i found out that noise~ seems to be seeded only with something like $0+noise_count and not with a better seed like the current time
which leads to the result that if you use noise to control things like in the syn-rain from diy2 the output is always the same and completely reproduceable (well i didn't check if you can even get the same md5)

so i did (a short and dirty) noiser~
i think i overdid it a little bit but at least it seams to give good results

to see what i'm talking about open up the noiser-testing maybe 3 times

for your usual noise~ needs (above 1hz) noise~ should still be enough

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

is it possible?
i need to send messages with $0 and similar... but
|$0( gets replaced to: 0 (as it should)
|$0( gets replaced to: $0 (like i want it, BUT if saved becomes |$\$0( )
|$\$0( gets replaced to: $\$0 (clearly not what i want!)

and it seems that it is even impossible to generate a message with adddollar and similar to do what i want to

really annoying that i use more time to work around pd than to make nice things

i have attached a example but don't give me a f $0 solution... i need the $0 in a (auto generated) receive!

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

how is it possible?
i guess its annoying to read 4 samples each output sample if you don't need exact timing but a lot delaylines

i know z~ but its not suitable for feedback

my collection on more or less shareable stuff
i think msia3 and perk-eso3string may be the most interesting ones... and maybe keykontrol if someone gets what it does

its all the most recent versions so maybe some things in abstractions i put on here may have changed

ana-h10* is the h10 example as an abstraction

filewriter~ is for easy saving as .wav it has a build in clock and adds time and date to the filename specified by argument 1

keykontrol can make your keyboard into a theremin
the key-split's are unfinished and not standardised. they split the keyboard into different layouts

mono-freqshift does frequency shifting
the F version tries to prevent reflections at 0hz or the nyquist

mono-gate is a complex gate with 3 inlets~
see help file

msia3 is my latest version of the oto-krümel patch. array based so it only works from -2 to 2 and it did give only ~ 10-20% performance improvement

oscil~ is a simple ugly stereo oscilloscope

perk-eso3string is a plucked string synth with 3 strings. basically a karplus strong, but extremely cpu hungry! If you can optimise it you are welcome.
Also i can't figure out if there is some usefulness in the feedback between different strings (Xfeed)

syn-wobbel do you remember mr oizo?

--- unfinished ones
arrayclip~ a complex wave shaping abstraction

beatshuffle a abstraction to confuse an audio stream

comments, suggestions, improvements (performance!), money, pron, death threats and viagra are welcome

the different files are redundant

http://www.pdpatchrepo.info/hurleur/slurs_crap_1.zip

my collection on more or less shareable stuff
i think msia3 and perk-eso3string may be the most interesting ones... and maybe keykontrol if someone gets what it does

its all the most recent versions so maybe some things in abstractions i put on here may have changed

ana-h10* is the h10 example as an abstraction

filewriter~ is for easy saving as .wav it has a build in clock and adds time and date to the filename specified by argument 1

keykontrol can make your keyboard into a theremin
the key-split's are unfinished and not standardised. they split the keyboard into different layouts

mono-freqshift does frequency shifting
the F version tries to prevent reflections at 0hz or the nyquist

mono-gate is a complex gate with 3 inlets~
see help file

msia3 is my latest version of the oto-krümel patch. array based so it only works from -2 to 2 and it did give only ~ 10-20% performance improvement

oscil~ is a simple ugly stereo oscilloscope

perk-eso3string is a plucked string synth with 3 strings. basically a karplus strong, but extremely cpu hungry! If you can optimise it you are welcome.
Also i can't figure out if there is some usefulness in the feedback between different strings (Xfeed)

syn-wobbel do you remember mr oizo?

--- unfinished ones
arrayclip~ a complex wave shaping abstraction

beatshuffle a abstraction to confuse an audio stream

comments, suggestions, improvements (performance!), money, pron, death threats and viagra are welcome

the different files are redundant

http://www.pdpatchrepo.info/hurleur/slurs_crap_1.zip

which wave editor do you use?
i am currently searching for one with as much options as possible, multiple tracks for mixing and it should be free-ware or similar

i found freqshift~ in extended but it does some strange aliasing and did even produced extreme amplitudes... which i currently cant reproduce
so i did a first experiment myself
it is samplerate dependant and has other little bugs but i think its somehow better than freqshift~

any suggestions ideas or whatever?

http://www.pdpatchrepo.info/hurleur/fft-freqshift.pd

i want to split a signal into high mid and low parts to process them differently and later mix them back together
also i want the 2 frequencys where the split happens to be changeable.
perfect would be a solution with which one could produce any number of different paths, but 3 should suffice.

how can this be done?

