1. The forum software has recently been updated. If you have comments or feedback, please let us know HERE! If you have arrived via a Google search, please use the forum search feature to locate specific information.

    Best wishes,
    The NI Team

control rate of 3000 not enough have to use ' a to e perm .module

Discussion in 'Building With Reaktor' started by gentleclockdivider, Mar 29, 2012.

  1. gentleclockdivider

    gentleclockdivider Forum Member

    Messages:
    135
    so I am rebuilding the osc. part of 'microtonic ( drum comp by sonic charge )' in reaktor on primary level
    The osc' in microtonic can be modulated by an lfo that goes up to 2000 hz and a pitch amount modulation range of +/- 4 octaves ,..but it is not using FM...just purely pitch modulation .
    In reaktor I modulate the pitch input of an osc .by another osc ..not using fm , but just the event .p input ..as we all know ..the event input is sampled at the control rate , but 3000 ( as a control rate ) really is not enough when the modulator and modulation amount is high = when it is an audio signal
    So I used an' a to e ' module with permanent control value set to a whooppping 44100 ...now the pitch modulation is exactly as the one in microtonic .
    My question is , it seems that the event input is able to process audio rate modulations ( when using a to e with control value verry high ) , why can't we choose a verry high control rate (similar to sample rate ) in the preferences .?
    Yeah o.k.3000 is more then enough when the event input only have to handele env. and lfo .
  2. salamanderanagram

    salamanderanagram NI Product Owner

    Messages:
    1,899
    it's hard to understand exactly what yr doing without a pic. if you're sending an lfo thru a perm a to e with 44100 as the sampling rate, that doesn't make the lfo itself run any faster. also, for a control rate of 3000 hz, 2000 is over the nyquist limit.
  3. sowari

    sowari Moderator

    Messages:
    24,182
    i don't understand why you are not use an Audio Oscillator as an LFO?

    sowari
  4. Chet Singer

    Chet Singer NI Product Owner

    Messages:
    656
    There are some instances where I want a high control rate, say 10 kHz or so, because it still uses much less CPU than a true audio rate. I use an A>E converter because I don't want the sound of my ensemble affected by someone else's use of the default control rate.

    But if you're doing FM, I'd really try to find some way of using a true audio modulation rate with audio oscillators. One way is to use a primary oscillator that has both P and F inputs. Tie the event-rate P input to a negative constant such as -200 and then use the audio-rate F input exclusively. Use a P>F converter to convert pitch to frequency.
  5. gentleclockdivider

    gentleclockdivider Forum Member

    Messages:
    135
    wel when using an lfo at audio rate ( as modulator ) I don't get the same results as when using an osc.as modulator .I don't know why , I think it's because the waveform of the lfo is not anti aliased at high freq.
    I just use a sine osc .as a modulator .controlling the p input of the carrier ...because the modulator is an audio signal I have to use an 'a to e ' module .setting the control rate of the ' a 2 e ' module to 44100 gives good results at the cost of high cpu oerformance .
    I know all this can be achieved with simple fm synthesis , just using the fm input , the reason why I use the p input instead of the fm is because I am rebuilding microtonic in reaktor , and microtonic uses pitch modulation instead of fm ,also the sidebands introduced ( harmonics ) with high pitch mod.are not the same as when using fm both are intersting
    ... .the results I achieved in reaktor are the same as with the microtonic ( similar results on osciloscope ) .It's just to bad that i have to use an insanely high control rate to achieve those results , using 3000 hz as control rate gives good results when the modulation is around 400 hz , any higher mod.rate suffers from artifacts that's why a very high control rate is used .
    @ salamander anagram , I know that the control rate does not affect the speed of the modulation , it affects the resolution at which the events are processed ..and In my case it affects the insanely high mod .speed
  6. herw

    herw NI Product Owner

    Messages:
    4,843
    .. and please use core (REAKTOR's future)
    ciao herw
    Last edited: Apr 1, 2012
  7. Big Gnome

    Big Gnome Forum Member

    Messages:
    105
    Either use a core oscillator with an audio rate pitch modulation input, or do something like this:

    Attached Files:

  8. gentleclockdivider

    gentleclockdivider Forum Member

    Messages:
    135
    I really have the feeling You are all taking me for an idiot
    , I know how to convert pitch scales to frequency scales etc..
    That was not the issue , like I said before , IT was my intention to NOT use fm synthesis but regular fast pitch modulation using the p input of an osc. and a verry high event rate for converting the mod audio sign to an event signal , and the only way to do that was to use a verry high res.a to e event rate ( at 44100 hz)
    As the modulator i use an fm osc ..setting pitch to -330 and just a freq.slider ...why did I use an osc and not an lfo ? , because a an lfo at audio rate is not antialiased ..
    @ gnome , if you had read my previous post , you would have noticed that I did not wanted to use FM synthesis ..;because the overall character of verry fast pitch modulation introduces a different kind of harmonics then whenh using fm synth .
    Also , the thing I am building is a replicate of sonic charge microtonic( just the osc and mod.parts ) and it does NOT use fm synthesis ..It is not because a modulator goes into audio rate into a carrier that the output is fm ..wel technically spoken there are both pitch/freq. modulation ...but the behaviour of the modulator signal is different
    Just try it yourself and you'll see.
  9. salamanderanagram

    salamanderanagram NI Product Owner

    Messages:
    1,899
    a picture is worth a thousand words
  10. colB

    colB NI Product Owner

    Messages:
    756
    Sorry, but this doesn't make any sense at all.
    Pitch is just a logarithmic mapping of frequency - so fundamentally, the only difference between frequency modulation and pitch modulation is a PtoF converter as suggested by herw.
    If there is a difference in your primary level implementation, modules then it will be either your 'code', or some quirk of the primary modules internal implementation relating to event rate issues or some such. In that case, the chances of that quirk being the same as any that might exist in uTonic are slim to none.

