Features request for next firmware versions...
  • I’m starting to feel like I’ve reached my request quota. Anyway, if not, it would be really cool if there was a way to make the arp retrigger on each keypress if no keys were currently held. Basically so that the arp does not keep running if there are no keys down. This way you can sequence the Shruthi externally but run the arp at a different tempo, and make deliciously sloppy sequences without worrying about a note getting dropped when it’s played before a Shruthi bar passes after the previous note.

  • ilcense: the arpeggiator currently behaves with this “release time” on the clock to make it easy to switch chords without super accurate timing. I’ve played a few synths whose arp or clock got confused when switching from, (for example) C to Dm – resetting their clock or counter in the interval between the two notes.

    So this would require a new menu to enable/disable the “clock release time” I guess…

  • Yeah, that makes sense, I could see many situations where it would definitely be preferable. I just like using my Shruthi with my MPC and its arp complements it really nicely. That’s why I would have rather it been an option than overwriting the current behavior.

    It looks like there are 7 sequencer modes, so probably 3 bits storing this behavior in the patch… meaning there is one other possible value there, so could it fit there? Of course that would mean that this was only available in the arp and not the step-sequencer though, which is a little sad. I can’t think of anywhere else to put this option.

    Sorry for so many requests. I love this synth and I just keep of thinking of more little things I want to do with it. I should solve my compilation problems so I can build some firmware with my little tweaks. I think it’ll be the same way when I get my DSP board built.

  • Apologies if this has been suggested before. If it has, I blinked and missed it!

    Is a paraphonic mode feasible? I realise there is only 2 main osc’s, but still, it might be fun if the sub and noise osc are brought in to play too.
    Naturally, I stole this idea when reading about the Waldorf Pulse 2 which will supposedly offer this feature. I hadn’t even heard of the word “paraphonic” before this thing came along. But I realised it might be something a Shruthi is potentially capable of..?

  • A paraphonic mode is feasible under some constraints:

    • one osc per voice ; and only a for cheap aliased saw / square, not the fancy interpolated wavetables, oscillator algorithms, fm, bandlimited saw/square etc…
    • it would require so many changes that it would have to be a distinct firmware, not something you could just enable/disable in the main Shruthi-1 firmware.
    • a great loss in sound quality – due to the fact that only one fourth (assuming a 4-voice mode) of the 8-bit dynamic range would be used when only one note is played.

    I have already tried the idea a long time ago on the Shruti-1 and it sounded like shit. Aliasing is already nasty in mono mode, in polyphony it’s horrible. 1 osc is thin – all the poly sounds we love (pads, strings) rely on 2-osc or chorus… And sending a stack of sawtooth waves into the same filter sounds very mushy… Also, pretty much everything in the synth – arpeggiator, sequencer, etc. would be broken since the assumption that the synth is mono is hardcoded in many places in the code for optimization reasons. I’d rather draw a line and set a standard of quality than cram more and more half-baked features…

    I am not wiling to invest time into that unless someone can fund the 2 or 3 weeks of development.

  • Eeek. We’ll scrap that idea then! Evidently it would be crap, at best.. I (foolishly) thought that each osc could perhaps be made to trigger separately, much as they stand. I had no idea a major code overhaul would be required, and large quality sacrifices would result.
    Thanks for the explanation though :)

  • If each osc is made to trigger separately, that’s only 2 voices. Would such a duophonic mode (in which notes are ping-ponged between osc1 and osc2, and the sub plays the last played note) be helpful?

  • Well.. The Arp Odyssey works something along those lines, right? (quasi duophonic, all piped through a single vca and filter) And no one claimed those sound bad.
    If it were possible, I figured throwing in the sub & noise oscs for voices 3 & 4 would be interesting in an experimental/fun way, of some kind..
    Anyway, don’t waste more time thinking about it, if you are quite confident it would be rubbish. Im not interested in any half baked compromised lash-up any more than you are :)

  • Such a duophonic mode might be playable. But 4 voices of aliased saw/square, sheesh!

  • I was thinking 2 ‘proper’ voices/oscs.. + possibly the sub osc & noise osc. Each doing exactly the same as they are currently. Except triggered in a polyphonic manner (Rather than 4 new oscs/voices all doing the same thing, like saw/square, as you appear to suggest?)
    So sort of 2.5 voices with the 4th, noise osc which could join in for some kind of rhythmic element perhaps.
    (Actually, this crazy idea is almost multi-timbral as well as quasi polyphonic!)

