004
4.648526E-003 2.813121E-004
4.671202E-003 2.323559E-004
4.693878E-003 1.769494E-004
4.716553E-003 1.281274E-004
4.739229E-003 7.188765E-005
4.761905E-003 1.688794E-005
4.784580E-003 -3.127137E-005
4.807256E-003 -6.956058E-005
4.829932E-003 -1.052631E-004
4.852608E-003 -1.405042E-004
4.875283E-003 -1.673679E-004
4.897959E-003 -2.061543E-004
4.920635E-003 -2.521440E-004
4.943311E-003 -2.997394E-004
4.965986E-003 -3.482784E-004
4.988662E-003 -4.078855E-004
5.011338E-003 -4.593985E-004
5.034014E-003 -5.184414E-004
5.056689E-003 -5.729494E-004
5.079365E-003 -6.281788E-004
5.102041E-003 -6.819656E-004
5.124717E-003 -7.400267E-004
5.147392E-003 -7.907668E-004
5.170068E-003 -8.480542E-004
5.192744E-003 -9.078326E-004
5.215420E-003 -9.675220E-004
5.238095E-003 -1.019161E-003
5.260771E-003 -1.081883E-003
5.283447E-003 -1.142441E-003
5.306122E-003 -1.211244E-003
5.328798E-003 -1.279667E-003
5.351474E-003 -1.360891E-003
5.374150E-003 -1.448537E-003
5.396825E-003 -1.563946E-003
5.419501E-003 -1.687978E-003
5.442177E-003 -1.826177E-003
5.464853E-003 -1.976581E-003
5.487528E-003 -2.121865E-003
5.510204E-003 -2.271140E-003
5.532880E-003 -2.413033E-003
5.555556E-003 -2.554629E-003
5.578231E-003 -2.689729E-003
5.600907E-003 -2.813375E-003
5.623583E-003 -2.930984E-003
5.646259E-003 -3.043323E-003
5.668934E-003 -3.154562E-003
5.691610E-003 -3.261334E-003
5.714286E-003 -3.366707E-003
  Reply With Quote
Old 7th May 2009, 06:18 PM   #5415
gedlee is offline gedlee  United States
diyAudio Member
 
Join Date: Dec 2004
Location: Novi, Michigan
Michael - that list of numbers sounds great!

Just to add to what Matt said, we have to be clear on the difference between "preference" and "accuracy" - they can be dramatically different things. I only ever deal in accuracy, not preference for the simple reason that preference has no answer, but accuracy does. So figure out which camp you are in first and then you can begin to understand how one could "prefer" ENable while its is clearly less "accurate" if the CSD has large decay ridges.

CSD is not meaningless, its just not everything either. Its shows nothing that the impulse response itself doesn't show.
  Reply With Quote
Old 7th May 2009, 06:26 PM   #5416
pjpoes is offline pjpoes  United States
diyAudio Member
 
Join Date: Jan 2005
Location: New York
Send a message via AIM to pjpoes
One interpretation problem I have with CSD is that people sometimes make big deals out of very small changes. For instance, using CSD to look at room dampening. Let's say you have two graphs, one with and one without XYZ room dampening, and you show that, at the 0 time point, there is an average smoothing of 1db. In other words, if there was a peak or null, it's now reduced by 1db in severity, on average. However, as 10ms out, you see that what was once a visible decay ridge, had now been reduced to non-visible or much less visible, what does that mean. Common interpretation is that this room dampening has created a massive audible difference. In reality, I would argue it's minute, and probably not audible. The decay ridge needs to be looked at in absolute db terms as well, no? If it's already down 30 or 40db's, what will this delayed sound do to the audible signal, I would think probably very little or nothing. If the signal, which is 40db's down, is then reduced to 45db's down, is that audible? Have we really reduced the "ringing" in the room in some important or significant way?
  Reply With Quote
Old 7th May 2009, 06:28 PM   #5417
diyAudio Member
 
auplater's Avatar
 
Join Date: Apr 2005
Location: KyOhWVa tristate
Default Numbers & such

Gee Michael... I could make more of that if you had listed it in hexadecimal...

