Sunday, June 28, 2009

Working on a side project...

I've had the opportunity to work on a very interesting and exciting project for a couple of months, so my current ReFill project will suffer a bigger delay than I was expecting.

Nonetheless, I'll try to release some free stuff 'till the final release of it :)

Any comments and suggestions about the ReGRAINZeR demo file are welcome, even if I don't comment back, I read'em all! :)

8 comments:

  1. You have bugs in ReGRAINNeR ,
    Use B1 or A#1 for long time and watch on the volume meter .
    It not works always , but many times volume is going up with time ,more than 2xOrginal Volume.

    And there is a trick with will help you remove 50% of CV cables and Spltters , do you want to know how to do it ? :P

    ReplyDelete
  2. hmmmm... I must check those then, thanks for the feedback! :)

    I don't know if you noticed this, but the volume of the loop (the dry/mix) is driven by the velocity you apply.

    By design, I made it this way, because you can on-the-fly, apply some effect where it applies the looping effect briefly while kicking in the original source (there's an example of it in the demo song, with the quick and brief stuttering of the guitar).

    This is why CV is used. My LoopBoxes simply use Mix Mutes, with no velocity sensitivity.

    Also, right now, the demo uses V0.6.2 beta, and as I'm sure you already know, betas are pre-optimization, so this is why there's still so much CV cabling. A lot of spiders and CV in there is used simply for clarity while still in development stage. Some cabling routes are still unnecessarily long, with extra hops that will disapear in the final version.

    The stage from V.0.9 to V.1.0 will be exactly that, optimization of the stable feature list implemented.

    I think I already mentioned, but there's still some features missing from ReGRAINZeR that I'll add in V.0.7 (external control of sister/companion FX boxes).

    V.8.0 is also planned to have some additional things I'm working on. If they work as expected, they get in on it too :)

    ReplyDelete
  3. ok, I've tried holding B1 and A#1 for 30 seconds with various velocity values, from 127 to 32 and the only thing that happens to me is that at lower velocity values, the loop fades while the original source returns (as its supposed to happen, by design).

    Can you tell me for how long do you need to hold those keys and at what velocity value to replicate that bug ?

    ReplyDelete
  4. OK , try it on high sample rate.

    1.I open demo song
    2.Delete Devices:
    a)ReGUITARZeR
    b)Nn-xt
    c)rex-3
    d)matrix
    e)guitar combinator
    3.Delete Automationtrack for ReDRUMZeR
    4.Play loop

    I don't change anything in mixer

    5. Use B1 and try glitch loop on 50% of velocity. (It will be easier when you fast click B1 2 times )

    Not always it want to grow , and it is hard to see it in reason mixer . ( probably reason have compression or something)
    Tracks in preview can sounds different after render. ( I hate mastering )
    Try connect reason output to audio analyzer or if you have very good years you will hear it.

    I find this bug in my own old combinator because i have in my WIN7 gadget with volume control and audio output display.

    If you will see it , probably you will hit your had in to wall asking where is that evil bug and how to fix it.
    I can say it .

    You can't make 100% mute using inverted vocoder cv.
    When you create/open vocoder it looks it is working , but when you play "voconote" and release it it will back to 98% or something like that.

    6.Mute channel 14 , play B1 one time , and when than maximize your sound output in os , or hardware. You should here your loop very quiet.

    Now in my combinator I don't control mute by cv. I found other way to do it perfect.

    Now because I explain you how it works, can you show me why you use thor with aftertouch .
    I know what is aftertoch but
    I don't have it in my keyboard ;(

    Also I still thinking about collab with you , because we are have very similar combinators.
    It will be problem if I and You will try sell it in the same time , or you will give it for free download. ( about this case , please contact with me on mail )

    ReplyDelete
  5. Thank you for describing the steps to replicate the bug.

    I did all of what you said, and again, I tried with various velocity values, from 32, to 64, to 100 to 127 and did the double and triple quick trigger of B1 and A#1 and hold it for more than 30 seconds and the device stays stable and does what I expect it to do.

    The mixer levels keep stable too, I also used an external Signal Analysis, and I saw no progressive raise of the dB's.

    Besides synthetically generating the velocity values (using a RPG-8 to force them, or using the sequencer automation) I also tried it with 4 MIDI keyboards, all with different velocity curves, and still no problem.

    One of the keyboards I can specify the value of the velocity, so I did force some values with it too, and still couldn't replicate what you describe :(

    I'll have to wait before someone else confirms what you describe and then I'll have a second look at it. With some luck, maybe it'll happen to me too :)

    As I said earlier, my LoopBoxes are much simpler and basic devices (so much so, that I'm thinking of releasing some for free, because there's really nothing new there), so with those I use no CV, it's all done with the usual channel mute technic and Combinator Modulation Routing mappings.


    About the BV512 not generating enough CV to mute a channel when inverted, that depends on the velocity you applied to it and also and most importantly, to the CV scale you set at the back of the device.

    Are you sure your keyboard can generate a velocity of 127 ?

    What I've seen is exactly the opposite: If I add 2 CV signals and feed them into a BV512 band input, it'll display a higher graphic value than the usual one with 127 CV max. ...which I already used once in a device for some tests I needed.

    So, about Thor:

    Thor can be many things. It can be a CV math/routing machine :)

    About the Aftertouch: its just another performance value to use, like you use the Mod or the Pitch wheel, so don't sweat about it if your keyboard doesn't have it :) ...you can even make the Combinator fake Aftertouch with the Modulation Routing if you really need to test it :)

    ReplyDelete
  6. Oh , my mistake.

    My midi keyboard is set to 100 , not 127 , I forgot about it :P

    but wait... I test it again on 127 , end this bug still exist.

    Did you create step 6. ?
    It is inside comment .

    Did you have almost muted sound ? o.1% or something.

    This must 100% works .


    Btw, abot increasing sound I found where is bug.
    Not in combinator , not in vocoder , not even in reason.

    LOL
    It is in windows :P

    I connect Audacity via virtual audio cable to reason , and record that broken glitch .
    There is not bug in output waveform.

    Probably windows is resampling and on some freq it do it wrong.
    Or this that reason output bug with I hate , when reason play in preview one thing , and after render it plays it different.


    Sory for problems , but btw. now you are sure that you have one free beta tester :p

    ReplyDelete
  7. LOL!!!! *phew* I'm glad we sorted that all out, eheheh :D


    Thxs for checking the ReGRAINZeR demo ^^

    ReplyDelete