Mike81

Hi guys. Was just reading about one osc feedback Fm. I understand the principle but for the life of me can't seem to be able to implement it in max. Does anyone have any examples?

one-oscillator-feedback

You could for instance connect to the phase input of a cycle~ object the output of the same object. Note that this works only if you include a send~/receive~ pair or tapin~/tapout~ in the network (it's forbidden otherwise to connect outs to ins). You should also include level control in order to control the amount of feedback. Things become really more interesting when you add a filter, for instance a reson~. I think the result can be called somewhat chaotic.

> Hi guys. Was just reading about one osc feedback Fm. I understand the

> principle but for the life of me can't seem to be able to implement

> it in max. Does anyone have any examples?

The only way to do that with plain max objects, is to place it into a 

Or you need to code it as external. To get the classic DX7 patches you'd 

need 3 variants of this, as the feedback is going over 1, 2 or 3 

Michael Andrews schrieb:
> Hi guys. Was just reading about one osc feedback Fm. I understand the
> principle but for the life of me can't seem to be able to implement
> it in max. Does anyone have any examples?

The only way to do that with plain max objects, is to place it into a 
poly~ with the vector size set to 1.
Or you need to code it as external. To get the classic DX7 patches you'd 
need 3 variants of this, as the feedback is going over 1, 2 or 3 
oscilators...

-- 
Stefan Tiedje------------x-------
--_____-----------|--------------
--(_|_ ----|-----|-----()-------
-- _|_)----|-----()--------------
----------()--------www.ccmix.com


> Stefan, i made some cool things feeding the phase of an oscillator with 

> the output of itself or another one, but i'd love to know the details of 

> the dx7 feedback you're talking about. Any more infos ?

http://www.adp-gmbh.ch/csound/dx7/algorithms.html)

There is always one feedback path. mostly for a single oscillator.

> Last, i came accross this "poly~ with vector size to 1" topic a few 

> times these last years, but i always said to myself to look into it 

> later. Now i'd love to know what the hell it is and what you do with that.

Whenever you create a feedback path in MSP, you need at least one signal 

If you'd use a cycle~ and feed the output of the oscillator back to the 

frequency inlet, you have to do that as well...

Usually that feedback oscillator was used to create noise. If you crank 

up the feedback level, you get a chaotic/noisy result.

It might be fine to leave a bigger delay (signal vector size), but I am 

There seems to be some DX7 orcs for csound~ - but i always said to 

f.e schrieb:
> Stefan, i made some cool things feeding the phase of an oscillator with 
> the output of itself or another one, but i'd love to know the details of 
> the dx7 feedback you're talking about. Any more infos ?

If you look at the DX7 algorithms 
(http://www.adp-gmbh.ch/csound/dx7/algorithms.html)
There is always one feedback path. mostly for a single oscillator.

> Last, i came accross this "poly~ with vector size to 1" topic a few 
> times these last years, but i always said to myself to look into it 
> later. Now i'd love to know what the hell it is and what you do with that.

Whenever you create a feedback path in MSP, you need at least one signal 
vector of delay, or MSP will shut up...
If you'd use a cycle~ and feed the output of the oscillator back to the 
frequency inlet, you have to do that as well...
Usually that feedback oscillator was used to create noise. If you crank 
up the feedback level, you get a chaotic/noisy result.

It might be fine to leave a bigger delay (signal vector size), but I am 
sure it will sound different...

There seems to be some DX7 orcs for csound~ - but i always said to 
myself to look into it later... ;-)

>>Last, i came accross this "poly~ with vector size to 1" topic a few 

>>times these last years, but i always said to myself to look into it 

>>later. Now i'd love to know what the hell it is and what you do 

>Whenever you create a feedback path in MSP, you need at least one 

>signal vector of delay, or MSP will shut up...

don't you achieve the same result with a tapin~/tapout~ pair with no 

>
>>Last, i came accross this "poly~ with vector size to 1" topic a few 
>>times these last years, but i always said to myself to look into it 
>>later. Now i'd love to know what the hell it is and what you do 
>>with that.
>
>Whenever you create a feedback path in MSP, you need at least one 
>signal vector of delay, or MSP will shut up...

don't you achieve the same result with a tapin~/tapout~ pair with no 
arguments???

kasper
-- 
  Kasper T. Toeplitz
noise, composition, bass, computer
http://www.sleazeArt.com

> There seems to be some DX7 orcs for csound~ - but i always said to 

That orc proposes something like below. If you find a manual of a DX7 you can now model one for yourself.