    I imagine sequencing something like that with 4 channels on Cyclic might just be good.
    If that is, unacceptable compromises were not involved.
    Anyway, even if that was possible, I have zero clue how one might best handle retriggering of the envelopes!? And thats on top of all the other issues you mentioned, like the inbuilt arpeggiator etc etc..

    Oh well, t’was a fun thought!

  • How would you decide if a note you play on the keyboard should trigger osc 1, osc 2, the noise or the sub?

  • I would think the mulitrimbral mode would have each voice on a different channel, or some sort of auto note assigning. The closest thing out there would be something like the midibox sid, even though that allows you to assign the filter to a voice as needed, or something like midines.
  • Lets not dwell too much on the multi-timbral thing, that was me being silly. I just mentioned it, because as I imagined this feature, it wouldn’t be totally unlike that because each ‘voice’ will be its own osc, and hence sound different. The first 2 oscs could be set up the same, as we can do now. Or they could be set differently, again, as we can do now. Plus of course no matter how you set up the noise and sub, those will always be very different anyway.
    The UI is probably going to become over tedious if we started trying to assign each osc a midi channel, and that would introduce new problems all of it’s own anyway..

    How would note triggering work? I guess if I were to play only 1 note at a time, then I should only expect to hear osc 1. I don’t think I’d want a monophonically played line cycling through each osc, for example.
    2 notes at once, then osc 1 & 2.
    2 notes at once/held, and a 3rd note repeatedly triggered (like 16th’s or whatever), then I’d expect to hear osc 1 & 2 together, and the sub osc doing it’s 16th thing.

    Is this something that might be workable then? Im thinking when I first mentioned it, we were not quite on the same page. But you seem to see what i’ve been getting at now?
    I still couldn’t suggest how envelope triggering would be handled though! Perhaps there is still too many issues like that to make this practical.
    It will be interesting to see how it will work on the Pulse 2. Not that i’ll get one, as I already have an original Pulse and far too many other monosynths scattered about the place as it is!

  • Time for a faster processor? :)

    Is there some fundamental problem with ARM boards? obviously they can be big and ugly but the Maple Mini is pretty compact.
  • A pseudo-duophonic mode with the two oscillators is OK, and is not very alien (Arp Odyssey, Polivoks). This 3-voices thing with the sub sounds very dodgy – every third note this would play one octave below?

  • > Is there some fundamental problem with ARM boards? obviously they can be big and ugly but the Maple Mini is pretty compact.

    • These boards solve a non-existent problem. The idea of a paraphonic Shruthi “for free” through a software upgrade is nice. The idea of having to upgrade the hardware just for this feature is not so nice. Paraphony is fine in the “we had some spare CPU cycles so we gave you this fancy mode”, but not as the starting point of a design.
    • They are expensive. I want to pay for an ARM chip on a board, not fund the development of yet another arduino libraries clone.
    • Some of them are not intended for resell / commercial use (these are devboards after all).
    • They tend to have short lifecycles, and manufacturers rarely address the issue of “pin compatibility” between generations of boards.
    • They are bulky.
  • Pseudo-duo power chords! AWWWW YEAH! +1

  • Perhaps some kind of "secret" key combination would be the answer, which would turn machine in to multitimbral mode while pressed simultaneously at power on.
    This way multitimbral mode would not mess normal UI.

    So in multi mode every OSC that could be used separately would have different MIDI channel.
    If using multiple MIDI channels is not the right way, but some sort of exotic note triggering logic would, how about using round robin? Ensoniq EPS 16+ had that kind of solution where was possible to point sound to special outbus that directed sound to all different outputs in turns. Like dry, effect bus 01, effect bus 02, dry, effect bus 01...

    How about random trigger? Would sound interesting with fast sequences.
    Possibility to set note ranges for every OSC would be more useful, though.. . . it would give lots of possibilities if making those zones overlap each others. Like at the bottom there would be just OSC1, little bit higher just OSC2 and rest free keys SUBOSC and/or Noise.
    And more: what if every OSC could have two or more key zones? First notes just OSC1, next ones OSC2, then OSC1&noise, OSC2&SUB.. . . Root key for every OSC should also be set in that case.

