Emu XL-7 & MP-7 User's Group

this description of z-plane filters sounds like fun:
 	What are Z-Plane filters and how do they sound ?
 	http://www.tweakheadz.com/es1faq.html#Z%20Plane%201

but it seems like the command station's implementation of the z-plane 
filters is neutered... no 3 dimensional cubes... no 2 dimensional 
squares... not even a 1 dimensional line that i've found... or am i 
missing something?


-- 
         ...atom

  ________________________
  http://atom.smasher.org/
  762A 3B98 A3C3 96C9 C6B7 582A B88D 52E4 D9F5 7808
  -------------------------------------------------

 	"There is a theory which states that if ever anyone discovers
 	 exactly what the Universe is for and why it is here, it will
 	 instantly disappear and be replaced by something even more
 	 bizarrely inexplicable. There is another theory which states
 	 that this has already happened."
 		-- Douglas Adams

--- In xl7@yahoogroups.com, Atom Smasher <atom@...> wrote:
>
> this description of z-plane filters sounds like fun:
>  	What are Z-Plane filters and how do they sound ?
>  	http://www.tweakheadz.com/es1faq.html#Z%20Plane%201
> 
> but it seems like the command station's implementation of the z-plane 
> filters is neutered... no 3 dimensional cubes... no 2 dimensional 
> squares... not even a 1 dimensional line that i've found... or am i 
> missing something?


I am wondering the very same thing myself. I'm looking at the manual,
and it's talking about a Morph parameter, which is how the filter
should change over time, but I can't find anything in the machine
itself. I get the feeling the manual section was copied from somewhere
else, and is rather misleading. Please, someone tell us that it is not
so!?

I remember reading about the morpheous and the emu, and as far as i am
able to tell, the 'Morph' stuff is actually a combination of the
frequency and the resonance controls, as there wasn't any other
options that seemed to make sense..

i could of course be completely off the mark with that though.

[zar]

> I am wondering the very same thing myself. I'm looking at the manual,
> and it's talking about a Morph parameter, which is how the filter
> should change over time, but I can't find anything in the machine
> itself. I get the feeling the manual section was copied from somewhere
> else, and is rather misleading. Please, someone tell us that it is not
> so!?
>

Z-plane doesn't actually have anything to do with "morphing" or interpolation.  They are filters designed around the mathematical construct of the z-plane, which is the discrete version of the continuous s-plane.  Gross oversimplification is that instead of representing filter coefficients in the time domain, they're represented in the frequency domain (by using the z-transform, the discrete equivalent of the Laplace transform).

Techie info here:
Z-transform - Wikipedia, the free encyclopedia
S plane - Wikipedia, the free encyclopedia
Z Transform and Laplace Transform, applet showing s plane to z plane mapping.

Unfortunately, the z-plane filters appeared first in a product called "Morpheus", highlighting the morphing ability of these new filters.  Since marketing something techie is usually pretty hard, E-mu pretty much said "Z-plane filters let you morph between X and Y!"  I don't blame them since can you imagine reading one of the above wikipedia articles on a piece of advertising copy?

That said, there *is* a grain of truth to the whole "Z = morphing" misconception.  By modeling your entire filter as individual poles and zeros, you could modify each pole or zero individually, letting you dynamically alter a filter into a completely different kind of filter.  The filter chip in the E-mu products has one hardware interpolator, allowing it to go from one set of filter coefficients to another in hardware.  This also means that it only has one hardware degree of freedom (hence the only dynamic parameter being Fc).  To do other realtime manipulations, the software has to load a whole other set of coefficients, then use the hardware to switch to them smoothly.  This takes a lot of CPU, and can sound choppy if not updated frequently enough.  It is, however, really really cool when it sounds right.

-Aaron



----- Original Message ----
From: Alwyn <zardac@...>
To: xl7@yahoogroups.com
Sent: Sunday, January 11, 2009 3:08:03 PM
Subject: Re: [xl7] Re: z-plane - is it "all that"?

I remember reading about the morpheous and the emu, and as far as i am
able to tell, the 'Morph' stuff is actually a combination of the
frequency and the resonance controls, as there wasn't any other
options that seemed to make sense..

i could of course be completely off the mark with that though.

[zar]

> I am wondering the very same thing myself. I'm looking at the manual,
> and it's talking about a Morph parameter, which is how the filter
> should change over time, but I can't find anything in the machine
> itself. I get the feeling the manual section was copied from somewhere
> else, and is rather misleading. Please, someone tell us that it is not
> so!?
>

Z-plane doesn't actually have anything to do with "morphing" or interpolation.  They are filters designed around the mathematical construct of the z-plane, which is the discrete version of the continuous s-plane.  Gross oversimplification is that instead of representing filter coefficients in the time domain, they're represented in the frequency domain (by using the z-transform, the discrete equivalent of the Laplace transform).

