Having mercy with all the deeply depressed ESL, OB, TL, (...) lovers that now – that CMP behaviour is debunked - have to realize that what they loved for a long time, being (CMP-) distortion only.
Well – lets see if there is any medicine in the form of :
Hey, lets look at the bright side : its a feature not a bug !
I mean – all those people (me included) simply can't be wrong, can they?
With stating „CMP framing“ instead of „CMP distortion“ (
) we give it a chance that there *might* be a sonic effect introduced by CMP that *might* be beneficial as well.
LOL
Looking over the fence into „picture reproduction“ we easily realize that not a single foto would make it to our attention without some brush up – at least this is true for all those that are heavily payed for in glossy magazines and pretty books.
One of the „standards“ in picture editing to make a picture looking (at least ) as real as reality, is the use of a tool called „unsharp mask“ - or more profane : „sharpening“.
In short it is cranking up contrast for details - and more specifically – especially exactly between two picture details, in order to set them apart form each other.
This technique is also seen most pronounced in pixel art and comics – where every part of the picture is heavily „framed“, which helps to add depth and palpability to otherwise poor and lifeless sujets.
Playing the game of „unsharp mask“ / contrast well, is definitely an art by itself for any picture post processing, no matter what.
If we now stop and think for a second what all people are so crazy about, we may come to the conclusion that reproduced sound or pictures simply never ever are the real thing and hence may need some means to „better“ convince us / out senses.
So in a way „CMP framing“ might be seen in the same light as „unsharp masking“.
„CMP framing“ adds a specific „distortion“ to sound – as does sharpening (!) with pictures - no question about that.
But CMP behaviour is also pretty special in that it „frames“ any sonic event in time – a pretty unusual behaviour not seen with any „normal“ audio distortion.
At the beginning of any sound there is – after delay time – a kink in the amplitude which expresses into a bunch of high frequency content.
Same is at the end of any sound – when the CMP tail happens. Also CMP' „native FR“ time slot and CMP tail by itself are kinda framing, if seen that way.
So, „CMP framing“ possibly can't be considered being an ideal tool like „unsharp mask“ in audio, but it nevertheless may play a comparable role.
Love my „sleep well lollipop“ ?
Michael
Well – lets see if there is any medicine in the form of :
Hey, lets look at the bright side : its a feature not a bug !
I mean – all those people (me included) simply can't be wrong, can they?
With stating „CMP framing“ instead of „CMP distortion“ (
) we give it a chance that there *might* be a sonic effect introduced by CMP that *might* be beneficial as well.LOL
Looking over the fence into „picture reproduction“ we easily realize that not a single foto would make it to our attention without some brush up – at least this is true for all those that are heavily payed for in glossy magazines and pretty books.
One of the „standards“ in picture editing to make a picture looking (at least
) as real as reality, is the use of a tool called „unsharp mask“ - or more profane : „sharpening“.In short it is cranking up contrast for details - and more specifically – especially exactly between two picture details, in order to set them apart form each other.
This technique is also seen most pronounced in pixel art and comics – where every part of the picture is heavily „framed“, which helps to add depth and palpability to otherwise poor and lifeless sujets.
Playing the game of „unsharp mask“ / contrast well, is definitely an art by itself for any picture post processing, no matter what.
If we now stop and think for a second what all people are so crazy about, we may come to the conclusion that reproduced sound or pictures simply never ever are the real thing and hence may need some means to „better“ convince us / out senses.
So in a way „CMP framing“ might be seen in the same light as „unsharp masking“.
„CMP framing“ adds a specific „distortion“ to sound – as does sharpening (!) with pictures - no question about that.
But CMP behaviour is also pretty special in that it „frames“ any sonic event in time – a pretty unusual behaviour not seen with any „normal“ audio distortion.
At the beginning of any sound there is – after delay time – a kink in the amplitude which expresses into a bunch of high frequency content.
Same is at the end of any sound – when the CMP tail happens. Also CMP' „native FR“ time slot and CMP tail by itself are kinda framing, if seen that way.
So, „CMP framing“ possibly can't be considered being an ideal tool like „unsharp mask“ in audio, but it nevertheless may play a comparable role.
