damu

I've been trying to make a universal 303 slide algo in gen~

You know the classic sound. I'm trying to make a gen version which listens to a stream of incoming notes into in1 then applies a slide to the tail of any note which is accompanied by a trigger to input 2.

I have used Graham Wakefield's gen~ implementation of ramp smooth to create the glide, which works great.

However, the logic has proven difficult as it involves making a clock which allows the triggered note and the following note to enter the ramp smooth algo.

It's nearly working, but still skips some of the glides. Could anyone help me finish it?

comment

sig~

live.dial

click~

button

sah~

edge~

gen~

change

noise~

switch

history

delta

counter

bool

codebox

scale~

abs~

phasor~

toggle

ezdac~

metro

However, the logic has proven difficult as it involves making a clock which allows the triggered note and the following note to enter the ramp smooth algo. 

I've been trying to make a universal 303 slide algo in gen~
You know the classic sound. 

303-style-slide-in-gen~

Very briefly, just a guess (since I'm on the move), but it might be due to the stuff in the Max patch where you use edge~ into click~ for the glide trigger. Anything that goes from signal domain to message domain (edge~ here) can lose sample accuracy, and you may want sample accuracy for your algorithm here since you are working with single-sample triggers (click~).

E.g. your random selection between ramp & not ramp can be driven by the metro->click~ into a gen~, and you can multiply this click~ by a combo of [noise] -> [> 0] to get random triggers synced to the clock. (And you can replace "0" with another value or a param to change the glide probability).

Another approach that doesn't even need clock input: In the gen~ that does the ramping, get a trigger whenever the note changes and sample the [noise] -> [> 0] there. Feed that, or the [in 2 glide], into a [latch] (which works a bit like sah~) and trigger it with the [change] -> [abs] pair you already have. That will give a gate that is either on or off for the duration of any note, which you can feed into the [switch] driving the rampsmooth codebox.

Very briefly, just a guess (since I'm on the move), but it might be due to the stuff in the Max patch where you use edge~ into click~ for the glide trigger. Anything that goes from signal domain to message domain (edge~ here) can lose sample accuracy, and you may want sample accuracy for your algorithm here since you are working with single-sample triggers (click~). 

E.g. your random selection between ramp & not ramp can be driven by the metro->click~ into a gen~, and you can multiply this click~ by a combo of [noise] -> [> 0] to get random triggers synced to the clock. (And you can replace "0" with another value or a param to change the glide probability). 

Another approach that doesn't even need clock input: In the gen~ that does the ramping, get a trigger whenever the note changes and sample the [noise] -> [> 0] there.  Feed that, or the [in 2 glide], into a [latch] (which works a bit like sah~) and trigger it with the [change] -> [abs] pair you already have. That will give a gate that is either on or off for the duration of any note, which you can feed into the [switch] driving the rampsmooth codebox. 

Thank you very much for this reply Graham. I'm looking forward to trying it :)

303 style slide in Gen~