Reaktor

[anarchosyn]

Hey,

anybody that feels experienced with reaktor want to take a newbie (well, sorta newbie) under their wing? I know how to set up basic patches, and I know the basics of design structure (i.e. start with an empty instrument and build inside, slowly recompiling the structures into macros).

What I need to understand are the subtler elements, like (for example) when setting up a portemento control why should I run the glide knob through an event exponential module before taking it into the UP/DN Iinputs of a slew limiter (example taken from a preset patch I'm attempting to reverse engineer)? I understand the event signal would need to be converted into a audio signal for input, and this module does just that, but why change from a logarithmic signal into linear? I'd like to learn the basic, foundational math behind this reasoning. Is there a reason why the designer chose a min value (of the glide control knob) to be 80 and a max of 20 (both positive values).

[edit: btw- I am aware of the relationship between exponential voltage increases and how we perceive pitch psychoacoustically. seems like exponential values would be prefered to linear ones, or am I wrong in thinking this module outputs linear signals ?]

Or, what does the "control shape" module actually do (I know it is a "Shaper for control signals. One breakpoint. (0), (50) and (100) adjust the levels for (In) = 0, (In) = 0.5 (breakpoint) and (In) = 1" but this makes no sense to me, how do the respective breakpoint inputs actually affect the primary input for processing to the output). In the patch I'm looking at, the midi gate and midi note numbers are mixed into a monophonic signal, via the "select min" module ("The active voice with the minimum value is selected, and its value is passed to the monophonic output. A gate signal of the same instrument is needed to detect the active voices."), and the singular output is fed into all three breakpoint inputs of the control shaper module. The primary input, which is being effected by the 3 breakpoint inputs, is a mix of midi velocity on and off signals (both midi input modules are feed directly into the input of the control shaper). The output is simply the midi note numbers which were fed into the module in the first place, seemingly unaffected.

or..

why do the midi On and off velocity inputs transmit values even when the sequence is turned off? I always thought velocity on and off values were not transmitted simultaneously and only resulted from keypresses/depresses...

etc..

These are not really specific questions as examples of the types of things I need to learn and understand from a concrete mathematical and/or operational basis.

I know I'll probably do little more then embarrass myself by asking such questions since answers are far and few between on this forum, but I can't find any place else to ask (the yahoo group's list is pretty dead).

(btw - yes, I am aware I could simply take modules and reuse them without needing to understand them, but that’s no fun. ;)

[anarchosyn]

Originally posted by anarchosyn
(I know it is a "Shaper for control signals. One breakpoint. (0), (50) and (100) adjust the levels for (In) = 0, (In) = 0.5 (breakpoint) and (In) = 1" but this makes no sense to me,

btw- I forgot to add that the quote above (which is from the horrid reaktor documentation) implies to me that the breakpoints .. well.. damn.. it doesn't imply a bloody thing.

Ok, we have 4 inputs:
----
Breakpoint in 100 - "Input for control of the output level at (In) = 1"

Breakpoint in 50- "Input for control of the output level at (In) = .5"

Breakpoint in 0- "Input for control of the output level at (In) = 0"

[and ]

in - "Input for the control signal to be shaped. The points (0) and (100) refer to the input values 0 and 1.
Typ. Range: [ 0 ... 1 ]"

-----------

We have only one output.

If the same patchcord is attached to all three breakpoints, meaning all three breakpoints are getting the same value, how is this changing the control signal at the input?


or am I just trying to reverse engineer a shitty use of this module..?

THIS is something I would really like some clarity on..

[c s]

actually i never used neither the 'control shaper' nor did i build gilde/portamento into my own seq yet, so at the moment i'm clueless as well, but i'll have to take a look at that later today as i also still strive to really understand reaktor completely. i never looked at too many of the ensembles in the library but when i did there were often those things i didn't understand completely, mostly exactly those small plus/invert/exp/log-stuff at the inputs of known modules. well, i'm generally quite helpless in maths i guess. which ensembles were it you were looking at ?

[CHIP TRONIC]

is anybody able to construct a good sequencer?????
does anyone like to exchange selfbuilt ensembles?

[CHIP TRONIC]

uhm at the moment i am analysing brinkmann´s sequenced ensembles.
Perhaps i will have finished some ensembles until the end of next week.