    IMHO, this all would make worth of effort only if sound quality would not be compromised. =)
  • >A pseudo-duophonic mode with the two oscillators is OK, and is not very alien (Arp Odyssey, Polivoks). This 3-voices thing with the sub sounds very dodgy – every third note this would play one octave below?

    Haha, No! 33.3333% weirder than that.. Every 4th note is noise! :D
    Honestly, it could sound like complete ass. But just maybe with a bit of careful setting it up, it could be equally brilliant. The noise osc, as I mentioned for example, potentially adding some kind of percussion style rhythmic element. Which would admittedly be a pain to make use of if playing by hand. But with fancy sequencers like Cyclic around, just maybe it would come into its own.

    Ok, so failing that. If we did drop the noise & sub osc from the equation, and just used the 2 main oscillators.. Would it still sound good enough? I think on a Shruthi it might sound a bit weedy by that point. Perhaps if going purely duophonic, you could add the sub osc on 1 voice, and the noise on the 2nd voice to make things more interesting again.

    Anyway, although this was my suggestion, it’s starting to look awfully messy! I half regret mentioning it.
    Fortunately I love these little synths just the way they are. I’ll cope without poor mans polyphony ;)
    Wait! What am I talking about?? I have 4 of these suckers. I can polychain (Well, I would if I had 4 power bricks and a mixer that worked!)
    In hindsight, I’d much rather you directed your efforts into your new projects than this.

  • powers chords are possible now! Its the dual melody thing thats tricky. Another cool thing youd get out of a poly mode is that you can sequence some cool chip tracker style faux delays and echos.
  • @ Pichenettes: " This 3-voices thing with the sub sounds very dodgy – every third note this would play one octave below?"
    -If it would have those user definable keyboard zones which would allow changing root note, it would allow SUB OSC to be played at the same octave as others.

    @ Luap: "But just maybe with a bit of careful setting it up, it could be equally brilliant.", "The noise osc, as I mentioned for example, potentially adding some kind of percussion style rhythmic element. Which would admittedly be a pain to make use of if playing by hand. But with fancy sequencers like Cyclic around, just maybe it would come into its own."
    -Exactly. But if creating a track where sequence needs some notes to be played by certain OSC -like noise, how to define that on sequence start? Ordering those OSC's that way (round robin) wouldn't be useful very often. It would be fun from time to time, though.

    If those OSC's could be spread over the keyboard by splitting it to user configurable zones would do the trick. And it would be 3-voice if SUB OSC is included. Like i said that would require also changing root key one/two octaves up. And UI would need some new pages i guess..
    Two OSC poly would be better sound wise of course because SUB OSC doesn't have all those parameters that first two have.


    BTW. About your polychain setup. I recommend one of this babys, if you can find it:
    http://www.midi-classics.com/p5944.htm

    Originally i bought it for Evolver polychain but after i sold couple of those it got useless. Now when i have two Shruthi-1's and 3rd. is coming i can bring it back to business.

    If you think this conversation is a mess already, think about polychained poly-shruthi-1's and how those should be organized. :P
  • Here is a suggestion. I tried to come up with something really simple and musically interesting. No zones, menus, user-configurable stuff, too boring; It has very little to do with the duophony / paraphony thing, but maybe this is what you wanted after all…

    Preview here

    I call it “step mixing”. It shows up as a new mixing operator, and the only thing you have to do is enable it.

    In “2 steps” mode, the balance between oscillators 1 and 2 goes back in forth between osc 1 and osc 2 at each new note.

    In “4 steps” mode, the mix is the following: 1st note = sub is playing ; 2nd note = osc 1 is playing ; 3rd note = noise is playing ; 4th note = osc 2 is playing. It’s a kind of 4 steps wavesequencer, with the 1st step locked to a kick-like sound (you can put “pop” or “pl2”), and the third step to a noisy snare. Try it with the arpeggiator.

    In “8 steps” mode, it steps at each note through a kind of cheesy drum pattern (home organ style), using sub as a kick, noise as a snare, and alternating between osc 1 and 2 in between. Try it with the arpeggiator.

