Emulate oto biscuit in PD
You know I was just thinking that the mute and invert sonically sound the same, the invert just sounds a bit louder...
Wait, where/how would you make the bit = 128? when all the bits added together = 255? Sorry, I'm still learning the math aspect of all this.
-Kai
@mod: But each bit can only have a value of 0 or 1. It's not meant to mute a sample; it's just to alter the value of that sample.
yeah, i know, but in pd you have 32 bits, so you can cheat.
if you make the mute default to 128 instead of 0, then you don't have to account so much for the DC offset you will get if mute is 0 or 255
Can you post an example of what you mean? I'm not really following. You get DC offset no matter what mute value you use.
I found that DC offset can be fixed easily by taking the 8-bit mute value, divide it by two, and adding it to the signal. Modifications attached.
yeah sorry, i was really twisting my words around and not being clear. what i meant, is that if you make the mute value equal to a signal of zero amplitude, then you won't get DC offset.
so, if a message value of 0 would be equal to a signal amplitude of -1, and a message value of 255 would be equal to a signal amplitude of 1, then a message value of 128 would be equal to a signal value of 0.
patch attached, should be clear.
also, i did a volume adjustment that seems handy.
only thing is, i ran some sounds through it, and it sounds bloody awful. guess that's why the analogue filter is needed!
don't think that totally worked.
revised patch seems to do it ok. the signal needed to be salted with noise to stop it going all crazy on low volume parts. i sort of understand why, because i have read about this sort of thing before (denormalization, i think it's called) but it's the first time i have run across it first hand.
anyone know how to make a nice analogue lowpass filter in pd?
mmb: you don't need the int() the & already takes care of that
mod: in my opinion mmb's version of doing this is much better and smaller
control can be easily changed to radiobuttons too
i would try lp?_bess because they attenuate the signal much better than the default lop and have a slow roll-off (is it called that way?) so that higher frequencys still get through but are much weaker
but i didn't even test it
pd redefining mathematics |expr fact(0)|==0
@Maelstorm said:
Can you post an example of what you mean? I'm not really following. You get DC offset no matter what mute value you use.
I found that DC offset can be fixed easily by taking the 8-bit mute value, divide it by two, and adding it to the signal. Modifications attached.
Added a [biquad~]-Filter to your bitwise distortion and tested it with a basic kick, sounds awesome, and with a simple string patch, sounds cruel. Good job anyway.
want to share your biquad?
i got the idea for a floatcrusher... incorporated mmbs basher into it and it got way to complicated
it has some obvious bugs but currently i can't get my mind to work on it any more
the floatcrushing effect is mostly visible at 1bit
http://www.pdpatchrepo.info/hurleur/floatcrusher_advancing.pd
pd redefining mathematics |expr fact(0)|==0
thx
what i found, in a short search, about analogue filters:
https://ccrma.stanford.edu/~jos/pasp/Wave_Digital_Filter_Models.html
http://www.dsprelated.com/dspbooks/pasp/Wave_Digital_Filters.html
pd redefining mathematics |expr fact(0)|==0
@slur said:
mod: in my opinion mmb's version of doing this is much better and smaller
control can be easily changed to radiobuttons too
yeah, after testing, i certainly agree. it's really neat.
Any ideas on how to dynamically change the samplerate?
mod: Ah, I think I see what you mean now! You're thinking more in terms of dealing with the decimal equivalent of what each bit represents. Interesting, though, I'm not getting denormal issues. That noise gets pretty noticeable as you start messing with the bits...
agroschim: You can use [samphold~] to simulate it. The frequency of a [phasor~] going into the right inlet will be the "new" sample rate, though it will still be quantized to the sample rate of the patch
@Maelstorm said:
Interesting, though, I'm not getting denormal issues.
Ha! That's because I'm not on an Intel machine right now. 
@Maelstorm said:
agroschim: You can use [samphold~] to simulate it. The frequency of a [phasor~] going into the right inlet will be the "new" sample rate, though it will still be quantized to the sample rate of the patch
Will this lead to aliasing effects? As far as i know, you would have to change the sample rate of the patch to achieve this.
yes but aliasing is what distortion is mostly about 
but if you want to you can filter all frequencys out which would reflect on the new samplerate
but sh goes only up to 1/2 the samplerate with a normal phasor
because it only changes if the value drops.. to do this it has to increase for one sample so 2 samples are needed
at least that how i understand it
pd redefining mathematics |expr fact(0)|==0
@slur said:
but if you want to you can filter all frequencys out which would reflect on the new samplerate
A highpass filter with a cutoff frequency at samplerate/2 and a strong falloff should do. Am i correct? I will try this later.
this won't do anything
look at H10 to understand why
i'm talking about 1/2 the frequency of the phasor because the phasor simulates a new samplerate
pd redefining mathematics |expr fact(0)|==0
I'm sorry, i confused highpass with lowpass.
Edit: Updated version of my kick with destortion attached.
Oops! Looks like something went wrong!
