I have seen this error ("can't load abstraction within itself") before, but it should show up in Pd as well as MobMuPlat. Is your patch named the same name as the abstraction? Have you got instances of the abstraction inside the abstraction itself? or in a subpatch [pd subpatchname]?
I have been able to load abstractions within MobMuPlat but the abstractions themselves are entirely Vanilla (and, as you say, in the same directory).
If the abstractions are utilizing some externals then these will not create but the patch will still load.
Is it possible that you have multiple copies of Pure Data on the Pi?
Perhaps you installed it and then moved it but did not delete the previous installation?
If the "address" it refers to is the port number, than you could keep trying with different port numbers, it could be that you are using a port number already in use by your system.
I realize that I am not sure how to view this on linux but for sure there will be a way if you research it...
however, if the problem is that for whatever reason you are opening 2 instances of Pd that have the same port number saved or loadbanged into the patch than maybe a test could be to delete the port number from the patch and make it so that you enter this manually after the patch opens
then try it and if it works but still shows two instances of pd running then the problem is that you have multiple Pds being opened at the same time somehow.
In that case, perhaps deleting any Pure Data installations on the Pi and re-installing would solve it.
Isn't the -minpower flag meant to set a lower db threshold by which the analysis will take place?
In other words unless the minpower value is reached the object won't analyse the incoming event.
Also 50 is the default so in fact you don't need to flag it if you want to set the minpower to 50. This could also be why you are not noticing any effect.
If npeak is working are you not getting the result you want?
(to see the results:
[route 0 1 2 3 4...]
[unpack 0 0 0 ...]
number boxes for 14 outlets
If you are getting too many/too fast values you can slow down and spread out the analysis by increasing the -ntps and -hop values.
You can reset flags by sending the object. a new argument value in a message box.
This is detailed in "pd setting parameters" in the sigmund~ help file, and in fact the -minpower flag is the one they use as an example.
This is a pretty common way of updating some arguments to some objects in Pd, super useful.
Hppe this helps!
hey there @polyphuckin
I realize this thread is three years old but it seems like I am working along similar lines although in the other direction. Did you complete this project and is there a way to see it or listen to it? What solution did you find for parsing the csv data?
I am working the other direction from csv data to FFT to audio...
Currently the IEMlib CSV object isn't working on my machine but I wondered what you came up with in the end, perhaps a solution like @LiamG suggested above is a good one, but I am sort of trying many options (haha, every option I can fine in other words).
I know it is in fact an infinitely complex thing, and I use software like the REW package (amazingly also open-source) to do it "right" but I also think that there could be a quick a dirty way to pull the main room resonances out with a simple test, and even make this an automatic part of another patch, such that if I have a piece with a broadband and quite loud set of continuous synth tones (haha, pretty common in my music) but for which I am also using a live mic, I could have my patch bring down the resonant room nodes by itself so as to prevent trouble.
Like an automatic feedback canceller (or creater along these lines, like what I think is being used in some of Agostino Di Scipio's music (the electronic side of which is often written in Pd I think)
have you all heard any of those?
thanks for all these suggestions, I will work on it and see what comes out! maybe it will be something worth sharing
@jameslo once your have your IR balloon recordings, have a look at the built-in reverb device in Reaper if you don't already know it, it has a lot of intereresting functions for customising how the IR is implemented. Pretty great.
I would like to build a quick and dirty analysis tool, such that I could sweep a sine wave or perhaps give a quick impulse response to a space and then, using a microphone analyse what the room does to this signal (thus revealing harmonic resonances of the space).
I think maybe the new (and amazing) AudioLab library could be useful for this, and there is the older Pd Spectral Toolkit that allows for working with complex numbers and is amazing, and of course William Brents TimbreID,
but since I don't really need this to do everything that programs like this do: https://www.roomeqwizard.com/
I was trying to think of the simplest way to do this, before I commit to a six month project
I am feeling lazy, it happens.
I wondered if anyone had any suggestions or new of some patches that already did this?
I thought the simplest way might be to sweep a sine wave and simultaneously record it back in, then play this recording into sigmund~
...but here is where it gets confusing because since room sounds are not musical instruments, the most resonant partials are not....well, the whole point is that one doesn't know where they are, so I wouldn't know which partials to tell sigmund~ to analyse ahead of time.
so, it might be better if the recorded sound was played into another spectral analysis tool (but which?) and then the output of this could be filtered to find x number of strong partials (or all partials about a certain strength).
Wouldn't that be awesome?
I am dreaming but maybe somebody has already started on this?
ok! for anyone else who is on a newer version of Windows and ever needs to trash the preferences file, I have found the solution.
It is similar to what is given on the Pd site ([https://puredata.info/docs/faq/pdsettings])
but not the same
you do use REGEDIT to get at it, which is the built in Windows registry tool (careful!)
because the registry address (at least for me on Windows 7) is not HKEY_LOCAL_MACHINE -> SOFTWARE -> Pd
as it says there, but rather UP in HKEY_CURRENT_USER
don't be fooled, there is an entry in HKEY_LOCAL_MACHINE -> SOFTWARE -> Pd, but this only contains default registry information
instead go to HKEY_CURRENT_USER -> Pure-Data ->
and there you find two sets of entries
clicking on HKEY_CURRENT_USER -> Pure-Data -> org,puredata.pd.pdgui
only gives registry entries pertinent to Deken
however clicking on the first, parent folder called Pure-Data
gives an itemized list of each entry you have made to your preferences
so you can just scroll until you find whatever the offensive preference is from the list and select delete with that one highlighted in REGEDIT
anyway, hope this helps someone!
I was editing startup flags trying (once again) to get the Cream library to run on Windows 7 and accidentally added a quotation mark.
Because of this I am getting the dastardly error message posted below and I can not longer enter the startup area of the preferences to delete or edit this!
I have reinstalled Pd but that did not help (as startup flags are carried forward from the old installation)
I have also tried the "Forget" option in preferences but that charmingly reports that there is "nothing to forget"
I think what I need to do is trash the preferences file, but I cannot figure out where it is. On mac it is called Plist. Is anyone familiar with where I find this on Windows 7? Following the recommendation on the Pd website, I have looked in the registry using REGEDIT or whatever it is called but only see reference to Pd-Extended (I also have extended on this machine but that should not effect this I do not think). I think that is probably old info up there.!
I am attaching a sad screenshot so you can see the error: