uModular Collection Beta
  • I guess it is a bug, The original logic relied on a linear display of notes and so the first was always the root. The same applies here, but I see what you mean. I arranged the mask like a keyboard to save space, but that does sort of imply that the first mask is "C". I added the scales primarily as a learning tool (and because I can never remember the pentatonic), since you can obviously mask out notes to make any scale you want, so it would definitely be better if the graphic reflected a real keyboard. I'll see if I can shift the display to match the root. I had an epiphany yesterday, regarding my approach to quantization, My current approach works pretty well when the intervals between notes are arbitrary, but is somewhat complex. I think I can simplify it in the case of an evenly divided scale.
  • That would be super cool as it would then serve as an educational tool for learning about scales :)
  • Cross-posts again!
  • @stschoen all of these will be documented! Which ones do you find the most cryptic though?
  • Here's (some) of the collection put into folders and categories - let me know what you think! Trying to keep the menu diving to a minimum but also want to leave room for big expansion.
    Screen Shot 2017-10-05 at 6.32.28 PM.png
    1158 x 460 - 124K
  • looks good to me! maybe switches are a type of utility though?

    Also, the subharmonic VCO is a out of the -1 to 1 spec. Attached is a simplified version that stays in the normal range.
    Screen Shot 2017-10-05 at 3.39.44 PM.png
    1206 x 562 - 47K
    Screen Shot 2017-10-05 at 3.49.45 PM.png
    1056 x 611 - 69K
    Subharmonic VCO.audulus
  • @robertsyrett - yeah that's a good point - and yeah these aren't finished per se but just was going crosseyed trying to make new ones so decided to start organizing what was there so i coul dhave it in the right click menu.

    The subharmonic version you made doesn't work really the same way that mine does - you might as well just have a sub square oscillator. What mine is doing is rectifying the wave and then passing every other wave period that comes through. It's a technique I came across in some dark DSP part of the web.

    Check out the difference:
    Screen Shot 2017-10-05 at 7.29.16 PM.png
    1380 x 872 - 146K
    Screen Shot 2017-10-05 at 7.29.39 PM.png
    1352 x 878 - 145K
  • Screen Shot 2017-10-05 at 5.29.07 PM.png
    862 x 480 - 81K
  • @robert syrett - the scope in the background is the unmodified sine, the next one "up" is the abs(x) and shifted down from DC to AC range (that's the part you're missing in yours right there) and the one on top is the sub octave generated by the flip flop alternating between passing/not passing the wave. It's not really a "sine" but it definitely isn't a square and sounds different on whatever waveform you put it through. this scheme doesn't work too great when there are several oscillators mixed in though, so that's why i built it into a single oscillator like this.
    Screen Shot 2017-10-05 at 9.05.13 PM.png
    728 x 366 - 83K
  • @robertsyrett - this one is 2 octaves down - it doesn't retain the same shape, but it does sound like it's 2 octaves down from the main oscillator.
    Screen Shot 2017-10-05 at 9.10.01 PM.png
    678 x 324 - 153K
  • @Robertsyrett just to add, you have to make sure you have a z-1 in your flip flop so it can flip fast enough to track the higher pitches.
    Screen Shot 2017-10-05 at 9.11.25 PM.png
    1658 x 1330 - 213K
  • Is the unit delay node preferable to the feedback delay node in all instances?

    Also, I still don't quite grasp the method to your madness but I assume you like the way it sounds. You should probably internally trim the signal though.

    I really like this new aesthetic. I'm finding that using these I can make pretty complex patches with much less cpu.
  • Hey I bought a braids when they said that they were going to discontinue them. It might be fun to make some waveshaping oscillators in the same vein as the braids that all have two knobs that change the timbre.

    You're gonna need a lot of oscillators to get to 1,000 modules after all.
  • Here's the revised quantizers. While working on the scale display for the larger one, I discovered a bug in the quantizer module. When an incoming note was exactly halfway between notes it would select both and output their sum. I revised the module so that wouldn't happen and it also simplified it considerably. @biminiroad, you might want to replace your current versions with the updated module. I still need to update the larger variable units.

    10-6-17 Files updated - see below.
    Screen Shot 2017-10-06 at 9.11.12 AM.png
    1802 x 945 - 241K
    Screen Shot 2017-10-06 at 9.39.16 AM.png
    1221 x 974 - 127K
  • @biminiroad, you asked which ones were cryptic. Here's an example. Without opening this up and puzzling out the logic, there's no way to tell what this does. Is it a filter, a sequencer or a TUFLM (totally useless flashing lights module)? Don't get me wrong, I love the design, but a name and some internal docs would go a long way to making this more useful. I also wonder about our Windows based friends who won't have a clue.
    Screen Shot 2017-10-06 at 10.11.09 AM.png
    683 x 899 - 67K
  • @stschoen - thanks for the feedback! As for some, the VCFs and VCAs are all labeled, and the VCOs will be too when they dont have waveform selectors. VCOs also have octave inputs instead of gates which distinguish them from LFOs. Both are also free running now so you can see the output of the VCO is scintillating purple and the LFO is red.

    Others are meant to be “discoverable” in a sense....Well for one they’re all “labeled” by the name of the module, so if a user is adding them to a patch they’ll do so theough a right click menu or through the new folder system on iOS thats coming.

    But part of the unified design sense is that modules are consistent from one to the other - so all filters have cutoff at the top, then Q, then input drive, then envelope amount/input.

    The R input is always “randomize” and rs is always “reset” for sequencers. But even if you didnt know that all you’d have to do is attach a gate to each input and watch what it does.

    And yeah inside each one will be a little summary of what it does and how it works node by node :)
  • I understand that this is a work in progress and I'm certainly not trying to be critical. How are you indicating a gate vs. a 1/octave vs. an audio input etc? Duh I think I get it now. Pardon me for being a bit slow.
  • It occurred to me as I was walking the dogs, what a great set of building blocks these will make if we get the ability to expose the UI of a sub-module. I envisage the module exposed pretty much "as-is" with a border drawn around it. How about if the inputs and outputs of an exposed sub-module automatically became "normalized" connections. Whatever was connected at the sub-module level would remain connected until you connected to the input or output at a higher level. That way you could create bigger, more complex synths without losing the ability to re-configure on the fly.
  • @stschoen - the audio inputs will be labeled "a" but outputs will be labelled by the lights, basically. And totally be critical! No offense taken, I definitely want all the feedback people can give on this so it's really a community effort.

    Speaking of which, check out this folder organization and tell me what you think of it. I'm trying to keep diving down to a minimum.

    Some naming conventions: A-R is two separate attack release knobs, while AR is one knob that controls both.

    Trying to label them in ways that people can search easily - so like, you know you want a random envelope, the envelope's name starts with random and you can then narrow your search in just the random X envelope portion of the list (rather than having a whole separate "random" submenu.

    For basic synth voices, I will name them based on their distinguishing features rather than weirdo names. So "1xVCO Curved A-R K35LPF Digital Delay" is a self-contained single voice synthesizer with A-R envelopes for VCF and VCA, the VCF is a K35 LPF, and it has a digital delay.

    Other more unique/difficult to describe succinctly synthesizers can of course have "wacky" names or whatever.

    VCOs - basic ones use the oscillator node and are mixing/crossfading/etc. Subharmonic VCOs are different than Suboctave in that the subharmonic is generated whereas suboctave indicates an oscillator 1 octave down.

    Wavetables are created from the spline or mapper nodes but are different from graphic VCOs - the wavetables should be selectable/fadeable, and generally remain "static" on the inside of the panel. The graphic VCOs are where the mapper/spline is brought onto the front panel and meant to be manipulated.

    LFOs are basic (just basic waveshapes) or complex (fm/am/pm, crossmodulation, etc), or long period (meant for installation 1+hr to many days or years drones, that kind of thing).

    Granular effects use the delay node in a granular way (like my icebox freezer).

    Some of these categories, like MIDI, will get more expansive when the MIDI out comes out.

