so i now tried atarikai's
as agroschim already corrected the vcf is a bandpass not a lowpass
also i think the samplerate reduction is done before the waveshaping...
but i'm not so sure if this does even make a difference
-
Emulate oto biscuit in PD
pd redefining mathematics |expr fact(0)|==0
-
@slur said:
also i think the samplerate reduction is done before the waveshaping...
but i'm not so sure if this does even make a differenceIt does not make any difference.
-
damn you are right....
also i suggest adding a oversampling of 2 or 4 to get more out of the lowpass
but yet again not sure if this is a good choice or could even be contra productivep.s. pd aaaaaaarrrrrrghh
pd redefining mathematics |expr fact(0)|==0
-
Here's a more descent version. I did add some waveshaping effects. The rectifyers work okay, i guess. The rectify down is somewhat of a cheap trick, but sounds fine. Bat fuzz doesn't work as expected, i would be glad if someone could take a look at it. Swap was easy (
).I took slur's advice and made the frequency button logarithmic. I also added an on/off button to the clock (aka change samplerate effetc), as it may be undesired sometimes
For testing and understanding the "Krümel M." i will upload a simple kick with a oscilloscope too.
-
Here's a more descent version. I did add some waveshaping effects. The rectifyers work okay, i guess. The rectify down is somewhat of a cheap trick, but sounds fine. Bat fuzz doesn't work as expected, i would be glad if someone could take a look at it. Swap was easy (
).I took slur's advice and made the frequency button logarithmic. I also added an on/off button to the clock (aka change samplerate effetc), as it may be undesired sometimes
For testing and understanding the "Krümel M." i will upload a simple kick with a oscilloscope too.
-
Here's a more descent version. I did add some waveshaping effects. The rectifyers work okay, i guess. The rectify down is somewhat of a cheap trick, but sounds fine. Bat fuzz doesn't work as expected, i would be glad if someone could take a look at it. Swap was easy (
).I took slur's advice and made the frequency button logarithmic. I also added an on/off button to the clock (aka change samplerate effetc), as it may be undesired sometimes
For testing and understanding the "Krümel M." i will upload a simple kick with a oscilloscope too.
-
hitting rec/down seems to kill the sound for good, for me.
overall I think the sounds achieved in the previous version are nicer. Couldn't really get any musical distortion out of this one
Dual 1.8 IBM G5: Mac OSX 10.4.11 -- Asus eeePC 701: Pure:Dyne / eeeXubuntu GNU/Linux -- myspace.com/thearifd
-
The filter knob now is logarithmic, try a higher cutoff. I made this mistake myself.
Edit: I made a stupid mistake in the bat fuzz subpatch.
-
i made some changes to krümel i thought where necessary or something else
- replaced the os dependent spigot~
- simplified some stuff
- complified some other stuff
- made the gui smaller and much more ugly
- added the bit reduction from my floatcrusher
- made the filter section (hopefully!) samplerate independent
- added experimental oversample capabilities for you to test if its a useful mod or not (feedback please!)
i didn't test much and especially not every change so keep your eyes open for bugs
@agroshim
what's the purpose of rect down? as far as i understand it it should be sqrt(pow(sig~,2)) which is exactly the same as rect up
the way it works now is something like clip 0 1 but with a light twist
the lower part of the waveform which is cut away is not at -1 after the sig*2-1 its at 0pd redefining mathematics |expr fact(0)|==0
-
@slur said:
[...]
the way it works now is something like clip 0 1 [...]That is exactly what it should do... omg i was so tired when i did this.
I will test your improved Krümel soon.
-
-bug-fixes!!
-simplification of bat fuzz
-filter changes
-other stuff
-rect down with clip
-(new) rect c to simulate the behaviour of the old rect downthe msi version is the most recent
in ms i tried to make the biquad based filters more samplerate independent....
well..... its better than before.........i stands for "iemlib filters"
the vcf are the only ones i could find which are samplerate independent so i used them
does anybody know a better solution?
which one sounds better?i know there are some things which could be done better.... like switching the inactive filters off
well... you do it!pd redefining mathematics |expr fact(0)|==0
-
-bug-fixes!!
-simplification of bat fuzz
-filter changes
-other stuff
-rect down with clip
-(new) rect c to simulate the behaviour of the old rect downthe msi version is the most recent
in ms i tried to make the biquad based filters more samplerate independent....
well..... its better than before.........i stands for "iemlib filters"
the vcf are the only ones i could find which are samplerate independent so i used them
does anybody know a better solution?
which one sounds better?i know there are some things which could be done better.... like switching the inactive filters off
well... you do it!pd redefining mathematics |expr fact(0)|==0
-
no feedback?
well next version
with level compensation for batfuzz and mute
not perfect but bearablepd redefining mathematics |expr fact(0)|==0
-
Oh, i am so sorry. I totally forgot this as i was strugeling with other things. Looks descent sofar, will have to give it a try with some samples and report.
-
New Version of mine:
* added slurs bat fuzz
* added slurs oversampling
* changed the maximum samplerate to 44000Hz (don't forget to oversample before using rates over 22000Hz!)
*clock scale is now logarithmic
* simplified the rect down
* kept my interface
* kept the [spigot~]Have fun.
-
You should really get rid of the [spigot~]. It doesn't work on Windows, and there's nothing it can do that [*~] can't.
-
* changed the maximum samplerate to 44000Hz (don't forget to oversample before using rates over 22000Hz!)
yeah right wanted to do this too... now its done in a good way i thinkhas anybody an idea how to make my version use less CPU?
i guess its the iemlib filters but i'm too lazy to think about a good (!) way to switch them on/off
can't put them into sub patches and switch them on the hard way. i guess this would lead to clicks and whatnotagroschim:
you have kept a crapload of other crappy crap... like the samplerate dependant biquadpd redefining mathematics |expr fact(0)|==0
-
You could [switch~] them off without clicks by crossfading them. [switch~] the new one on and fade-in while fading the old one out and [switch~] it off at the end of the fade. You'd have to get rid of the [mux~], but since only one will be on anyway it shouldn't be a problem.
-
@slur said:
you have kept a crapload of other crappy crap... like the samplerate dependant biquad
I admit it, i've been a bit lazy with this one. I'm sorry, i had different things in my mind the last weeks. I promise to make a major update this weekend, with a new bitwise distortion feature.
-
uhuuhh this sounds nice....
and always remember the filters....
i think it would have been easier to just put your gui back into my stuff... or whatever
if you want to use biquad for whatever reason look at my ms version... its not perfect but at least a starting point
there even exists a msia a=array where the whole first part is done by an array to reduce cpu load but it only brought 20%? for a lot side effects so i kept msimaelstorm:
yeah i'm just to lazy... but i think i now have a good idea to make it simple and effective
maybe i'll implement itpd redefining mathematics |expr fact(0)|==0
