@Yar You have a couple of [gemhead]'s
You could try fixing the rendering order...... see the [gemhead] help file..... as the mask needs to be rendered after the original for every frame (I think.... it depends on the effect you are looking for).
David.

@mezko Guard points can be added using this handy patch (made by @lacuna if I remember correctly).
Because [tabread4~] interpolates the samples a copy of the last samples in the array needs to be added to the start..... Array Guard.zip

Here I have maybe softened the [switch~] .... that might help.......

EDIT... Error in previous upload........... per.zip
David.

@mezko I am missing too many externals to test your patch.... but I will guess that the resize message will halt audio momentarily, and the [switch~] object, if it is being triggered by [adr~] will be switching audio on and off.
As you have called your array $0-noise I assume that it is noise, and so the lack of guard points for [tabread4~] will not be the cause of your clicks.

But resize and [switch~] could both be causing clicks..... [switch~] definitely if it is being triggered.
"resize", in your patch, is not causing clicks as you are sending the message to noise-$0 and the array is named $0-noise.

So [switch~]
You are probably trying to soft fade as notes change and avoid clicks that way, but [switch~] is abrupt.
It would be better to soft fade using [line~] controlling the [*~ ] as below to fade out and in on each side of the switch.
David.

@stefano_zorzanello Yes...... some here for audio, and Gem already has a recursion folder (13.recursion) with some ideas... https://forum.pdpatchrepo.info/topic/10897/l-system-iterations
David.

@Balwyn I wondered about [file cwd]
Does that work? and does Pd still find objects once another directory is set if you then use [declare]?
I have used [file glob] to find directories and search files in the past, but never changed the working directory.
David.

@jamcultur As @jameslo says...... something like this....... this.pd
Although you probably want each array in an abstraction with the file name as an abstraction argument or you will have problems with multiple arrays with the same name.... so ignore that....
So see the zip below which uses abstractions.
Dollars in objects are translated as the patch is opened so then you could replace [pack s s] with [pack s $1] and remove the [loadbang]
[EDIT... you cannot as the $1 needs to be a symbol so [loadbang] is still required.... I have changed that in the zip]

The directory selection could be in your master patch and a message sent to the abstractions to change directory.
Note that the $2 in the message is the same as the $1 in the objects in this case.....

Try doing something like this to organise your files and the setting of the tables...... this.zip

You will need to put all the abstractions in your Master patch but you only need to do that once. Future edits of the abstraction where necessary will be easy.
David.

@jamcultur
Replace "disp" in the read message with a $1 and instead of banging the message send the directory name accel vel or disp into it,,
David.

@raynovich You are not doing anything wrong.
Some external libraries are individual binaries for each external, so putting the path in preferences allows Pd to find them by name as they are created in the patch.

But zexy and some others are a single binary containing all the externals. Only the help files are separate and can be found by name.
That means that the whole library needs to be loaded into memory as Pd starts..... and that is what happens using the -lib flag at startup.
The library can be loaded instead just for your patch using [declare].
That means that all the externals are available for Pd in exactly the same manner as all standard objects are available because those are part of the Pd binary itself.
David.

@Yar You will find this useful....... https://curiousart.org/digital_proj/pd_eBook.pdf

I have not tried with Gem. Tuio was an alternative in the past.
David.

@cfry You might find some help here........ https://github.com/agraef/pd-lua/issues/75
There was a commit at the bottom of that page.
David.

@tildebrow Not getting such errors in windows with pd extended or vanilla..... but I think it is not working as you expect either.

It is easier to do this though to hide the window......... hmm.pd
To remove edit focus before closing also send "editmode 0"

So
[;
pd-$2-subpatch editmode 0, vis $1(

If you want it in editmode as it opens..
[;
pd-$2-subpatch editmode $1, vis $1(

David.

@mezko I am fairly certain that in Pd fft objects will be used in small sample filters so an understanding is useful.
It is not my subject so I could be talking rubbish....
However..... https://www.katjaas.nl/minimumphase/minimumphase.html
David.

@mezko You should be able to remove the header using the -raw flag for [soundfiler]
This......

should help ypu, although I am unsure because I have not watched it (yet).
You can probably put (a) header back by saving the result as a wav file.
David.

@mezko
https://forum.pdpatchrepo.info/topic/5927/building-filters-from-difference-equations-using-pd-s-raw-filters

You will see that @katjav chips into that thread.
You will find more information from her here....
https://www.katjaas.nl/home/home.html .... with great explanations of the Fourrer Transform among other things.... and her site will help understanding of the maths involved.

Also.... fft.zip
David.

@hansr Pretty sure a wrong order of operation...... as [expr sin....etc] looks for the value of w$0 but it is not set by [unpack] until after [expr ($f3...etc] operates.
So a one time only divide by zero as the patch is first run..
Swapping the first and second outlets of [unpack f f f] should fix it...
David.

@Orangery ChatGPT will not help...it will not have devoted much resource to Pd..
As you mentioned compiling @dreamer probably thought you were trying to compile CC from pure data using Heavy.
Pd does not normally need to be compiled.
You could send a screen shot of the errors you are getting, which would help us understand maybe..?
But the pd patch that is producing those errors would be more useful for seeing how to help you.... it doesn't matter that it is not working.... it will show us how far you have got and save us repeating the work that you have already done.
You can upload it to the forum using the "up arrow" symbol above your post as you are typing.
David..

@Penguino Do you get a result using this.zip test patch...? (unzip first)
David.

@Penguino Android supports midi since 6.0 and Midi2 since 13.0.
But I think devices need to offer support for a USB OnTheGo connection.
The Hexler TouchOsc page ... https://hexler.net/touchosc/manual/setup-android-usb-midi suggests that there are no naming conventions or fixed port numbers.
This page might help understanding of any hurdles... https://source.android.com/docs/core/audio/midi

I can find no reference to a successful PdDroidParty implementation but I think it could be a question of finding a midi device that claims to work with Android.... class compliant.
Some might work but need a splitter OTG cable to provide extra power for the device.

The TouchOSC app is cheap and IMHO a useful investment...... so might be worth trying.
It can easily communicate with Pd on a computer over a network using OSC messages.
David.

@jamcultur My best guess....
[savestate] in the main patch (as that works correctly) and global (not $0) messages to/from the [toggle] in the 2nd level abstraction.
Might need a small delay to restore the [toggle] correctly on recreation as the creation order might perturb the restore.
David.