Love my „sleep well lollipop“ ?
For what it is worth - enjoy !
http://www.kinotechnik.edis.at/page...ine/CMP_Behaviour–Bridging_Points_of_View.pdf
Michael
Michael
Last edited:
Hi Dave
I've been sparse at Elias thread as I think CMP tis quite OT there - same as I think it possibly was at Lynn's thread.
Anyway.
I know your position pretty well, but in case of you'd just like to battle a little bit - welcome !
One thing I ask you for though - even if falling again in "heavy bubble burst mood", please stay with arguments that have more substance than "it's proven by now" - at least at those points where I have proven otherwise.
No problem to question my "prove" though - but you have to become far more specific than what I have read from you by now
If this looks to be too much of a burden for you
- *I'd* rather prefer not to discuss CMP issue with you. I assure you that I'm pretty well aware that what I claim with my conceptualization of CMP behaviour plus its widespread involvement plus its implications on "frequency response concept", puts a loooot upside down in audio - so - no need to tell me that over and over again.
Michael
Last edited:
From what has been presented on this "CMP concept", there is no "proof" yet to question.
Dave
Great
With all due respect, Micheal, starting new threads smells a lot like the avoiding the issues as was the case with the Enable nonsense. You leave me openly baffled.
An externally hosted image should be here but it was not working when we last tested it.
Not sure what you mean - how many dedicated CMP threads do you count for now ?
At Lynns thread I was told pretty clearly that CMP is way OT - though not by Lynn.
At Elias thread measuring / analyzing / visualization with wavelets it the main thing - so CMP is slightly OT there too.
Also - this is a "quiet corner" not in the focus of most - so no accuse of prestigious acting
- good place for elaborating on CMP - no ?Michael
Last edited:
FR is just one way of looking at data, and supports the design process. It is not a way of evaluating whether a system is good or not. So how does CMP work in the design process? This is what I consider very impractical. Only people doing that need to publish some kind of thesis might consider something impractical as this for practical purposes.
Yes I fully agree on „FR is just one way of looking at data“
But you have to be aware that FR – with non-CMP systems is a determination of spectral ditribution along the time line too.
With CMP systems this is no longer the case.
The practical conclusions out of this are possibly marginal – as said – ESL lovers havent noticed any bad for long time – OB lovers same thing though less „old“ .
I know you are after „results“.
But regarding CMP anybody is an „early adapter“ - right now its no more than a differnt point of view.
Michael
Love that illustration. It certainly presents good picture of systematically revealing different characteristics by using different data presentation.
With EnABL, I don't think it's totally nonsense, just not optimized in a way to have clearly measureable improvements. Same with cables, I have found characteristics of interconnects clearly measureable, tuneable, explainable corelation with listening impressions.
Well, first step in scientific research is presenting measurement data explaining CMP in it's actual existance. If you can pass that point, then you've at least gotton to first base.
There is actually nothing more to do in this respect - no more than I have presented in my paper and is well known for long time under the the synonym "constructive and destructive interference" .
So CMP already is proven to be real - actually for a long time - its the acceptance of my point of view that is under question - and, to some extent - the impacts concerning correctability.
The point that is "new" is to put - and see - that behaviour as a "CMP system ".
You know - usually any reflections get cut out of the picture as "ill room interaction" that has nothing to do with "excellent speaker performance" - so to say
Well, having learned that with horns there is a loooot of looped reflections involved and with OB (ESL, Tapped Horn, TM ...) as well, it deserved a concept (IMO) to describe and pin point those effects and its impacts.
"CMP concept" is just that.
Practically any speaker design has some CMP distortion mixed in - simply due to non optimal diffraction alignment in this respect.
To find out what that means - in terms of audible perception - will probably become just another never ending story.
Michael
First of all, your simulations have nothing to do with the what is happening in real world audio certainly it is not possible to take a measurement and model it using the examples you have shown.
Second, in order to prove that what you call CMP exists, you have to take any section of a measurement and prove that it's minimum phase. This has not been done.
So basically, you are not on first base yet. If I am incorrect, point to the specific section in your report that addresses the two issues above. Otherwise, you are still in the imaginary stage.