John L.
__________________
"...His brain is squirming like a toad..." Jim Morrison
  Reply With Quote
Old 7th May 2009, 06:35 PM   #5418
diyAudio Member
 
auplater's Avatar
 
Join Date: Apr 2005
Location: KyOhWVa tristate
snip
__________________
"...His brain is squirming like a toad..." Jim Morrison
  Reply With Quote
Old 7th May 2009, 06:37 PM   #5419
ScottG is offline ScottG  United States
diyAudio Member
 
Join Date: Feb 2003
Location: US
Quote:
Originally posted by auplater


CSD is not "basically a series of SPL curves at different times"...

this is an erroneous interpretation in my opinion.

What it does seem to show is how an original impulse at time zero decays over a set time interval over the bandwidth allowable under testing constraints.

To state that it's a bunch of spl curves at different times implies that the time > 0 curves can be generated independent of the time = 0 curve, which is not the case.

John L

As a statement, that is a LOT better.

However,

"..a bunch of spl curves at different times *implies*" (emphasis on "implies")

..that soongsc's statement "implies" this - is a matter of perspective. Note the word "basically" in his statement, it's a caveat in relation to the context of the discussion.
__________________
perspective is everything
  Reply With Quote
Old 7th May 2009, 06:40 PM   #5420
mige0 is offline mige0  Austria
diyAudio Member
 
mige0's Avatar
 
Join Date: Jan 2005
Location: Austria, at a beautiful place right in the heart of the Alps.
Quote:
Originally posted by Jmmlc
Hello mige0,

It seems that the window rise time of the CSD you obtained using Arta on Earl's pulses is not the same that you used previously for the CDS in your message:

http://www.diyaudio.com/forums/showt...79#post1818279

Could you indicate the parameters used to obtain those CSD:

FFT size
Window rise time
Slice shift
Smoothing

(you said in your CSD comments that the decay was excellent, could you specify the frequency above which you consider the decay being excellent...)

Best regards from Paris, France

Jean-Michel Le Cléac'h



Sharp observer you are.
Yes its been my mistake (1/3 octave smoothing at CSD processing) as I did processing of Ears IR fils at a different PC – hence not the same settings for the CSD .



Settings used now (1/12 octave smoothing at CSD processing) - same as done for Earls OS CSD plots:

Click the image to open in full size.

same pix form same JA8008 twin in OB as before

acoustic target:

Click the image to open in full size.

on axis

Click the image to open in full size.


15 deg

Click the image to open in full size.


30deg

Click the image to open in full size.


45 deg

Click the image to open in full size.



Conclusions remain the same – pretty good at the point of equalisation getting worse towards wider angles.

The point here is that we *add decay* at other than 0deg – as of the optimised filter used for close to ideal on axis response.

Thinking about it – its not exactly the same as comb filtering out of that very reason. Comb filtering would not have a (PC-XO) filter behind that "stores energy" with the intent to release it at exactly that moment where it is necessary to get close to ideal impulse shaping on axis.

Regarding Earl decay time I think about roughly 2kHz its excellent.
At least to me anything below 1msec for the 25dB shown is OK – no need to be picky about another 1/10 of a msec.

I admit that the falling FR is in favour to Earl – but I didn't want to be unfair in my comparison OB versus OS.
I – of course - had a look way down the dB scale and it looks as expected - nothing strange happening there...

The main difference is that Earls CSD are made from an (small) compression driver where as the twin Jantzen OB is a 8" – so not exactly comparable anyway – but good enough for me to get a handle on the subject as I especially was interested if we could see any mouth reflections that translate to CSD decay.


Michael
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



New To Site? Need Help?

All times are GMT. The time now is 03:35 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2