Techie info here:
http://en.wikipedia.org/wiki/Z-transform#Zeros_and_poles  -  Z-transform - Wikipedia
http://en.wikipedia.org/wiki/S-plane   -  S plane - Wikipedia
http://dspcan.homestead.com/files/Ztran/zlap.htm  -  Z Transform and Laplace Transform, applet

Unfortunately, the z-plane filters appeared first in a product called "Morpheus", highlighting the morphing ability of these new filters.  Since marketing something techie is usually pretty hard, E-mu pretty much said "Z-plane filters let you morph between X and Y!"  I don't blame them since can you imagine reading one of the above wikipedia articles on a piece of advertising copy?

That said, there *is* a grain of truth to the whole "Z = morphing" misconception.  By modeling your entire filter as individual poles and zeros, you could modify each pole or zero individually, letting you dynamically alter a filter into a completely different kind of filter.  The filter chip in the E-mu products has one hardware interpolator, allowing it to go from one set of filter coefficients to another in hardware.  This also means that it only has one hardware degree of freedom (hence the only dynamic parameter being Fc).  To do other realtime manipulations, the software has to load a whole other set of coefficients, then use the hardware to switch to them smoothly.  This takes a lot of CPU, and can sound choppy if not updated frequently enough.  It is, however, really really cool when it sounds right.

-Aaron

(updated with links)

----- Original Message ----
From: Alwyn <zardac@...>
To: xl7@yahoogroups.com
Sent: Sunday, January 11, 2009 3:08:03 PM
Subject: Re: [xl7] Re: z-plane - is it "all that"?

I remember reading about the morpheous and the emu, and as far as i am
able to tell, the 'Morph' stuff is actually a combination of the
frequency and the resonance controls, as there wasn't any other
options that seemed to make sense..

i could of course be completely off the mark with that though.

[zar]


> I am wondering the very same thing myself. I'm looking at the manual,
> and it's talking about a Morph parameter, which is how the filter
> should change over time, but I can't find anything in the machine
> itself. I get the feeling the manual section was copied from somewhere
> else, and is rather misleading. Please, someone tell us that it is not
> so!?
>



      


------------------------------------

Yahoo! Groups Links

so... the way it's implemented on the command stations, it comes down to 
_one_ dynamic (aka: tweakable) parameter: filter cutoff. and instead of 
being able to set our own filter types/parameters at either end of the 
knob, each filter in the drop-down list is built around one or two types 
of filters hard-coded at the factory...? and the way some of the filter 
models are designed we can expect results that are very different from a 
traditional VFC as we turn the "cuttoff" knob...?

did any of the emu synths implement it as described as lines, squares & 
cubes with arbitrary (user configurable) filters at the corners? or are 
they all implemented like the command stations?


On Sun, 11 Jan 2009, Aaron Eppolito wrote:

> That said, there *is* a grain of truth to the whole "Z = morphing" 
> misconception.  By modeling your entire filter as individual poles and 
> zeros, you could modify each pole or zero individually, letting you 
> dynamically alter a filter into a completely different kind of filter. 
> The filter chip in the E-mu products has one hardware interpolator, 
> allowing it to go from one set of filter coefficients to another in 
> hardware.  This also means that it only has one hardware degree of 
> freedom (hence the only dynamic parameter being Fc).  To do other 
> realtime manipulations, the software has to load a whole other set of 
> coefficients, then use the hardware to switch to them smoothly.  This 
> takes a lot of CPU, and can sound choppy if not updated frequently 
> enough.  It is, however, really really cool when it sounds right.







-- 
         ...atom

  ________________________
  http://atom.smasher.org/
  762A 3B98 A3C3 96C9 C6B7 582A B88D 52E4 D9F5 7808
  -------------------------------------------------

 	"[The] feminist agenda is not about equal rights for women.
 	 It is about a socialist, anti-family political movement that
 	 encourages women to leave their husbands, kill their
 	 children, practice witchcraft, destroy capitalism and become
 	 lesbians."
 		-- Rev. Pat Robertson, 1992

Thank you for that, Aaron....I almost forgot that is what the z-filter it referred to and all those years of engineering of sampled data theory...

Glad to see you back again....Happy New Year !

Were you able to help Jane with her question on detecting available output ports.....she just finished an editor for the P2K that is extends it's usefulness into the rest of the XL-X family....