Wow

I'll give some of the questions a try......

Originally posted by anarchosyn
when setting up a portemento control why should I run the glide knob through an event exponential module before taking it into the UP/DN Iinputs of a slew limiter ?... Is there a reason why the designer chose a min value (of the glide control knob) to be 80 and a max of 20 (both positive values).

first of all: the input value for the up and down ins of the limiter is "units per second" (1/sec), so the value you attach here is inversely proportional to duration (the time it takes for a value at the In input to ramp up or down to the next, for porta: the glide time from one pitch to another) This is why minimum (few or no glide) needs to be a higher value than max (longest glide).
The exp shaper between knob and inputs has got nothing to do with exp vs log pitch control; after all, the pitch information enters the In input, the Up/Down inputs are just there to set the TIME for rising/falling slopes (i.e. the SHAPE of the slope, or transition between pitches, is not altered exponentially); it simply converts values that are convenient for a knob (with limited resolution) to usable values for the limiter's inputs.
I don't know about this particular example, but compare it to this: if you want to send pitch values to an osc's F input, you can do 2 things: give the knob values from say 100 to 10,000 Hz in 100 steps; you now have very coarse resolution in the lower frequencies (100 -> 200 Hz is one octave), and a resolution that is probably too fine in the higher. If instead you would give the knob a value from 43 to 123 and run it through a exp shaper (F), you get the same range but with a more equal distribution, at least in terms of pitch. You probably knew this already, but once you've grabbed this concept, there's no reason why you shouldn't use this trick in other circumstances than P -> F.

Or, what does the "control shape" module actually do (I know it is a "Shaper for control signals. One breakpoint. (0), (50) and (100) adjust the levels for (In) = 0, (In) = 0.5 (breakpoint) and (In) = 1" but this makes no sense to me, how do the respective breakpoint inputs actually affect the primary input for processing to the output)... The output is simply the midi note numbers which were fed into the module in the first place, seemingly unaffected.

By attaching the same constant value to the three breakpoints, a variable input is indeed changed into that constant at the output (maybe that was exactly what the designer wanted, he could have used a value module instead though).
The Reaktor manual is not horrid, I won't have that (), it's just.... euhm more of a reference guide than a tutorial; it's even very 'gründlich' I would say, very German... ,

By the way, the control shaper is a good example of a module whose secrets are easily unravelled by moving the mouse over the output cable (great feature, use it!! use oscilloscopes for audio signals!!); attach different values to the different inputs and see what happens, a logic pattern should emerge eventually
What the manual writes is correct actually, the shaper transforms values BETWEEN 0 AND 1, the breakpoints are (rather confusingly I admit) expressed in PERCENTAGES (50 means '50% of 1' = 0.5, so input values between 0 and 0.5 will be shaped into new values between breakpoint 0 and breakpoint 50)

why do the midi On and off velocity inputs transmit values even when the sequence is turned off? I always thought velocity on and off values were not transmitted simultaneously and only resulted from keypresses/depresses...

because it is what it says: a vel on (or off) module, not a on AND off module (if that where the case you would indeed get different values for key on or off); the vel on module 'doesn't know' the key is depressed so to speak, it will only react to a new keypress with different (or even the same) vel on value. The gate is a module that changes value at key off (vel on as long as the key is pressed, 0 for note off).

I know I'll probably do little more then embarrass myself by asking such questions since answers are far and few between on this forum, but I can't find any place else to ask (the yahoo group's list is pretty dead).

Questions are never embarrassing, but pleassssse don't inundate me with them because I'm always tempted to answer, and it keeps me from producing music aaaaaargh The Reaktor forum at the official NI website is probably a better place to ask for help. It's very much alive and there's enough of a mix of newbies and senior users (haha) to not feel intimidated and yet get the help you need for specific questions. Also, there's a Wizoo guide on Reaktor that comes highly recommended, I've never used it myself though.

is anybody able to construct a good sequencer?????

you can use my ensemble named "drummachine", it consist: 8channel sequencer, 2 hat synth, fm percussion, bassdrum, clap, tom, cowbell, snare synth and i can send you some sequencers

http://www.native-instruments.net/fi...chine_n1 .zip

[anarchosyn]

Thanx trip!