Second, in order to prove that what you call CMP exists, you have to take any section of a measurement and prove that it's minimum phase. This has not been done.
So basically, you are not on first base yet. If I am incorrect, point to the specific section in your report that addresses the two issues above. Otherwise, you are still in the imaginary stage.
I do not agree -
My simus show CMP behaviour in its most pure form - anything else would be a waste of time.
That CMP may *usually* not show up in this pure form "in real world audio" does not stand against at all.
You see *I can* - and you also see that measurements and simus are perfectly correlating.
Ease to tell that simus - as I set it up - are valid representations to visualize CMP behaviour hence
As said - no further prove needed - its all at the table.
And its there actually for a long time.
Michael
Last edited:
The simulation you show are only result of delay of the same spectral content. This is never the case in audio. Of course you can present what you consider a pure form, but it is only imaginative ideal condition. This information cannot help any development work in the real world simply because you cannot extract minimum phase data. If you can take a set of measurement data and show that minimum phase data can be extracted from that at at least two points along the data, you at least have one person other than youself convince.I do not agree -
My simus show CMP behaviour in its most pure form - anything else would be a waste of time.
That CMP may *usually* not show up in this pure form "in real world audio" does not stand against at all.
You see *I can* - and you also see that measurements and simus are perfectly correlating.
Ease to tell that simus - as I set it up - are valid representations to visualize CMP behaviour hence
As said - no further prove needed - its all at the table.
And its there actually for a long time.
Michael
If you can specifically point out the pages you are referring to, certainly it would help the discussion, and point out what I have missed. I also beg for someone else that sees a corelation between theory and measurement in your report to raise a hand.
I can wait for a week to look for a raised hand. If no hands are raised, and no specific pages of the report to focus discussion on, then I guess it's not worth further discussion until there is more information available.
I can wait for a week to look for a raised hand. If no hands are raised, and no specific pages of the report to focus discussion on, then I guess it's not worth further discussion until there is more information available.
Dont hold you breath !
-----------
As for all that min phase buzz:
Go back into the discussion if you are interested in.
For me the whole output of that discussion about CMP being classified as min phase or non min phase is pretty moot.
You will see some proponents declairing CMP being „min phase“ and some propenents decraring CMP being „non min phase“ end even some proponents that flip betwee camps within months.
LOL
All of that people are way more skilled in signal analysis in math terms than I am – so I simply leave the head banging regarding min phase or non min phase to them
*Consecutive* min phase tells it all IMO - its my „Solomon solution“.
The input signals are well defined - the systems behaviour is well defined - the output is well defined and all correlates well with measurements – also implications are outlined clearly - nothing else to prove!
In my point of view you are hunting wind mills – simply wrap your head around the statement that „FR concept“ is void with CMP.
As soon as you have accepted that – spectral distribution along time line tells a different story to you – be it CSD or wavelet analysis or whatever time frequency plot.
Michael
Last edited:
I see many things in data, I just don't give them funny names.Dont hold you breath !
-----------
As for all that min phase buzz:
Go back into the discussion if you are interested in.
For me the whole output of that discussion about CMP being classified as min phase or non min phase is pretty moot.
You will see some proponents declairing CMP being „min phase“ and some propenents decraring CMP being „non min phase“ end even some proponents that flip betwee camps within months.
LOL
All of that people are way more skilled in signal analysis in math terms than I am – so I simply leave the head banging regarding min phase or non min phase to them
*Consecutive* min phase tells it all IMO - its my „Solomon solution“.
The input signals are well defined - the systems behaviour is well defined - the output is well defined and all correlates well with measurements – also implications are outlined clearly - nothing else to prove!
In my point of view you are hunting wind mills – simply wrap your head around the statement that „FR concept“ is void with CMP.
As soon as you have accepted that – spectral distribution along time line tells a different story to you – be it CSD or wavelet analysis or whatever time frequency plot.
Michael
Different data show things in different ways, I'm not so narrow minded as to think that only one set of data tells everything. I accept anything that I find usefull and practical.
So, since you have some fit about that "minimum phase" thing, are you going to remove it from your theory? Call it CP?
Even Jesus had to perform miracles to get followers.
Last edited:
Call it CP?
A good one !
