[PD] [PD-announce] bassemu~: transistor bass emulation

Christian Klippel ck at mamalala.de
Tue Jun 20 03:35:26 CEST 2006


hello padawan12,

Am Dienstag, 20. Juni 2006 13:24 schrieb padawan12:
> Lovely. Really nice sounds.
>
> One or two humble suggestions to make it a truly authentic acid box...
>

you are welcome, any input is welcome for that matter ;-)

> A zero value note should behave as a rest (as well as -1?) Makes sequencing
> easier in most scenarios.
>

you are right. dont know why i made it -1 to stop originally....

> Glide (the secret sauce for proper acid lines)
>
> i)  switchable on and off by a control signal while keeping its range
> fixed, it should be truncated by a rest not dive to zero. (see iii)
>
> iii) instant zero - reset to zero slide "instantly" (without clicking -
> about 30ms)

ok, thats a good idea. altough it could be done with external objects, i still 
prefer to have that builtin, for performance- and ease-of-use reasons....

> ii) premptively * - by previous note, a glid note in position 14 slides all
> the way to note at step 15 then stops. A "lagged" glide doesn't always work
> right because if it's not constant octaves/time it doesn't work in
> different places on the keyboard.
>
> v)  an option for constant time slide, it changes its speed
> depending on the note distance to always be constant 1 step period.
> * possible for one step but see below
>
> * it's actually impossible for a sequenced instrument because you need to
> precompute the slide by looking ahead in the pattern. It's a subtlety that
> many people don't get when aiming for an authentic 303 sound. This way you
> can slide at constant pitch/time whether there are one, two or more tied
> notes or rests in between.
>

there could be a helper-message in the form of [next note note note ...( that 
is sent instead of a single note, if such a gliding is wanted. that way the 
engine can know of the next few notes ahead of time .....

in any case, the whole gliding stuff in the bassemu is not really the way i 
want to have it finally. but it worked somewhat back then, and i just forgot 
to work more on it....

>
> iv) exponential as well as linear - original RC circuits are asymptotic so
> up glides always fall a tiny bit flat while down glides remain a teeny bit
> sharp of the target note.
>

that should be easy to implement....

>
> vi) with filter cutoff tracking the glide too
>

speaking of that, i think i will add an option to route the glide to the 
different sections (note, cutoff, reso ...) as wanted.... so, one could have 
a glide on the note only, or note & cutoff, or whatever ....

> Accent (choose from the below list which are all hallmarks of good
> acid instruments, not necessarily all TB303)
>
> i) shortens decay time by about somewhere between 0-20%
> ii) boosts env mod and resonance by 10% and cutoff by 20%
> iii) boosts amplitude the same
>

ok ....

> Listen to anything by "Hardfloor" or "a guy called Gerald" to really get
> it.
>
> Can anything be done to stabilise the filter? It went mad a couple of times
> when I was tweaking the filter env parameter.
>

in a short form: yes, just limit the slider to not go below 0 or above 1 ..... 
in fact, not all parameters are thought to become negative, also, some are 
meant to be +1 maximum. i just did not limit the input's because in some 
settings you can have nice effects. for example, look at 
http://home.mamalala.de/bass.jpg ....
set the waveform & limitng type first, then cutoff and reso .... then the 
decay to -0.07 .... the sound will first start to modulate, then fade into a 
noise, but it will come back after a short while .... after that, it will 
produce a kind of bass-soundscape that will never repeat in pattern ... it 
modulates itself ....

oh, and i found a small bug for the reset message ... it doesnt recalc the 
coefficents after resetting, keeping the old ones ....

but as you can see, there may be uses for extreme settings, when you know what 
you do .... for save live-tweaking, just limit the sliders, or split a 
parameter up into serveral sliders, avoiding "bad" ranges ...

thank you for all the hints, will include that some day (have some other stuff 
to do right now, urgently ...) when i release the other objects as well...


greets,

chris





More information about the Pd-list mailing list