Let me ask you though, can you give me examples of potential uses for the control shaper? I've toyed with it a bit since I read what you've posted here, and I think I understand it a bit better, but what's the mathmatical logic behind it?

For example, if for breakpoint 100 (bp100) I put in the value of 60, for bp50 I put 30 and for bp0 I put in the value of 10, inputs over 1 will come out 60, and values at .5 will come out at 30, and values of zero come out at 10. but if I put in the value of .3 i get the output of 22. What's the math behind this (i.e. the rational)? Wait.. I guess the math is dependent upon the specific values input (since, with a bp50 of 50 and bp0 of 0 outputs 25 with a input of .25, midway between the two values. If the bp50 was 30 and bp0 was 187, an input of .25 would equal 108.5.. let me test that, ok cool - it worked).

So, I guess my only question would be why somebody would apply this?

[btw- yes, I know there is not one "right" way to apply this module, but could you give me a few examples of its use?]

Ironically, with what I know now, it seems that the module servered no real purpose in the patch I was looking at. It could have been omitted, as it was simply passing the values recieved at the BPx inputs to the output, regardless of the input values involved (since the same value was connected to all three BP inputs). I felt this might be the case from just cable hovering (yes, that is quite the nice option ;), but I wanted to make sure.

thanx again!

[CHIP TRONIC]

well in the meantime i´ve built some ordinary synths using "normal"
sequencers.
your drummachine ensemble is really good.
Yet at the momrent i regard those tables as a little difficult.
let´s wait and see! I´ll work on it!
would be cool if you sent me some sequencers!


By the way.martin brinkmann´s lvd_21 is also a very good drumsequencer

[anarchosyn]

Wait.. so, one could actually say that the name "control shaper" is quite misleading. It's actually a "control changer" and doesn't actually shape the cv wave (or floating point numbers) in any particular fashion, no (when I say "shape" i mean like an audio signal shaper, as found in the nord which turns audio signals into more exponential or otherwise)?

Originally posted by anarchosyn
Wait.. so, one could actually say that the name "control shaper" is quite misleading. It's actually a "control changer" and doesn't actually shape the cv wave (or floating point numbers) in any particular fashion, no (when I say "shape" i mean like an audio signal shaper, as found in the nord which turns audio signals into more exponential or otherwise)?

right! it doesn't apply any hidden math like an exp or log shaper, it just linearily changes/scales values between breakpoints.
Since you seem to know the nord modular, you should compare it to the Note and Velocity Scaler (last of the control modules), except that the breakpoint in Reaktor is fixed: when you move the L/R gain knobs (in the modular), the little graph shows how values below and above the breakpoint are scaled in a linear way.

Uses.... euhm let me think.... right now I can't think of anything .
'irregular' types of scaling, but each example i come up with seems to be possible without using the ctrl shaper

Check out Travellizer (in the New in Reaktor 3 folder) for an ensemble that makes use of a couple of shapers. Brinkmann uses it for some math inside the 'songlfo' macro (see his LfOz1 ensemble from the user library).

Now you got me playing around with it... I attached an lfo module (sine out) to it, to create wildly irregular lfo shapes (you have to do some additional math since the lfo outputs values between -1 and 1).

[c s]

Originally posted by Triptonizer
Now you got me playing around with it... I attached an lfo module (sine out) to it, to create wildly irregular lfo shapes

hehe i did exactly the same thing today.

how about discussing a more fascinating (to me at least) module as well, the great tables ? it's been a year since they've been introduced and i already used them here and there, but now i just started to do more complex stuff with it. i think the possibilities of these modules alone are great and possibly unlimited. writing to and reading from the table at the same time, with a changing delay etc. but i don't know what other people have been doing with it already (except for controller graphs).

c s,

I rarely use the audio tables, I use the event tables for sequencing or to create complex controller envelopes; there's something I hate about them though: if you're not carefull they get overwritten very easily. I have the habit to play around with an ensemble and save a snapshot as soon as I've found something usefull; but if you change the contents of the table file in the process these snapshots come out all wrong afterwards. So you always have to make sure you make the table file large enough to include several patterns, and then change pattern each time you make a new snapshot -- undo doesn't work on table files either.

Perhaps you could give some examples of how you use the audio tables?

