Hi,
New pd user here,
I've used Max MSP for some time now.

I was wandering if one sample delay processing is possible in pd as in gen~ in MaxMSP ?

also, is there anything similar to the gen~ cycle object with 0 phase attribute?

creating something like this in pd is possible?
a feedback oscillator with input to the phase of the oscillator?

First of all, you're using a control delay and not a signal one. You're missing the tilde. Use [delwrite~] instead of [delay]. You'll need a [delread~] to read the signal
Also not that you can't make a feedback connection in the signal domain because it creates a DSP loop. Max probably works under the hood to make this function properly. But if you think, the execution order goes like this: the object that sends audio to its outlet does its computations before the object that receives the audio signal. If you make a feedback connection, which object is supposed to make its computations first?
To achieve this you have to send the output of [cos~] to a [tabsend~] (don't disconnect it from [*~ 0.4]), and connect a [tabreceive~] to [delwrite~]. Also, if you want feedback in your delay, you'll need to do this in the same way, so use another pair of [tabsend~]/[tabreceive~] to connect the output of [delwrite~] back to its input.

@alexandros said:

But if you think, the execution order goes like this: the object that sends audio to its outlet does its computations before the object that receives the audio signal. If you make a feedback connection, which object is supposed to make its computations first?
To achieve this you have to send the output of [cos~] to a [tabsend~] (don't disconnect it from [*~ 0.4]), and connect a [tabreceive~] to [delwrite~]. Also, if you want feedback in your delay, you'll need to do this in the same way, so use another pair of [tabsend~]/[tabreceive~] to connect the output of [delwrite~] back to its input.

By using a [delwrite~] [delread~] or [delread4~] pair, the DSP loop is already detached, - you won't need [tabsend~] [tabreceive~] additionaly. You can use either or to break the loop.

Make sure that the whole chain feeding (and including) [delwrite~] is written in the .pd file before [delread4~],
(Check .pd file in a text editor. In Pd you can select > cut > paste parts of the patch, to change the order) -
then you will be able to go as low as min 1 sample delay.

You can measure latency with this latency-tester2.pd
https://forum.pdpatchrepo.info/topic/13814/catch-would-like-to-set-the-name-programmatically/20

A thread on this topic:
https://forum.pdpatchrepo.info/topic/6185/feedback-fm-algorithm

@lacuna is correct about [delwrite~]/[delread~] and the DSP loop being dettached, didn't occur to me while writing the post.

delread~ is not a sample delay it is a tapping buffer.

a tapping delay does not delay "the samples", it delays their content.

plus its minimum delay time is the vectorsize.

i think in pd you would replace [delay~ 1] with [fexpr~]... something like x = x1, no idea about the correct synthax.

You can perfectly use the delwrite~ delread~ , write~for 1 sample feedback
Test case :operator feedback phase modulating itself will turn a sine into a sawish waveform and then ultimately noise .
-Create subpatcher , route phasor~ into +~ into cos~
Inser block~1
Insert delwrite~ , insert delread~
Route cos~ to delwrite~
Route output of delread~ into miltiply *~( this is our feedback control)
Route this back to the adder before the cos~ .

Voila , phase modulation feedback , Sine will turn into saw into noise , the shawish like shape would not have been possible with longer delay times.
So delread~,write~ are perfectly capable of doing 1 unit sample feedback.

Here's a PD patch ( plugdata version ) which uses the tab~send , tab~receive~ and ofcourse ' block~1'
I have noticed that it doesn' t make any difference when using tabread~ or tabreceive ~.
I do have noticed that while the delread~ and delwrite~ are capable of 1 unit sample feedback , the part where the foldover-collapse starts sounds better when using tabsend~ method instead of delread~-write~
This patch is euivalent to a two operator feedback loop , modulator going into carrie and carrier back into mod .

feedbacker2.pd

And here using the two techniques , one operator feedback
Left channel using del~ method , righ channel tab~method

You need the Else sine object or just replace with phasor~ +cos~
comparisons.pd

@alexandros

What am I doing wrong ?

@KMETE you have to put all objects you want to run at a 1-sample block size into a subpatch. [dac~], [s~ ], [r~ ] and the likes can't handle block sizes other than the default 64. Only [inlet~] and [outlet~] can up/downsample signals.
Also, you don't need [delwrite~]/[delread~] there, but a [array define fed 1]. Get rid of the objects on the right and the top part of your image (with the [delread~] and [delwrite~], along with their connected [tabsend~ fed1] and tabreceive~ fed]), and change the argument to the [tabreceive~ fed1] to "fed".