Hi,

so finally I get to the point where I dare share an abstraction.

As the topic implies, [temperament] does basically the same thing as [mtof], only that it will let you switch between 137 different tunings. I also gave it a gui that will show you what tuning you are in and what the deviations from equal temperament (in cent) are for each pitch class. (Unfortunately, this makes it take up quite some space on the canvas - but I suppose you could just turn off the graph-on-parent. Maybe I'll find a way to control that conveniently via a tgl button?)
The temperaments.txt file needs to be in the same folder as the abstraction.

This is the page where I found the cent deviations.
http://www.instrument-tuner.com/temperaments.html

It must be said that this can never be "truly" historically accurate, since midi notes work with enharmonic equivalence (i.e., e-flat is exactly the same pitch as d-sharp), and historical tunings often or always don't (I have to read up on the details myself...). So, you would probably need seventeen different pitch classes per octave - and midi note numbers provide us with only twelve.

If you look at the tables on the above website, you might notice that the deviation for A is always zero - that is because one needs some kind of reference tone to build the temperament on. However, this needn't necessarily be A; so the third inlet in my abstraction lets you set that too (integer midi note nrs). At this point, though, this reference tone must always be an equal-tempered one - another historical inaccuracy. I will probably rebuild the mechanism a little so this inlet will accept Hz input instead, thus allowing you to set the reference pitch to, e.g., 415 Hz.

But I suppose these are rather marginal issues, since who would want to use pd as a "period instrument" anyway? So, I hope this will be useful for somebody's creative messing around, even if it is not "historically sound".

Have fun, if possible; and if you do, I'd be happy to know.
Cheers!

EDIT: Just noticed that I wrote VIII/2013 into the abstraction. I meant July, of course. Big deal.
Plus, the above webpage actually states that "the values are relative to the keynote C", not A as I assumed from the figures. I will have to do some thinking so as to be able to say whether this makes any sense.

http://www.pdpatchrepo.info/hurleur/temperament.zip

Hi,

this might be a dumb question, but I can't get msgfile to write a text file. Whenever I send it a message [write whatever.txt(, it will only give me this message:

msgfile : failed to open whatever.txt

and it does not create the file.

I tried the same thing with [textfile] and [qlist], they do create the files, and if I add some content, then do a debug printout, these contents show up; however, if open an external text editor, they don't; and most notably, if I close the patch containing the qlist object, the data is lost.

I suppose there is a way to actually save a textfile (so I can have pd write something into it and e.g. look the file up with an external text editor) , but I am unable to find it. Could someone please help me out?

Your help is, as usual, much appreciated.

Hi again,

I'm looking for a way to have pd change tuning/temperament (as in equal temperament, zarlino, kirnberger, meantone etc...) of its midi notes.

On the homepage of a computer program called scala (which I don't want to install if I can avoid it, because it requires Xcode in turn, which is a 2GB installation I'll never need for anything else) pd is listed among the "supported tuning dump formats".

"[Scala] Exports tuning data to a variety of synthesizers with an internal tuning table. You can save tuning data in a MIDI file or send it directly to your instrument or sound card. Supported tuning dump formats include among others:"
... a lot of stuff and pd among them.

(http://www.huygens-fokker.org/scala/#features)

So, does this mean that maybe somewhere in my pd installation there is a txt file or something which contains a tuning table? Possibly even for non-standard tunings? Because that would save me a lot of work.

And if there isn't, does anybody know of a way to do this apart from writing that tuning table from scratch?

Thanks, your help / attention is much appreciated!

Hi everyone, I've been getting error messages in the console when accessing the help browser. Apart from the message itself, there doesn't seem to be any problem. I just don't understand what the message is about:

WARNING: duplicate 'helmholtz~' library found!
'/Users/t1/Library/Pd/helmholtz~' is active
'/Users/t1/Library/Pd/helmholtz~' is duplicate
WARNING: duplicate 'honk' library found!
'/Users/t1/Library/Pd/honk' is active
'/Users/t1/Library/Pd/honk' is duplicate

and so on...

So, I'm on a mac and I've installed my externals and libraries where this says I should:
http://puredata.info/docs/faq/how-do-i-install-externals-and-help-files

And I wouldn't be so puzzled if the two paths followed by "is active" and "is duplicate" weren't actually one and the same.

Google came up with this:
http://lists.puredata.info/pipermail/pd-list/2012-02/094871.html;

Hans-Christoph Steiner's response to that email, respectively, the part I'm talking about, was
"I'm no longer supporting setting the path and startup libs via the
preferences. Anyone else is welcome to take that on. "

... and I don't understand the connection, since, as far as I remember, I didn't fiddle around with the preferences.

Is this problem worthy of anyone's attention at all, or should I just ignore these messages altogether?