    I looked on the uTonic website and couldn't find any info about how it is implemented. Where did you find out that it doesn't use FM? It seems much more likely that it doesn't use what yamaha calls FM - ie Phase modulation.
    (DX7 and clones use Phase modulation because it sidesteps massive tuning difficulties caused by true FM.)
    True FM is much more suited to inharmonic sounds like those in percussion where tuning is less important. I'm just guessing that there might be some confusion in the information that you have about uTonics implementation.
    Sorry if this is way off base, but I can't understand how there could be any fundamental difference in the results from 'pitch modulation' compared with true 'frequency modulation' other than those caused by the implementation.
    Because there would be no point in having a control rate close to the sample rate. The point of control rate is to save cpu by allowing you to process stuff that doesn't need to be at audio rate less frequently.

    If you really do need an event rate that is say a quarter of the sample rate, then use core where you have much more control.
    It is easy to set up a very efficient control rate event pump that is a power of two division of the sample rate, or a slightly less efficient one that divides the sample rate by any integer. You can then use its output as a clock signal within your core macro to provide your own custom 'control rate'.

    IMO, anything that is modulating the audio stream should be at audio rate. Particularly if you are attempting to emulate a non-reaktor device in reaktor. Use control rate for midi, arps, generative music processing, voice management etc.

    cheers

    Col
  11. herw

    herw NI Product Owner

    Messages:
    4,843
    As you said pitch modulation is a kind of frequency modulation but with linear curve.
    So please use it as described excellently by col above. Frequency modulation like DX7 uses instead a fix frequency ratio. The principle is the same: modulating the frequency with audio rate.
    read this (pages 5 to 8):
    http://www.doepfer.de/a100_man/A111_man.pdf

    ciao herw
    Last edited: Apr 1, 2012
  12. Big Gnome

    Big Gnome Forum Member

    Messages:
    105
    That certainly wasn't my intention.

    I did read your post, and I'm familiar with the difference between audio-rate pitch- and frequency modulation; the picture I posted is an example of pitch modulation. I'd like to be able to help you out--am I missing something?
  13. herw

    herw NI Product Owner

    Messages:
    4,843
  14. gentleclockdivider

    gentleclockdivider Forum Member

    Messages:
    135
    I will upload the ensemble later .
    Herw ..the manual of microtinc says that the lfo can create effects much like fm synthesis when the rate is high enough ..but it is actual pitch modulation .microtonic does not use FM synthesis ..as used in regular fm synthesis and it doesn't use phase synthesis either ..it it just plain linear pitch modulation from a lfo that goes into audio range .
    I recreated it in reaktor and the results are identical with microtonic , I chop off my head if anyone here can recreate the results with fm synthesis ...it's all really obvious when you do a side to side comparison and look at the waveforms on osciloscope
    I will upload the ens..later , it is just a verry verry simple oscilator ....as I have described above
    ---
    here it is
    http://www.box.com/s/e6742912e5df86975c55
    ---
    if anyone has microtinc , just dial in the same freq.settings and compare , an osciloscope comes in handy ...( I use renoise built in one ) they will sound the same ..and the waveforms will look the same , the dial in the same freq.setting on a synth that does fm synthesis ....( makes sure that the modulator is not keyboard scaled to frequency...results willnot sound the same .
    The reason why I am pretty sure microtinc uses fm synthesis , is because I can't replicate the sound on a fm capable synth , u-he bazille also does verry high pitch modulation and there the sound can be replicated , but not when using fm synthesis ..

    Attached Files:

  15. salamanderanagram

    salamanderanagram NI Product Owner

    Messages:
    1,899
    i think what people are suggesting is this (your method on top, mine on bottom)

    seems the same to me? uses way less cpu.

    chop off your head time? heh.

    Attached Files:

  16. colB

    colB NI Product Owner

    Messages:
    756
    This is probably due to some implementation difference - maybe the fact that utonic uses oversampled oscillators allows it to handle higher frequencies better ?

    Anyhow, as Salamanderanagram has shown - you are incorrect in your assertions.

    Another simple test is to modify the structure in the image you posted as follows:
    Remove the A to E perm, and connect the switch output directly to the + module.
    Attach the output of the add to an Expon(F), then connect that to the F input of the sine osc. Remove all connections to the P input.
    Notice that the sound is identical, but you have a cleaner structure - no more daft audio rate event stream.
    Now, to avoid serious injury, use a spork for the decapitation.
  17. salamanderanagram

    salamanderanagram NI Product Owner

    Messages:
    1,899
    FM synths vary in implementation.
    operator doesn't work like this, for example. i'm not sure what the 'standard' is, if there is one.
  18. herw

    herw NI Product Owner

    Messages:
    4,843
    i think you misunderstand the word FM-synthsis and the name of the F- and the P-input:
    FM-synthsis is a special methode to modulate oscillators.
    Using the F-input of any REAKTOR-oscillator doesn't mean that you are using such synthesis.
    P- and F-input determine only the step-width of the basic ramp oscillator inside of every REAKTOR oscillator.
    As mostly the P-input is used to change the basic frequency by keyboard it is made as an event input to lower the CPU usage. It is not neccessary.
    So using only the F-input of an oscillator, you have to calculate such events to get a proper frequency change (P2F-module). You have to hold in mind that an unconnected P-input means that it is connected to 0 which means a frequency a ca. 8Hz. So if you want to have the „exact” frequency you have to add a -300-constant to the P-input (means a very very low frequency).
    The rest is only basic mathematics.
    BTW you can test pitch modulation with my uploaded ensemble RUHR.

    ciao herw

Share This Page