But actually its not under question that the basis are min-phase systems
Even Jesus had to perform miracles to get followers.
BEWARE from followers !
Last edited:
NOTE: This post and many after it have been moved from another thread. If the discussion is disjointed, remember that some posts (over 100) have come from another thread___________________________________________________________________________
Another term that IMO pops up way too often in this thread without reflecting about.
*Stored energy* - uff.
This term actually is a junk yard nothing more.
I mean - only because it was coined by "SL audio icon", no one seems to have the slightest issue with that B*S ? - where are all the technicians and experts and scientists that go for nothing less than a precisely pin point definition ???
Resonance effects sub-summed under "stored energy" fundamentally differ in cause as they differ in effect
And as for the CSD before and after EQing - well it should be no surprise that often this has been discussed already.
Boys, I really don't get it !
Michael
Last edited by a moderator:
Michael, what do you not get ? Yes, energy storage from driver missbehaviour, crossover problems and cabinet related items differ in cause and effect. Still, we have being able in Essex to build a speaker that was nearly perfect over a very wide area of the frontal hemisphere. Wherever we measured ( eccept in the low bass and the uppermost treble ) we got a perfect copy of the impulse response of the B&K measurement mic under unechoic conditions. Needless to say, that the waterfall looked like El Capitan. All 3 causes for energy storage could be ameliorated with digital EQ. And
yes, this is not new too but that was in 1993, long before this forum even existed. What i am trying here is to do something similiar ( in result ) but without the complexity of an active, digital system. Oh, yes, the Essex speaker imaged like crazy too and we all found the effect accurate and natural. Call me a fool if you wish but do not judge my work until i am ready.
yes, this is not new too but that was in 1993, long before this forum even existed. What i am trying here is to do something similiar ( in result ) but without the complexity of an active, digital system. Oh, yes, the Essex speaker imaged like crazy too and we all found the effect accurate and natural. Call me a fool if you wish but do not judge my work until i am ready.
My posting was not "only" aimed at you - but for the part you brought up:
I was not referring to "practical" differences regarding this unbelievable "stored energy" term.
I was referring to the basic and fundamental difference in resonant behaviour as such - those parts that show up in the before mentioned cepstrum analysis (and which I call CMP behaviour) versus that parts that do *not* show up in such analysis.
more precisely: real resonance and "layman resonance" - or in other words: resonance that builds up continuously versus "resonance" that has discontinuous behavior along the time line (standing waves from a transmission line speakers for example)
"ein himmelweiter Unterschied"
But as may times - I have my difficulties with the very explanations that are brought up to support a position.
Michael
I was not referring to "practical" differences regarding this unbelievable "stored energy" term.
I was referring to the basic and fundamental difference in resonant behaviour as such - those parts that show up in the before mentioned cepstrum analysis (and which I call CMP behaviour) versus that parts that do *not* show up in such analysis.
more precisely: real resonance and "layman resonance" - or in other words: resonance that builds up continuously versus "resonance" that has discontinuous behavior along the time line (standing waves from a transmission line speakers for example)
"ein himmelweiter Unterschied"
The result is your choice that shall fit your taste - I certainly would not question that.
But as may times - I have my difficulties with the very explanations that are brought up to support a position.
Michael
Last edited:
Yes, its groundbreaking but sadly I did not find out how to cure common cold with it
But to be serious:
Possibly *if* we look closer at the underlaying mechanisms we might avoid to continue in making a loot if wrong assumptions and statements (like the "least diffraction bogus" for example).
Lets take that always said statement in #352
http://www.diyaudio.com/forums/multi-way/178350-zdl-8.html#post2444347
Its simply wrong put.
EQ does *not* need any *experienced hand* - it needs *understanding* whats going on!
Make FR as flat as flat as can be and you are done. Definitely.
Except:
When there are effects involved that are *not* "real resonances" like the mentioned TM "resonance" or any other effects caused by delay.
So - *also* to understand what this cepstrum analysis thing actually is telling does not hurt - quite in contrary IMO
In the end its not exactly "mucking up" but rather a call to come back to more solid ground instead to just throw a lot of "name dropping" on the subject - I mean - Joachim and most of the contributors are certainly no novice audio guys.
Michael
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.