Hello again everyone,

I'm currently in the process of reordering/cleaning a bigger patch, and one thing that keeps annoying me is that whenever I paste an object (as in cut-and-paste, for example to move things into a subpatch that were previously on the main patch), it will paste to a seemingly random location on the target canvas, so I often have to scroll around about half or three thirds my screen width to see the object and move it were it belongs.
I've also not yet found a pattern as to where it will paste (sometimes to the far left, most often the lower right, and sometimes even where I want it). Why won't it just paste on cursor location? This is SO annoying. Is there anything I can do to change this behaviour?

Help will be very much appreciated!

Hi,

I've just noticed something odd (at least to me) about the threshold~ object: namely, if I select 1 and 0 as its trigger/rest values respectively, it won't output anything. Of course, if I enter 0.9 as trigger and 0.1 as rest, it does work. Why does it behave like that?

Aren't 0 and 1 permitted? Or am I doing something wrong (look at attached patch)?

Thanks!

http://www.pdpatchrepo.info/hurleur/threshold~_1_and_0.pd

Hi,

I'd need something similiar to the [gate] object which sends data received in one inlet to an outlet specified by a number received in its second inlet - only that it should work for audio signals.
I suspect such an object does already exist, but could someone please kindly tell me what it's called?

And in the unlikely case this hasn't been built already, I've attached my personal version of it. However, since I've only been messing around with pd for half a year or so, it might not be the most intelligent, elegant or efficent way to achieve this.

Especially I wonder if there might be a way to get rid of the clicks when changing outlet; and if the [t f f] object I use is necessary in this place.

Thanks for your help!

http://www.pdpatchrepo.info/hurleur/Gate_for_signals?.pd

Hi everyone.

Two days ago, I installed jack on my mac, and since that, pd crashes regularly, stops reacting or just doesn't input/output anything: I guess I must have done something wrong in the setup.

So, if someone could tell me how to best configure jack and pd to work together (or point me to a link where that is discussed), I would be very grateful.

by the way, I also updated pd-extended to the current version just before installing jack, so it might even be a pd-alone-problem. But I can't tell, because I'm rather new to these digital audio issues.

The reason I installed jack in the first place was that I wanted to play around with my old electric guitar, using wavetables for distortion; it turned out that latency without jack was quite abominable; I then installed jack, and indeed latency decreased audibly. But for some reason, while yesterday everything worked, today it won't. (I admit I did play around with the audio settings of both programs, mainly sample rate and block size, to see how it would sound. Must have messed up something in the process.)

Another thing I noticed is that when I started up pd the first few times (yesterday, when it still worked) without having activated the jackserver, I got this message saying "you should start jackpilot now" or something like that, and it would allow me to start the jackserver directly from the pd window, more or less.
But if I start pd now without jackpilot activated, this message doesn't show up. This wouldn't be a big problem in itself, since I can activate it manually, but maybe it is a clue to somebody as to why pd now does not work anymore.

Well, I hope I could make myself understood, even if my post may sound a bit confused... anyway, thanks in advance for any help.

Hey sebfumaster,

cool that you actually took the time to go and improve my patching - I appreciate it!
However, I have to postpone taking a closer look at it, because I should actually be doing something else now - deadlines ...

Sorry, but I did add contents before sending the write message.

I've attached a patch that produces an error message for me:
msgfile : failed to open file.txt

I hope you can tell what I did wrong, but it's not that I expect msgfile to write an empty textfile.

Thanks so far...

http://www.pdpatchrepo.info/hurleur/msgfileproblem.pd

Look here: http://puredata.hurleur.com/sujet-5664-state-saving-abstractions

thanks!

I think you want

zexy/demultiplex~

(... I posted a very similiar question here not too long ago. I don't find it particularly intuitive that the same thing is called gate with control data and demux with signals - anyone care to explain why actually?)

This is how I'd try it: (see attached patch)

(I just discovered the list-abs library a few days ago, and it turns out to be rather useful).

http://www.pdpatchrepo.info/hurleur/column2list.pd

You need a counter to the second inlet, see attachment. The index needs to be set before the random number goes into the tabwrite, hence the [t b b].

Edit: Or you could probably even use a second urn object instead of the counter.

http://www.pdpatchrepo.info/hurleur/random_list.pd

Thanks, that helped. I really appreciate your taking the time to answer even questions as close to pointlessness as this one...

Hey, thanks for the hint. By the way, for me it was enough to delete just the .msg file for my language.
I don't really understand why anyone would want to translate the pd interface in the first place, since all the objects do stay in english and most of the online resources are in english as well... For me this was just a source of additional confusion.

Thanks, now I understand.

But now I've still got a problem with what I want to do: I think I need something like [sel] for signals, so it bangs strictly when it _reaches_ one or zero. Because in the patch above, the output of the vline~ object should always stay between one and zero. How could I do that?

EDIT: Oh my, I just figured I could probably do that with [==~] and [edge~].

If I understand you correctly, there is an easy answer to this: [sel 1].

Thanks for the answers!

(Looking into the matrix objects just now, I guess it'll take a while until I understand...)