  • @stschoen "It occurred to me as I was walking the dogs, what a great set of building blocks these will make if we get the ability to expose the UI of a sub-module."

    This is something taylor is working on :) That and group UI move. Exposing I/O is something that would be complicated, so he might have to pass on that (you could still make "extra" I/Os, but this would get rid of redundant knobs, buttons, etc).

    "How are you indicating a gate vs. a 1/octave vs. an audio input etc?"

    gate = green light on input/output
    modulation = red light
    audio = scintillating purple, with green light as a hard clip indicator (haven't implemented this yet, but will)
    octave = "o" on the input
    random = R
    reset = rs
    audio input = "a," to distinguish it from m inputs. it should be assumed if an "a" is not present at an input (except for the very top of the module) then only modulation/gate signals should be plugged in.

    In general, the audio path takes precedence above all others at the top of the module. If a module has audio, octave, gate, modulation inputs, this is the priority from top to bottom:

    1. audio
    2. octave
    3. gate
    4. modulation/"a" modulation - at bottom of module under input attenuator/attenuverter/positive domain attenuverter). If there are lots of modulation inputs, they should still go at the bottom, since the audiopath can remain compact going from left to right while modulation happens "below" the audio path.

    again, this is all stuff you don't necessarily need to worry about when building your modules here - I'll adapt them for you if you don't want to.
  • I like the folder organization. You might want to add a subfolder somewhere for percussion and similar (kicks, snares, bird calls, etc). FYI, Robert's quantizer is mis-named as a sequencer.
  • @stschoen whoops forgot to include the new version lol here it is
  • Oh and nice catch on the chromatic quantizer lol I was sleepy when I did that
  • Much clearer with some explanation! I would suggest changing the audio colors so that red is clipping if possible. Also how about s for set for modules that have it. I assume R is a gate that generates something random internal to the module?
  • V2 is even better. I see you added a folder for drums. Are you planning to add this to the context menu (right-click)? If so you might want to keep these separate from the earlier modules (maybe a uModules folder at the top. Obviously the iPad version already needs some work as far as organization so I'm not sure what makes sense there. BTW what do you think of the new files application on iOS 11? It might be a good way to organize on the iPad if you didn't have to maintain backward compatibility.
  • @stschoen the problem with red=clip is that it would require marginally more CPU since it would require the clipping indicator you see in the output module rather than the green one you see in the VCAs and VCFs under the overdrive parameter. It wont be super obvious what it means but once you get used to it it’s saving on an extra separate node (no more light node for clipping) and using just 1 expression for clipping. Also check out the wat I do the audio light in the modules I made its less CPU intensive than he one you’ve been doing. It takes advantage of the fact that the rgb node only responds between 0-1 by inverting the signal as it goes to blue.
  • @stschoen - the files system is what we're going to use for the new organization system - so basically you'd see the same folder organization on iOS as you do in the right click menu. I think (hope) the idea is to eventually deprecate the old module library and integrate all of its modules into the uModular library, so there's only that one.
  • As for keeping them on top, there's no way to organize that folder system at the bottom since that's alphabetical by whatever folders are in there.
  • @biminiroad I do find it tedious opening each file in the new structure, whereas opening up the collection was more like cracking open a travel watercolor palette. Is there room for an aggregate sub collection, such as all the VCA's or all the VCO's?