    Now try this on all the presets. Adjust the filter to make it more percussive. Free chachachipcore drum machine!

    Maybe I should make this controllable with the step sequencer?

  • to tired to try it right now, but sounds like a good idea.

  • Holy, I can’t get the kids in bed fast enough! Gotta try this!!!!!

  • "Maybe I should make this controllable with the step sequencer?"

    Interesting! If controllable with step sequencer would it mean that it could be possible to add, let's say, two noise notes in row, then sub, then noise again?
  • A couple of simple requests after all these interesting ones… Mod destinations for filter Envelope Amount, Filter LFO Amount as well as Modulation Amounts.
    thanks.

  • This is really cool! I think it sounds great :) The osc mixer seems to misbehave though? The mix, sub & noise controls largely don’t work. It depends on each of the 3 new modes.. Perhaps its by design? But personally I think it would be better (if possible) to control each element still.

    Damn fine evenings work though! I like it a lot..

    dr3tri, yes I’ll get me a mini mixer of some kind soon :)

  • Great times testing this out. Is there any way to keep the noise volume control in tact with step mixing? It on full blast right now and it’s a bit much.

  • > Mod destinations for filter Envelope Amount, Filter LFO Amount as well as Modulation Amounts.

    Use operators/modifiers. You can multiply a mod source by another one and use it as a new mod source, which achieves the same result as what you describe.

    Regarding levels for the new step thing: it totally takes control over the mixer, so it makes levels impossible to adjust.

  • I have just pushed a new version:

    • The sub and noise levels are taken into account.
    • A new option, “seqmix” allows the on-off state of oscillators/sub/noise to be taken over by the step sequencer.

    1 = osc 1
    2 = osc 2
    4 = sub
    8 = noise

    For example, basic drum pattern :

    4000804000408000

    These values can be added up for combinations (9 = osc 1 + noise, etc.)

  • This new addition is very nice,I am getting some very nice bass seq's
  • This all is very cool! No time to try any of it yet, but in re-reading the last few pages of posts, the one thing that I think would still be very interesting, and which would really open the Shruthi to lots of new possibilities, is the ability to trigger each of the 4 sound generating sections (osc1, osc2, sub and noise) to a distinct incoming MIDI channel…. Possible, or no?

  • or even just expand the trigger feature from earlier from note trigger to note +channel trigger, and add a destination to be the:

    1 = osc 1
    2 = osc 2
    4 = sub
    8 = noise

    Format?
  • >> Mod destinations for filter Envelope Amount, Filter LFO Amount as well as Modulation Amounts.
    >Use operators/modifiers. You can multiply a mod source by another one and use it as a new mod source, which achieves the same result as what you describe.
    Yep. That works fine. Thanks.
    New request: The mod. destination of ‘attack’ only works for me with faster attacks could this be updated to include slower attacks / negative values too?
    Thanks.

  • Am I being a chump? I don’t seem to have the new revision.. Im going here:
    http://mutable-instruments.net/static/firmware/
    And using the .96preview.syx version, just like last night. But it’s functionally the same as last nights build confuzzled
    No ‘seqmix’ addition to the mix operators here..

  • Also, this new mixer option brings me back to my previous desire to make the sequencer/clock/arp screen settings savable as a sequence preset because having to take notes and reprogram this isn’t fun and surely ain’t gonna fly in a live setting. I don’t think this is one of those firmware features that will sell you more units but I’d assume it’d greatly improve custom morale once they start really playing! :D

  • > This all is very cool! No time to try any of it yet, but in re-reading the last few pages of posts, the one thing that I think would still be very interesting, and which would really open the Shruthi to lots of new possibilities, is the ability to trigger each of the 4 sound generating sections (osc1, osc2, sub and noise) to a distinct incoming MIDI channel…. Possible, or no?

    Too complex I think. Would need a big overhaul of the code + some new menus.

    > or even just expand the trigger feature from earlier from note trigger to note +channel trigger, and add a destination to be the:

    It is already possible through the mod matrix.

    > New request: The mod. destination of ‘attack’ only works for me with faster attacks could this be updated to include slower attacks / negative values too?

    I don’t see how I can change this without breaking existing patches.