Bob
El Segundo, CA

  ----- Original Message ----- 
  From: Aaron Eppolito 
  To: xl7@yahoogroups.com 
  Sent: Sunday, January 11, 2009 4:50 PM
  Subject: Re: [xl7] Re: z-plane - is it "all that"?


  Z-plane doesn't actually have anything to do with "morphing" or interpolation. They are filters designed around the mathematical construct of the z-plane, which is the discrete version of the continuous s-plane. Gross oversimplification is that instead of representing filter coefficients in the time domain, they're represented in the frequency domain (by using the z-transform, the discrete equivalent of the Laplace transform).

  Techie info here:
  http://en.wikipedia.org/wiki/Z-transform#Zeros_and_poles - Z-transform - Wikipedia
  http://en.wikipedia.org/wiki/S-plane - S plane - Wikipedia
  http://dspcan.homestead.com/files/Ztran/zlap.htm - Z Transform and Laplace Transform, applet

  Unfortunately, the z-plane filters appeared first in a product called "Morpheus", highlighting the morphing ability of these new filters. Since marketing something techie is usually pretty hard, E-mu pretty much said "Z-plane filters let you morph between X and Y!" I don't blame them since can you imagine reading one of the above wikipedia articles on a piece of advertising copy?

  That said, there *is* a grain of truth to the whole "Z = morphing" misconception. By modeling your entire filter as individual poles and zeros, you could modify each pole or zero individually, letting you dynamically alter a filter into a completely different kind of filter. The filter chip in the E-mu products has one hardware interpolator, allowing it to go from one set of filter coefficients to another in hardware. This also means that it only has one hardware degree of freedom (hence the only dynamic parameter being Fc). To do other realtime manipulations, the software has to load a whole other set of coefficients, then use the hardware to switch to them smoothly. This takes a lot of CPU, and can sound choppy if not updated frequently enough. It is, however, really really cool when it sounds right.

  -Aaron

  (updated with links)

  ----- Original Message ----
  From: Alwyn <zardac@...>
  To: xl7@yahoogroups.com
  Sent: Sunday, January 11, 2009 3:08:03 PM
  Subject: Re: [xl7] Re: z-plane - is it "all that"?

  I remember reading about the morpheous and the emu, and as far as i am
  able to tell, the 'Morph' stuff is actually a combination of the
  frequency and the resonance controls, as there wasn't any other
  options that seemed to make sense..

  i could of course be completely off the mark with that though.

  [zar]

  > I am wondering the very same thing myself. I'm looking at the manual,
  > and it's talking about a Morph parameter, which is how the filter
  > should change over time, but I can't find anything in the machine
  > itself. I get the feeling the manual section was copied from somewhere
  > else, and is rather misleading. Please, someone tell us that it is not
  > so!?
  >

  ------------------------------------

  Yahoo! Groups Links



   

[Non-text portions of this message have been removed]

During a given note's lifespan, yes, there is one tweakable parameter, usually tied to Fc.  That said, every new note can have a completely different set of filter coefficients.  There are plenty of ways around this, though, the easiest being putting different filters on different layers...

-Æ


----- Original Message ----
From: Atom Smasher <atom@...>
To: xl7@yahoogroups.com
Sent: Sunday, January 11, 2009 5:02:35 PM
Subject: Re: [xl7] Re: z-plane - is it "all that"?

so... the way it's implemented on the command stations, it comes down to 
_one_ dynamic (aka: tweakable) parameter: filter cutoff. and instead of 
being able to set our own filter types/parameters at either end of the 
knob, each filter in the drop-down list is built around one or two types 
of filters hard-coded at the factory...? and the way some of the filter 
models are designed we can expect results that are very different from a 
traditional VFC as we turn the "cuttoff" knob...?

did any of the emu synths implement it as described as lines, squares & 
cubes with arbitrary (user configurable) filters at the corners? or are 
they all implemented like the command stations?

On Sun, 11 Jan 2009, Aaron Eppolito wrote:

> That said, there *is* a grain of truth to the whole "Z = morphing" 
> misconception.  By modeling your entire filter as individual poles and 
> zeros, you could modify each pole or zero individually, letting you 
> dynamically alter a filter into a completely different kind of filter. 
> The filter chip in the E-mu products has one hardware interpolator, 
> allowing it to go from one set of filter coefficients to another in 
> hardware..  This also means that it only has one hardware degree of 
> freedom (hence the only dynamic parameter being Fc).  To do other 
> realtime manipulations, the software has to load a whole other set of 
> coefficients, then use the hardware to switch to them smoothly.  This 
> takes a lot of CPU, and can sound choppy if not updated frequently 
> enough.  It is, however, really really cool when it sounds right.

