Beyond the Ariel - Page 553 - 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 12th May 2009, 03:15 AM   #5521
diyAudio Member
 
Robh3606's Avatar
 
Join Date: Nov 2002
Location: Destiny
"But does it move the end in the process of calculating the CSD? What I've seen is "no". As a matter of fact, CLIO could never show the resolution required from one of our suppliers."

Hello soongsc

Not sure what you mean. It clearly says in the manual that when you run a CSD plot the start and stop times determine the data processed so only the data in the selected time window is processed. Could be when they did an update they changed it?? I have Clio 7.13 running now.

You select a widow and the program runs a CSD for that window. You can trim off the initial delay from the standard 1 meter spacing and then select a point before the first reflection. So you effectively trim both sides. You can move this window anywhere in the measurement and run a CSD plot for that section of the measurement.

As far as resolution the version I have seems just fine. I can enter a time shift value and effectively zoom into the first couple of milliseconds to see what the initial decay looks like if I want more detail.

Rob
  Reply With Quote
Old 12th May 2009, 03:48 AM   #5522
soongsc is online now soongsc  Taiwan
diyAudio Member
 
soongsc's Avatar
 
Join Date: Mar 2005
Location: Taiwan
Quote:
Originally posted by Robh3606
"But does it move the end in the process of calculating the CSD? What I've seen is "no". As a matter of fact, CLIO could never show the resolution required from one of our suppliers."

Hello soongsc

Not sure what you mean. It clearly says in the manual that when you run a CSD plot the start and stop times determine the data processed so only the data in the selected time window is processed. Could be when they did an update they changed it?? I have Clio 7.13 running now.

You select a widow and the program runs a CSD for that window. You can trim off the initial delay from the standard 1 meter spacing and then select a point before the first reflection. So you effectively trim both sides. You can move this window anywhere in the measurement and run a CSD plot for that section of the measurement.

As far as resolution the version I have seems just fine. I can enter a time shift value and effectively zoom into the first couple of milliseconds to see what the initial decay looks like if I want more detail.

Rob
Normally the end of the window is not moved in the process of generating the CSD, and it is always placed before the first reflection. Only the beginning of the window is moved. How much it is moved when it generates a slice in the CSD normally depends on the sample time interval. However, if you move the window manually, the different software might operate differently. For example, in SoundEasy, once you have the window width set, if you move the beginning, then the end will move with it, which will cause the end to include the reflection at some point in time. If I remember correctly, PRAXIS moves the beginning and end independently; how CLIO does it, I cannot remember. One thing about CLIO I remember is that we could never get it to show good CSD resolution in the 0.4ms range.
__________________
Hear the real thing!
  Reply With Quote
Old 12th May 2009, 04:18 AM   #5523
diyAudio Member
 
Robh3606's Avatar
 
Join Date: Nov 2002
Location: Destiny
"For example, in SoundEasy, once you have the window width set, if you move the beginning, then the end will move with it, which will cause the end to include the reflection at some point in time."

OK I see what you are getting at.

"If I remember correctly, PRAXIS moves the beginning and end independently;"

Clio is like Praxis.

"One thing about CLIO I remember is that we could never get it to show good CSD resolution in the 0.4ms range."

OK I don't see why?? You can manualy set a Time Slice widow and use that to Zoom in so to speak. I can set up a 50usec slices and run 30 of them so I see the first 1.5msec in 50usec slices.

Here is an example.

Rob
Attached Images
File Type: jpg zoomed 05 csd.jpg (85.9 KB, 393 views)
  Reply With Quote
Old 12th May 2009, 05:49 AM   #5524
soongsc is online now soongsc  Taiwan
diyAudio Member
 
soongsc's Avatar
 
Join Date: Mar 2005
Location: Taiwan
Well, that's still 3 times the time interval for one, and at sampel rate of 96KHz, it should be possible to have a time slice of about 10.4uS. Second, it still samples part of the main impulse. So the hill you see is not exactly starting when the impulse hits zero. This is why it becomes very confusing. Idealy we want to look at how things respond starting from when the main impulse first zeros out.
__________________
Hear the real thing!
  Reply With Quote
Old 12th May 2009, 07:21 AM   #5525
xpert is offline xpert  Afghanistan
Account Disabled
 
Join Date: Apr 2009
Quote:
Originally posted by ScottG
Completely wrong.
...
I guess basically what I'm saying here is that if you are going to drop an "expert" *opinion* into the discussion, that you should probably read that discussion a little more carefully.
Don't take that xpert attitude to serious. Don't let it bring You into anger. I'm not different from You in that I'm quite convinced that I have to say a thing about the topic. I read Your post.

CSD seems to be a concept not known thoroughly enough to derive conclusions from it. CSD is completely based on the impulse response. The impulse response is completely derived from amplitude/phase over frequency. To make a CSD one needs not more than a frequency response! CSD is a representation of STEADY STATE frequency response of amplitude and phase! Check it out.