> There seems to be some DX7 orcs for csound~ - but i always said to 
> myself to look into it later... ;-)

> From what i guess, MSP objets are all using the vector size set in the 

> dsp status. So without any arguments, the tapin~/tapout~ pair should use 

> this vs as a minimum delay, correct me if i'm wrong.

That is correct, but in a poly~ you can have a different vs, and a 

tapin~/tapout~ pair will use the vs inside of the poly~. vs 1 usually is 

too heavy on the cpu to run a whole patch on it...

> Best crunchy sound i got with a cycle~ is with a vs 64. Vs 1 is dull and 

The sound should be compared to a real DX 7 to find out whow that is 

different. For just creating noise with feedback any delay might be 

f.e schrieb:
>  From what i guess, MSP objets are all using the vector size set in the 
> dsp status. So without any arguments, the tapin~/tapout~ pair should use 
> this vs as a minimum delay, correct me if i'm wrong.

That is correct, but in a poly~ you can have a different vs, and a 
tapin~/tapout~ pair will use the vs inside of the poly~. vs 1 usually is 
too heavy on the cpu to run a whole patch on it...

> Best crunchy sound i got with a cycle~ is with a vs 64. Vs 1 is dull and 
> above 64, it's too noisy.

The sound should be compared to a real DX 7 to find out whow that is 
different. For just creating noise with feedback any delay might be 
fine... (Or use noise~ instead... ;-)

Was I confused when I read somewhere that max 5 text will be readable by humans???

Quote: jvkr wrote on Wed, 07 May 2008 06:14

--------------------------------------------------->

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

Quote: jvkr wrote on Wed, 07 May 2008 06:14
--------------------------------------------------->
>
----------------------------------------------------


Fm noise is distinctly different from noise~, as you can have an envelope over the feedback amount and wind up with the sound turning from a chaotic noise to a clean tone over time. It can be a very powerful effect.

Fm noise is distinctly different from noise~, as you can have an envelope over the feedback amount and wind up with the sound turning from a chaotic noise to a clean tone over time.  It can be a very powerful effect.


On May 8, 2008, at 10:01 AM, Nicholas C. Raftis III wrote:

> Was I confused when I read somewhere that max 5 text will be 

This is a compressed interchange format, intended for posting on the 

mailing list. The native Max 5 format is based on json, and is pretty 

readable. Max 5.0.2 add the ability to take one of these compressed 

blocks, and expand it into a text window before instantiating it as a 

On May 8, 2008, at 10:01 AM, Nicholas C. Raftis III wrote:
> Was I confused when I read somewhere that max 5 text will be  
> readable by humans???

This is a compressed interchange format, intended for posting on the  
mailing list. The native Max 5 format is based on json, and is pretty  
readable. Max 5.0.2 add the ability to take one of these compressed  
blocks, and expand it into a text window before instantiating it as a  
Max patcher.

Chris Muir
cbm@well.com	
http://www.xfade.com

Quote: Chris Muir wrote on Thu, 08 May 2008 19:30

> On May 8, 2008, at 10:01 AM, Nicholas C. Raftis III wrote:

> > Was I confused when I read somewhere that max 5 text will be 

You mean you can't? Where did you go to school?? :P

Quote: Chris Muir wrote on Thu, 08 May 2008 19:30
----------------------------------------------------
>
> On May 8, 2008, at 10:01 AM, Nicholas C. Raftis III wrote:
> > Was I confused when I read somewhere that max 5 text will be  
> > readable by humans???

You mean you can't? Where did you go to school?? :P


One oscillator feedback