One oscillator feedback
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?
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.
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
> 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
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
Kasper T. Toeplitz
noise, composition, bass, computer
> There seems to be some DX7 orcs for csound~ – but i always said to
> myself to look into it later… ;-)
That orc proposes something like below. If you find a manual of a DX7 you can now model one for yourself.
----------begin_max5_patcher---------- 1026.3oc0YsriaBCEccxWAxKqRG4WXftqeGUipHDmYnJ.Q.ocdnlu8B1lDZZ xXCJ1E1PT3w0m6wGe8gKuubAXcwK7Jf2W79l2hEuubwBwoZOwB0+W.xheIYW bk31.721DmbDrRdo7CYEGp2wqEWDpN6935jmSye56k7jZYvwnvGfq7XLR6OT +tidOpdlzMhvWr9Ge1mzK7o4cQG2dteubY6gUFh1mhSyuAXwe.XwQs3yGJ9A CaOhnvdnUFm5W2ykOAnJ8o73cfUdfz7lnbsrBe2xpb9uZhXW3p4uHvMnIA3o +jeza65qmxnamxTnbhQLKE.6xasI7oaXaQdcU5ahKivO.uBA.uJAf5Ef73LQ ..esLsI32Opohmu417BTKufnREQv+PL5yaZjKx6CYq4kCWoyXh7hH065m1aU 2Mh70w4OMj4dJpCZ6K4MSF0w0oE48.BgDbFHP0gSCfEIts6JZh9fINJqMO8B 8Mk3ZFm3wQcmpazvAoYMXsUwd5NuAe5SHmwmY7YCszEdzoveeH4jhrLdqx4h kkE64kw0EkdrAuvjDIRMDK5FUrLfYIZDkz.TuwvkpxaTFKqtX6fKsSnTgHUv WDpdo5fTmTcbHx+7nOAnvOcbr6MJWKQvO3e+2aj5qUJRMc8L1Ez39miqJJON VwnptI0F9LnLyjjSFtL40jc7wPkxJf93yVpu6bYfVtj0CDtjL2Gmy2M7sMPh Z4RUHhJquSYW+EQng5bsDAOGMME2bmoNjzOBQ5V0ht5H51.ECkHgENSr0QnB 4LILx595Hzw3qiDI.HEBmEF6FdeBBEqknAhsbYnQXqi3q8cMjiQHcNaqCiCk qxirguNhtMQwAxgO.NeM1gCC5sZ2NN6HAl8luFshdRasqSPppcZEucjPyjkS G1bjl6Tsfj5yrm4NRjVxTBBVv7vbGV9hZDYWL05tCpK8kciRUgcxXuCCCkpA eaauCClt8eCJquPBrtQMb3GZTa1a4hPwxs+Pi1xE9+Xq8Gm4oPlLMCrRSw.y IWPp9HnVJYIWPf4mcFkDQUgwJ1YvfYmuDkkYY4W63KAAlrFLjUGQxWBWmACF PmSAwSA1kle42fWL1sm+uSgphCkIcDa2GU167fugWUmlK7wz+lZV84gNcSOm tYCOuellktYeQiIAEH7d7pD58GSPmgoVqt5wD1oXp86gqESseX.Gho198pGS QCCSjlhmM6ofj8zQ0AqK9mruaTjX2G0JJ6mItUD19Qv0StT2hIrohP2U.4BM +MvTfawD0DLwb6bGa5wSlHwIAts1enIXJxozDwjotK.t0wTvDjmnlpmbHlHl fI2VJfXTYbmBIbnga.6vYNCPDxsaqX.hbqqTrk4nl+76k+AfKKtCK -----------end_max5_patcher-----------
> 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
----------begin_max5_patcher---------- > 1026.3oc0YsriaBCEccxWAxKqRG4WXftqeGUipHDmYnJ.Q.ocdnlu8B1lDZZ > xXCJ1E1PT3w0m6wGe8gKuubAXcwK7Jf2W79l2hEuubwBwoZOwB0+W.xheIYW > bk31.721DmbDrRdo7CYEGp2wqEWDpN6935jmSye56k7jZYvwnvGfq7XLR6OT > +tidOpdlzMhvWr9Ge1mzK7o4cQG2dteubY6gUFh1mhSyuAXwe.XwQs3yGJ9A > CaOhnvdnUFm5W2ykOAnJ8o73cfUdfz7lnbsrBe2xpb9uZhXW3p4uHvMnIA3o > +jeza65qmxnamxTnbhQLKE.6xasI7oaXaQdcU5ahKivO.uBA.uJAf5Ef73LQ > ..esLsI32Opohmu417BTKufnREQv+PL5yaZjKx6CYq4kCWoyXh7hH065m1aU > 2Mh70w4OMj4dJpCZ6K4MSF0w0oE48.BgDbFHP0gSCfEIts6JZh9fINJqMO8B > 8Mk3ZFm3wQcmpazvAoYMXsUwd5NuAe5SHmwmY7YCszEdzoveeH4jhrLdqx4h > kkE64kw0EkdrAuvjDIRMDK5FUrLfYIZDkz.TuwvkpxaTFKqtX6fKsSnTgHUv > WDpdo5fTmTcbHx+7nOAnvOcbr6MJWKQvO3e+2aj5qUJRMc8L1Ez39miqJJON > VwnptI0F9LnLyjjSFtL40jc7wPkxJf93yVpu6bYfVtj0CDtjL2Gmy2M7sMPh > Z4RUHhJquSYW+EQng5bsDAOGMME2bmoNjzOBQ5V0ht5H51.ECkHgENSr0QnB > 4LILx595Hzw3qiDI.HEBmEF6FdeBBEqknAhsbYnQXqi3q8cMjiQHcNaqCiCk > qxirguNhtMQwAxgO.NeM1gCC5sZ2NN6HAl8luFshdRasqSPppcZEucjPyjkS > G1bjl6Tsfj5yrm4NRjVxTBBVv7vbGV9hZDYWL05tCpK8kciRUgcxXuCCCkpA > eaauCClt8eCJquPBrtQMb3GZTa1a4hPwxs+Pi1xE9+Xq8Gm4oPlLMCrRSw.y > IWPp9HnVJYIWPf4mcFkDQUgwJ1YvfYmuDkkYY4W63KAAlrFLjUGQxWBWmACF > PmSAwSA1kle42fWL1sm+uSgphCkIcDa2GU167fugWUmlK7wz+lZV84gNcSOm > tYCOuellktYeQiIAEH7d7pD58GSPmgoVqt5wD1oXp86gqESseX.Gho198pGS > QCCSjlhmM6ofj8zQ0AqK9mruaTjX2G0JJ6mItUD19Qv0StT2hIrohP2U.4BM > +MvTfawD0DLwb6bGa5wSlHwIAts1enIXJxozDwjotK.t0wTvDjmnlpmbHlHl > fI2VJfXTYbmBIbnga.6vYNCPDxsaqX.hbqqTrk4nl+76k+AfKKtCK > -----------end_max5_patcher-----------
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
> 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
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