I did. That's why I am an "xpert" (just kidding)
  Reply With Quote
Old 12th May 2009, 10:02 AM   #5526
diyAudio Member
 
Join Date: Apr 2007
Hi,

Quote:
Originally posted by john k...
The FFt length, which is completely unrelated, sets the frequency resolution.
it does not. Or, in other words, the real - I call it effective - frequency resolution is determined by the length of the unpadded impulse response. The zero-padding (to get a power of 2 for FFT) does only add interpolated samples to the result.

Bye


Baseballbat
  Reply With Quote
Old 12th May 2009, 10:41 AM   #5527
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.
I'm happy to see that we (most) finally came to a point where the resonances seen in CSD got accepted as such – and also some cevelets in taking CSD measurements were outlined.

In fact - a measurement / processing method that basically relies on infinite duration of sine waves that are processed – isn't particularly designed to show what's happening when they disappear (to cite someone, I highly respect).


Anyway – I'd like to do an experiment.

As I have the tools available to change acoustic IR of a speaker at will by "a simple click" – I'd like to know if there is a benefit we can gain for open baffles from that.
(it also might put some shade of light on any other multi way or line array speakers further on)

What I outlined some postings ago is that I got off axis decay above the first baffle peak (best seen at the 1/3 octave smoothed pix).
My explanation then was that its originating from the fact that I shaped impulse response in a way to integrate the delayed back wave close to ideal with the front wave to get perfect IR at on axis (also shown by measurement).
- Meaning - I added "energy storage" by equalisation.

Now – if I'm right here – we could ask if it might be better *not* to introduce that specific equalisation
- meaning – that we rather accept some impulse smearing on axis in favour to non-added decay off axis.

I simply could try - but really would like to hear your thoughts first.


Michael
  Reply With Quote
Old 12th May 2009, 11:27 AM   #5528
xpert is offline xpert  Afghanistan
Account Disabled
 
Join Date: Apr 2009
Quote:
Originally posted by mige0

In fact - a measurement / processing method that basically relies on infinite duration of sine waves that are processed – isn't particularly designed to show what's happening when they disappear (to cite someone, I highly respect).
...
Michael
Michael,

Every sine wave is in its own concept lasting forever. It is its very mathematical/physical heart. To describe changes within a spectrum needs example given a Fourier analysis. But this would work on frequencies too, that rely on its infinite duration. Thats the concept of "frequency".

Your conclusion cited above is misleading. As far as the system is linear, there is NO conceptual difference between a steady state and any other state, example given "shut off". This is known for hundreds of years now. It is part of the education of every engineer in Europe ever since.

Quote:
Originally posted by mige0
I'm happy to see that we (most) finally came to a point where the resonances seen in CSD got accepted as such – and also some cevelets in taking CSD measurements were outlined.
That's quite bold. I've shown that effects that are non resonant will make a ridge in the CSD by introducing group delay. Example given different pathlengths in a source (the dipole baffle, righty right). You didn't differentiate such from true resonances! Which made people accept Your claims then?

But - I give up not to make that fool out of me ...
  Reply With Quote
Old 12th May 2009, 12:25 PM   #5529
dlr is offline dlr  United States
diyAudio Member
 
Join Date: Mar 2005
Location: Canton, MA
Quote:
Originally posted by Robh3606
[B]"For example, in SoundEasy, once you have the window width set, if you move the beginning, then the end will move with it, which will cause the end to include the reflection at some point in time."

OK I see what you are getting at.

"If I remember correctly, PRAXIS moves the beginning and end independently;"

Clio is like Praxis.
I like systems that have independent time markers, since there are times when some adjustment is warrented. Once set, though, the window is fixed during any post processing for most systems. Soongsc is correct about that. I can't see how a fully sliding window would be useful in typical acoustic measurements.

One thing to note is that some software, MLSSA, LAUD and some others (not Praxis I think), indicate that loss of resolution in the successive CSD slices due to window length reduction by "wrapping" it so-to-speak. The low end is cut off in the display, example below.

Click the image to open in full size.

Another often overlooked issue is the vertical scale. A lot of software by default sets this to less than 30db maximum and I've seen 24db used. I like to set it at 40db as seen in graph above. Some resonances get buried under the "floor" if the scale is reduced like this.

That's part of why it's difficult to compare measurements from different tests, aside from all of the other confounding factors. Simply using a window doesn't make them all equal. CSDs are too prone to misinterpretation for several of reasons. To me they are of limited use.

Dave
  Reply With Quote
Old 12th May 2009, 12:25 PM   #5530
Rudolf is offline Rudolf  Germany
diyAudio Member
 
Rudolf's Avatar
 
Join Date: Mar 2003
Location: Germany
Quote:
Originally posted by xpert

I've shown that effects that are non resonant will make a ridge in the CSD by introducing group delay. Example given different pathlengths in a source (the dipole baffle, righty right). You didn't differentiate such from true resonances!
Isn΄t resonance in itself a special manifestation of group delay? Like comb filtering with a constant 360° phase delay at the resonant frequency? If that would be the case, we would have gone full circle: Everything in a CSD would be group delay. Or: Every group delay could be interpreted as some special expression of resonance.
__________________
www.dipolplus.de
  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 08:58 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