[c s]

well you can save the contents of the table (in a strange file format), but yes it's a bit too uncomfortable to look for the right table file that goes with a certain snapshot.

i haven't used it for static data, except for controller graphs in live ensembles (of course it's a great thing for that also), i'm doing an effect instrument (for use with any synth instrument) at the moment. for now it's writing the input and reading it with an adjustable delay of x units. it could also be the other way round, also up to now only write x is used, i wonder what will happen if it start attaching things to write y as well. or if reading and writing points cross each other... but it already sounds quite strange. often you suddenly have 50 knobs when playing with reaktor and don't really remember which one to turn...

[c s]

hmm i think it some other things with it too during this year but i don't remember. i would have to look at the ensembles.

[anarchosyn]

Hey trip (or anybody else).. I have one more question.

This time it concerns a macro that comes stock with reaktor, the ADSR.

Within the adsr is another macro called Vel amt, which seems to scale the gate signal to act as velocity for the ADSR ENV module.

I understand how this operates, except for the inclusion of the "Order" module.. why is that there?

Removing it doesn't seem to change the values being output, and I can't understand the rational for its inclusion. If there isn't one, I'll feel better hearing it from somebody more experienced. If there is one, please let me know so my knowledge of reaktor can blossom.

[quick, crack open a copy of the macro in question, as what follows makes little sense without a visual reference ]

Seems like you can simply run two cables from the HI output of the seperator module into the two locations which run from the Order.
Since the second output of order, which is slightly delayed from the
initial, is being used solely as a trigger to pass the Value of the first output (after slight modification) through the merg, why not have them fire at the same time? Or, is the slight delay there to compensate for the time it takes for the first output to cycle through all those other modules? Couldn't they have simply split and delayed the signal without using Order for a cheaper design?

[c s]

just looked at it and don't get the specific sense either. let's just wait for tripton, i guess he'll tell us...

@ cs: help luckily Anarchosyn seems to be very capable to come up with the answers himself, he just needs confirmation

@ Anarchosyn: I don't THINK (stress think) the Order module is really, really necessary here, no. And yes it's there to make sure the value at Val is in place when (or just before) the trigger reaches In.
The Value module acts as a sample & hold, meaning that it sends to Out whatever is at Val only when a trigger arrives at In. Out won't change until a new trigger reaches In. So if a new value at Val is just one sample late, In will still trigger the 'old' value and the envelope is started with the 'wrong' velocity value. (You could check whether Order is necessary by leaving it out, then gate the envelope with different velocity values, and check the cables at the beginning and the end of the chain, what you should be watching is whether the value at the Envelope's G input is indeed the current value at the Gate output).

For another example: the sampler modules that have a separate input for Selection and Pitch. P will not give timing problems, since a sample's pitch can obviously be changed after it is triggered (pitchbend, vibrato etc.); the Sel input however acts like a sample & hold again: the value there has to be in place when the sample is triggered, else the 'wrong' sample is started. This will almost certainly give problems when connecting sequencers to both G and Sel (the current G starting the previous Sel), unless you insert an event delay with a delay of 0 ms (rounded to the smallest possible delay, i.e. 1 sample) before G, so that the new Sel is in place just before G arrives.

[anarchosyn]

Simple question, how can I increase the range of a bipolar cv?

For example, let us say I want an lfo to output a range of -80 to +80, but the stock module only outputs -3 to +3.. what increases this?

Damn, I know it is something really easy, but I can't fathom it for the life of me.....

[c s]

multiply ?

multiply

happy newyear!

chill out anarchosyn, it'll all be much clearer tomorrow

btw I read in this other post you were building a system around a mixed hard/soft setup, with complex midi routings... is it taking shape already? sounded interesting...

[anarchosyn]

Heh, thanx guys.. I was a little drunk when I posted that (in passing to a rather .. well.. interesting new years bash). Just wanted to make sure theren wasn't something even simpler (i.e. coded into the modules themselves) that would allow for this.

btw- trip, I'm making headway, but she won't be complete for another month or so. I've got the funds together, but I'm waiting to find a new flat before I start getting the studio setup (my girlfriend is all up in arms to relocate, so I'm sitting on my thumbs until we locate something).

Anyhow, Happy day to all..