    @stschoen I love the new uModular Quant v4, it's endlessly pleasing to watch the scales move up and down. Is that the same on the really tiny v2? I have to admit one of the things I've started doing with it is just typing in random numbers to see what scale it makes.
    Screen Shot 2017-10-06 at 11.31.10 AM.png
    1103 x 488 - 91K
  • @robertsyrett yes! Good idea - will add a collection for each type of module and put thay in a separate folder
  • @RobertSyrett the really tiny V2 has the same logic as the larger one without the display or mask. The input digits have not changed. The lowest binary digit is still the root, the next is the first semitone up etc. The number is selecting which scale degrees are enabled. I "rotate" the number to shift it relative to the root note selected so that the correct notes are selected for each key. Previously I shifted the input note then shifted the output of the quantizer in reverse.
  • @biminiroad I think you might need to double check your sequencer knobs. The 8-step basic sequencer is pretty off, and I think the others might be also if you were cutting and pasting the knobs.
    Screen Shot 2017-10-06 at 11.47.18 AM.png
    591 x 815 - 73K
  • @stschoen Cool, this is a welcome upgrade in the quantizer department.
  • I've been working on a VU meter and I could use some help. I think I've got the meter ballistics and scaling about right but currently the clip indicator is not the best.
    Screen Shot 2017-10-06 at 2.47.22 PM.png
    738 x 559 - 48K
  • If you want the clipping light to be less flickering, try the uGate smear
    Screen Shot 2017-10-06 at 12.02.09 PM.png
    692 x 362 - 62K
  • Perfect! I set the clip at .995 instead of 1 since I believe that Audulus will clip audio at 1. I think the meter should be fairly close to the VU standard with 0 being -1 to 1 peak to peak. It didn't make much sense to set 0 to +4dB as is typically the case since we have no headroom above 1. The standard calls for a 300 millisecond rise time which I think I've modeled with a LPF. Should be pretty close anyway.
    Screen Shot 2017-10-06 at 3.57.57 PM.png
    1015 x 703 - 66K
    VU meter.audulus
  • Word to the wise. Don't get curious about what happens when you turn off store in iCloud on your iPad then turn it back on. Seems to have deleted several files from iCloud. Fortunately I was able to download them from the Forum.
  • @stschoen - it shouldnt delete them - might just take a few minutes to load all of them back on over WIFI. Happened to me before too, just waited a while and it all came back.
  • Smaller tiny quantizer with I/O at the top. No functional change
    uModular tiny quantizer V3.audulus
    Screen Shot 2017-10-06 at 4.35.22 PM.png
    935 x 720 - 112K
  • In this case they don't seem to have returned. I was investigating the new folder system on iOS 11. Turns out that you get a browse option in Audulus even if the "store in iCloud" switch is off. Seems like I've done this before as well without any ill effects. I think this might be new behavior. The files that were deleted were all things I had worked on in the last couple of days.
  • @stschoen you may have forgotten to connect the active step to the LED display for the v3 medium sized quantizer. The tiny quant is very nice too, it definitely is getting optomized!

    Here are some reduced functionality sequencers as so much of the fun of these is creating complexity from simple parts.

    So here is an 8 step sequencer that can slew steps, a sequencer that can skip specific steps, and a bare bones gate sequencer.
    Screen Shot 2017-10-06 at 2.27.12 PM.png
    1845 x 752 - 311K
    uSkipStep uSlewStep u8GateSeq.audulus
  • Oops! You're right. Here's the fixed copy. I like the little sequencer modules, very inventive.

    10-6-17 fixed the "fixed" copy - see below
  • @stschoen - current note seems to disappear in this little range?
    Screen Shot 2017-10-06 at 5.56.15 PM.png
    1670 x 868 - 240K
  • Screen Shot 2017-10-06 at 6.15.44 PM.png
    634 x 596 - 50K
  • Those are some insane hairstyles!
  • Screen Shot 2017-10-06 at 6.22.41 PM.png
    630 x 600 - 53K
  • btw @dcramer - if you're around, man I was checking out your poly krell patch online today, very nice!
  • we should definitely have a krell patch module in this collection
  • @biminiroad, thanks for the heads-up. It's fixed! (at least I hope so)
    uModular Quantizer V4.audulus
  • Going to make DLX versions of all the filters that looks roughly like this.

    The feedback is a control to feedback the audio output to FM the frequency cutoff. There are also two attenuators for envelope inputs. I'll probably change them to positive domain attenuverters so you can invert the envelope (without needing an offset).

    This one isn't hooked up yet, but just wanted to share the format to get some feedback :) Gonna go back to the real world for the evening, see you on Saturday!
    SEM LPF DLX.audulus
    Screen Shot 2017-10-06 at 6.38.53 PM.png
    1052 x 1410 - 186K