Scale-Invers mapping with Exponential


    Sep 26 2008 | 1:09 pm
    Hello,
    I am controlling Max with a bcf
    I use scale with the fifth argument
    because you manage in a better way
    the volume with a ctlin
    but when i send the value back to ctlout
    (invers mapping)
    my fith argument becomes wrong
    could you help me ?
    thanks
    s.gaxie

    • Sep 26 2008 | 1:36 pm
      If I were going to do some scaling and
      wanted to invert the outputs (high to low),
      I'd use the zmap object.
    • Sep 26 2008 | 2:08 pm
      thank you very much
      but with your solution
      we do not have the same result
      on bith sliders...
    • Sep 26 2008 | 4:40 pm
      That's because I looked at this before my
      morning coffee. Sorry.
    • Sep 26 2008 | 5:09 pm
      Check out the Litter Pack:
      The starter kit is free and has a better scaling object; lp.scampf. The arguments for the input would be:
      lp.scampf map 0 127 -1. 1. pow 1
      and for the output
      lp.scampf map -1. 1. 127 0 pow -1
      Notice the negative exponent on the output. The max scale object never seems to get the exponent thing quite right...or at least it never makes sense to me. I am enjoying some coffee as I type.
      Lewis
      Quote: s.gaxie wrote on Fri, 26 September 2008 06:09
      ----------------------------------------------------
      >
      > Hello,
      >
      > I am controlling Max with a bcf
      > I use scale with the fifth argument
      > because you manage in a better way
      > the volume with a ctlin
      > but when i send the value back to ctlout
      > (invers mapping)
      > my fith argument becomes wrong
      > could you help me ?
      > thanks
      >
      > s.gaxie
    • Sep 26 2008 | 7:53 pm
      This patch might help give some ideas or clear things up about the scale exponent, and could be useful in other situations too. Using other functions (like sine) to generate value curves can provide interesting data for parameters like volume, frequency, contrast, etc. Plus you can simply draw in the curve you want.
      I like scale too, but I've also run into some weirdness with the exponent (though that definitely could be in my head). You can see some values in the table that aren't quite what one might expect...
    • Sep 26 2008 | 10:33 pm
      thanks both of you
      fot those answers
      it is late in Paris
      and the Litter pack does not want to run for the instance
      but it seems to be my solution
      thanks seejames for your nice scaling patch
      I will probably use it in special effects
      but the topic is very interessant
      I do not like this systematic way in max to scale values in a linear manner.
      but it seems to be a bit heavy
      when you have 60 ctlout...
      I am sure we could use an other
      fifth argument in scale to get
      a good invers mapping...
      cordialement like we say in french
      s.gaxie
    • Sep 27 2008 | 7:44 am
      s.gaxie schrieb:
      > I use scale with the fifth argument
      > because you manage in a better way
      > the volume with a ctlin
      > but when i send the value back to ctlout
      > (invers mapping)
      > my fith argument becomes wrong
      > could you help me ?
      I made some research on the fifth parameter of scale. My conclusion was
      since then: don't touch it, its worthless, even the original programmers
      don't know what rule it follows.
      Instead, use dbtoa and atodb to create linear to exponential behaviour
      for levels, if you want the zero to be at complete silence, use a sel 0
      for this special case...
      If you deal with pitches, there are mtof and ftom...
      And of course Peters Litter package will have some options too...
      --
      Stefan Tiedje------------x-------
      --_____-----------|--------------
      --(_|_ ----|-----|-----()-------
      -- _|_)----|-----()--------------
      ----------()--------www.ccmix.com
    • Sep 27 2008 | 9:57 am
      thanks Lewis
      the solution with the Litter pack
      seems to be the better one
      if I send general parameters with
      messages I can grab them on the
      bcf with ctlout with the same value
      in an exponantial mode
      you have so more steps in little
      values to choice your volume
      thanks
      s.gaxie
    • Sep 27 2008 | 5:44 pm
      Quote: s.gaxie wrote on Sat, 27 September 2008 00:33
      ----------------------------------------------------
      > it is late in Paris
      > and the Litter pack does not want to run for the instance
      ----------------------------------------------------
      Litter Power will run in Paris any time of day or night.
      For that matter, it will run in any time zone at any time of day or night. N'importe ou, n'importe quand, n'importe quoi.
      It seems from your follow-up message that you got it running, but if there are questions let me know.
      I'm glad you've found lp.scampf helpful.
      Best -- Peter