thanks

the abstraction made from this thread

i did what i could do
maybe the audio pros do know how to enhance the sound more
filtering maybe?
we could make a community project out of it

have fun

even oversampling just doubles real block size

i want to build a feedback loop imitating in analogue circuits so i need low block sizes. 1 would be best. 64 would have resonant frequencies in the audible range.

should i report this as a bug?

http://www.pdpatchrepo.info/hurleur/testing-switch.pd

little cluttery thingy to make writing sound files easier

argument defines path/file-name
-start starts writing a new file with date and time attached
-auto starts and then stops after the given time

you can put your default output path into the msg within the patch

example_filename_2010.1.30_6-4-59

please post your improvements and if you like it

http://www.pdpatchrepo.info/hurleur/filewriter~.pd

if there already exists a good one please let me know!

with the version from E01, to get the whole range up to 20kHz, you have to use an array with around 900 points. but this will clutter half of your screen and if you shrink it, some spikes wont appear correctly making it useless.

i think what a spectrum view needs is
1. ability to shrink it down to a width of 150pixel without making it useless
(2.) logarithmic scaling

how to do 1 i cant figure out
2 may be possible with some quiet computational intensive hacks (converting signal to msgs and working on them)

also does anybody know how to get the block-size? (like samplerate~)

i noticed some abstractions seem to use another way than the preferences-> draw on parent function
lets say it shows some numberboxes on the parent but in the patch they are distributed all over the patch
how is this done?

its a wrap for keyname

key -> output

space -> 0
1 .. 0 -> 1 .. 10
q -> 11
y .. - -> 31 .. 40

F1 .. F9 -> -1 .. -9

! .. _ -> 40 .. 80

had to make some ugly workarounds bacause of pd

currently only german qwertz keyboard supported but its easy to implement different layouts

dont make money with it or ill bite you!
if you implement different layouts share! or ill peel your skin off... with a rusty spoon

making it better & suggestions welcome

hf

http://www.pdpatchrepo.info/hurleur/keykontrol-0.1.pd

i'm currently using pdext 40.3.rc3 but all errors were present in whatever 41.4
version i had before

-select wtf?
sel s wont recognize |symbol 1( even if you send it exactly this msg 2 times. i want
to use it to detect key 1 with keyname
interestingly a select f recognizes a |symbol q ( but not a float 1 combined with a
pressed 1

-unpack s s s
wont convert a |1 2 3 ( to symbol why? numbers should be able to be symbols too!
or what about at least a unpack a a a ?

-selecting keyname , and ;
keyname seems to put out \, and \; but you cant use \ in pd. the main window says keycode 92 dropped... duh :/

-dsp loop
if you make a patch / pd subwindow with a
inlet~ -> delwrite~
delread~ -> outlet~
pd wont see that there is no loop if you feed the output back to the input
should be easy to implement just one boolean after reading and building a graph for the patch. and i know you do some analysis becouse how would you detect the loop whithout

-slow redraw
why is there a redraw at almost every mouse event, while moving objects?
just use the last one or of you want to make it float around, make a list and use the position of event at 25% 50% 75% and 100% or whatever

-better looking connections?
i bet the developers already think about doing this but how about adding dots to connections instead of making workarounds with +~ and whatever to make a patch look better
and maybe even bspline or bezier-curves?
even better! autoalign /-organize like in http://www.graphviz.org/ ? i know its SLOW!

-not realy a pd problem
on my keyboard _ is uppercase for - but with capslock the output is -
but this seems not to be a direct pd error because windows does this mistake... why i wonder

-ET phones home?
why does pd.exe and wish84.exe regulary try to phone home? and what is wish84 anyway? mal-/adware?

enough nagging for today
thanks for working on pd @all devs
i realy like the graphical changes in 40.3ext-rc3 looks much better and everything
and thanks to all who answer me beforehand

p.s. i could make a patch with some of the bugs if you want

maybe take a look at timeroll

hm over here the help file for sys_gui dosnt work correctly... if i give it colour values in hex like discribed in there they won't work...
but thanks for asking a question i wanted an answer to too

im not good with fft too... i just know that i don'T know enough... currently thinking about building something phasevocoder based... the link from katja was http://www.dspdimension.com/admin/pitch-shifting-using-the-ft/ uses real frequencys... i think there was a link to a description how this is done... or at least its hidden in the source (im really put off by only thinking about porting a cpp to pd)

check out snapshot~ , unsig~ and pack~

didn'T look into it jet.. but did you account for bucket smearing? and maybe calculate the true frequency? (i hate fft):
but well a blocksize of 1s should be quiet accurate.. but you know the blocksize is always 2^x so i guess its really 32768 samples