--- In xl7@yahoogroups.com, Aaron Eppolito <synthesis77@...> wrote:
>
> Z-plane doesn't actually have anything to do with "morphing" or
interpolation.  

So, when the manual says this.. 

"Changing a single parameter, the Morph, changes many complex filter
parameters simultaneously. Following along the Morph axis you can see
that the \ufb01lter response smoothly interpolates between the two \ufb01lters.
This is the essence of the Z-plane @257;lter. Through the use of
interpolation, many complex parameters are condensed into one
manageable entity."

... it is simply a bold faced lie? Especially the 'condensed into one
manageable entity' phrase, considering, on my command station at
least, they seem to have condensed all this functionality into
absolutely nothing at all.


The manual section tries VERY hard to suggest that there are two
filters, that can be interpolated between using a Morph parameter. It
even cites a lot of (entirely irrelevant) mumbo jumbo about how the
human vocal tract morphs between vowels. Was this function dropped at
the last minute, or did Emu deliberatly intend to mislead its customers?

So, if you had managed to get past my first sentence, you'd have noticed the rest of my post which explained that.

From a mathematical sense, Z-plane = frequency domain instead of time domain.  It's as simple as that.  What E-mu *does* with those filters that are defined in the frequency domain is enable the morph between filter coefficient sets.

Your XL-7 (and any other H-chip based E-mu product) does do this morphing.  For most of the filters, Fc is accomplished by loading two sets of filter parameters, one corresponding to a low-pass filter with the Fc at a very low frequency and one with a similar filter with the Fc at a very high frequency.  If you used this approach and *didn't* use a frequency domain parametrized filter, fading between the two would simply turn down one peak and bring up the other.  Since the modulation actually affects all the parameters in the frequency domain, fading from one set to the other actually *moves the filter peak*.

Get it?

For the other filter types, you'll notice that Fc isn't necessarily a strict interpretation of cutoff frequency.  Filters other than the standard highpass/lowpass/bandpass work in somewhat mysterious ways in response to this parameter.  Since you're into reading the manual, check out page 202-204 where they explain what the parameters do for each filter.  An example:

AahAyEeh - 6th order - type: VOW
Vowel formant filter which sweeps from “Ah” sound, through “Ay” sound to “Ee” sound at maximum frequency setting.. Q varies the apparent size of the mouth cavity.

Still seem like a bold-faced lie?  Still seem irrelevant?

-Aaron

----- Original Message ----
From: innovine1 <innovine1@...>
To: xl7@yahoogroups..com
Sent: Tuesday, January 13, 2009 10:00:38 AM
Subject: [xl7] Re: z-plane - is it "all that"?

--- In xl7@yahoogroups.com, Aaron Eppolito <synthesis77@...> wrote:
>
> Z-plane doesn't actually have anything to do with "morphing" or
interpolation.  

So, when the manual says this.. 

"Changing a single parameter, the Morph, changes many complex filter
parameters simultaneously. Following along the Morph axis you can see
that the filter response smoothly interpolates between the two filters.
This is the essence of the Z-plane filter. Through the use of
interpolation, many complex parameters are condensed into one
manageable entity."

... it is simply a bold faced lie? Especially the 'condensed into one
manageable entity' phrase, considering, on my command station at
least, they seem to have condensed all this functionality into
absolutely nothing at all.


The manual section tries VERY hard to suggest that there are two
filters, that can be interpolated between using a Morph parameter. It
even cites a lot of (entirely irrelevant) mumbo jumbo about how the
human vocal tract morphs between vowels. Was this function dropped at
the last minute, or did Emu deliberatly intend to mislead its customers?

In fact, the command station is not just that.
I really appreciate to work with several synthesizers with different characteristics each one.  That brings me synergy.
At least, the command station filter's are very different from the  average.  Not just LPF, HPF and the other two, and a filter , a resonance,. and a depth envelope  knobs.  
Same sh*t on a lot of synths.  Add a little bit of resonance, and then move the filter know to get a swept effect.   Overused....
IMHO the most relevant characteristics of the Command S are the PATCH CORDS. 
Managing the filters, with experimentation on the Patch Cord sections, rocks... 
I love to create a patch, with 4 different layers, with four different filters each ones.  That's the point! 
the possibility of delaying the start of a layer, on a tempo sync mode, is a good idea. 
Is like created an scalated, or "stairs" sound, that sound like a pseudo LFO whe you keep pressed the note, and then, the 4 four layers appears after a 1/16, 1/8, and 1/4 steps due to the delay you specified.
And then, create a patch cord, that adds a little random on the filter, different for each one of the layers.   You can activate that random, if you leave that patch cord on zero, and assignate a MIDI KNOB to activate it. 
Something interesting, is that each filter reacts in a different way to a knob assigned to the filter cuttoff.  Some filters barely react to it, the..... flanger???.. Others, like tue "classic"  react dramatically with a little bit of resonance (Q???) on it.  
I want to mention that some instruments samples, seems to be not compatible with certain filters.   I have thrown away my headphones several times for the terrible ultrahigh resonance feedback sound that results with some filters when browsing them.  That seems unevitable. 
There are so much posibilities to create BIZARRE and NEW sounds, that you cannot perform on other synths (in the same way there are sounds from other synths that you cannot create with the EMU) that grow in a square ratio, that I still having fun programming sounds, after 4 years of using two command stations. 
And I still finding things that I didn't used before.  That's cool. 
Anyway, I find useless to keep on spoiling a good product, on a minor issue.  I think that the command station has another inconvenients, like any other product, that are more relevant than its z-plane filters. 
Forget about the nomenclature or jergue gaven to this programmation  section, and let's find what you can do with them that sounds cool.
A LOT of very creative and competitive sounds, can result from this instrument. 
And I'm not just impressed because I use it a lot.  I have a VIRUS TI, a Waldorf Q,  an  Evolver, JP8080, and other synths, classics like the JUNO 106.  
I mean, I'm a hardware freak.    And I like some features of the EMU synths that you progrrammed, and I believe that you, implemented good ideas on them
And I'm grateful, that you dedicate part of your time, to help the users of your inventions, with no relevant reward from you, than the feedback obtained from them...
AMEN. 

--- On Tue, 1/13/09, Aaron Eppolito <synthesis77@...> wrote:

From: Aaron Eppolito <synthesis77@...>
Subject: Re: [xl7] Re: z-plane - is it "all that"?
To: xl7@yahoogroups.com
Date: Tuesday, January 13, 2009, 2:07 PM






So, if you had managed to get past my first sentence, you'd have noticed the rest of my post which explained that.

From a mathematical sense, Z-plane = frequency domain instead of time domain. It's as simple as that. What E-mu *does* with those filters that are defined in the frequency domain is enable the morph between filter coefficient sets.

Your XL-7 (and any other H-chip based E-mu product) does do this morphing. For most of the filters, Fc is accomplished by loading two sets of filter parameters, one corresponding to a low-pass filter with the Fc at a very low frequency and one with a similar filter with the Fc at a very high frequency. If you used this approach and *didn't* use a frequency domain parametrized filter, fading between the two would simply turn down one peak and bring up the other. Since the modulation actually affects all the parameters in the frequency domain, fading from one set to the other actually *moves the filter peak*.

Get it?

For the other filter types, you'll notice that Fc isn't necessarily a strict interpretation of cutoff frequency. Filters other than the standard highpass/lowpass/ bandpass work in somewhat mysterious ways in response to this parameter. Since you're into reading the manual, check out page 202-204 where they explain what the parameters do for each filter. An example:

AahAyEeh - 6th order - type: VOW
Vowel formant filter which sweeps from “Ah” sound, through “Ay” sound to “Ee” sound at maximum frequency setting.. Q varies the apparent size of the mouth cavity.

Still seem like a bold-faced lie? Still seem irrelevant?

-Aaron

----- Original Message ----
From: innovine1 <innovine1@yahoo. com>
To: xl7@yahoogroups. .com
Sent: Tuesday, January 13, 2009 10:00:38 AM
Subject: [xl7] Re: z-plane - is it "all that"?

--- In xl7@yahoogroups. com, Aaron Eppolito <synthesis77@ ...> wrote:
>
> Z-plane doesn't actually have anything to do with "morphing" or
interpolation. 

So, when the manual says this.. 

"Changing a single parameter, the Morph, changes many complex filter
parameters simultaneously. Following along the Morph axis you can see
that the filter response smoothly interpolates between the two filters.
This is the essence of the Z-plane filter. Through the use of
interpolation, many complex parameters are condensed into one
manageable entity."

... it is simply a bold faced lie? Especially the 'condensed into one
manageable entity' phrase, considering, on my command station at
least, they seem to have condensed all this functionality into
absolutely nothing at all.

The manual section tries VERY hard to suggest that there are two
filters, that can be interpolated between using a Morph parameter. It
even cites a lot of (entirely irrelevant) mumbo jumbo about how the
human vocal tract morphs between vowels. Was this function dropped at
the last minute, or did Emu deliberatly intend to mislead its customers?

 














      

[Non-text portions of this message have been removed]