    > Am I being a chump? I don’t seem to have the new revision.. Im going here:

    Make sure that your browser hasn’t renamed the freshly downloaded file .96preview.syx (1), and that you are not trying to reload the older version :)

    > Also, this new mixer option brings me back to my previous desire to make the sequencer/clock/arp screen settings savable as a sequence preset because having to take notes and reprogram this isn’t fun and surely ain’t gonna fly in a live setting

    This requires a change to both the sequence format and the eeprom layout, and will cause an incompatibility between the old and new firmwares. I’ll add that to the list of things to change if I ever consider doing a major, backward-incompatible upgrade.

  • >> or even just expand the trigger feature from earlier from note trigger to note +channel trigger, and add a destination to be the:
    >It is already possible through the mod matrix.

    But arent you only limited to two triggers then? How would you trigger all four sound sources?
  • Ah, I see what you mean. Well, I think it’s time for me to learn to say no to your requests. I don’t want the Shruthi-1 firmware to become this kind of half-assed multitimbral mess. The code is already getting ugly and harder to maintain because of all those “hacks” – it’s not properly wired internally for all these polyphonic tricks and I don’t think it’s worth the effort to structure it more polyphonically. Also, I’m quite sure that if I start going down this road, you’ll be asking me in no time digital filters or digital VCAs to make each voice behave more independently of the others :)

    All this time shoehorning a drum machine or multitimbral sound module into a Shruthi could be spent developing, you know, the real thing :)

  • I totally agree with Olivier. Please let the Shruthi be as what its designed: a superb Hybrid Mono Synth. If you want Polyphony, get 2 ore more of them ;-)

  • Sorry, im a chiptune dude. Old voice stealing habits die hard! If its off the table, ill stop pestering about it :D
  • So from now on Voice stealing means someone stole your Shruthi ;-)

  • that being said, ive been live been playing with the seqeuncer and the mnew setting its and its super fun!
  • >> New request: The mod. destination of ‘attack’ only works for me with faster attacks could this be updated to include slower attacks / negative values too?
    >I don’t see how I can change this without breaking existing patches.

    Sorry, just to clarify… setting up a Modulator, for instance, as.. src:whl dst:atk amt:-63 will break existing patches? I looked and failed in finding the code for the Mod Matrix :(

  • >This requires a change to both the sequence format and the eeprom layout, and will cause an incompatibility between the old and new firmwares. I’ll add that to the list of things to change if I ever consider doing a major, backward-incompatible upgrade.

    Totally understandable! In the meantime, I’ll just have to be disciplined about keeping a Shruthi notebook/diary.

  • funkyfill: At first sight yes, because it will reduce by half the dynamic range of the modulation applied to the attack. Will give it a try tonight.

  • > Make sure that your browser hasn’t renamed the freshly downloaded file .96preview.syx (1), and that you are not trying to reload the older version :)

    I did actually check it wasn’t that. I really don’t know what the problem was though, because I tried it 4 or 5 times, each time making sure I wasn’t installing the older one. I think my browser was somehow caching and repeatedly giving me last nights version though.
    I just downloaded it on my old macbook instead, moved it to my usual iMac and installed it again with C6, and now I definitely have the newer version :) That caused much head scratching!

    I’ve not properly tried the seqmix thing yet (but at least it is there now!). But it is already much better now that the mixer controls work much as I expected them too.

    This is a really cool addition to the firmware! So a big thanks and a thumbs up from me. I think it is genuinely better than any take on poor mans polyphony :)

  • Oooh, can’t wait to try this. That said I totally agree with keeping development focused on the Super Hybrid Monosynth.

    I’ve made more than my share of requests so I’m just going to throw some more ideas out there without any expectation whatsoever:

    - Clock step sequencer by LFO rather than tempo/clock (would be cool if there was a way to use the step sequencer at a different rate from the arp)

    - Tempo clock as a destination in the mod matrix (esp. cool if you could still send the modulated tempo via MIDI clock – imagine pressing one of the Trigger keys and making a slaved drum machine suddenly speed up and gradually slow down!)

    - Increase modulation range for LFO destinations – I was playing with this over the weekend and it’s not quite freaky enough for my taste. With the full range I can go from Pretty Fast to Pretty Slow OK, but I can’t go from Super Fast to Super Slow like I want. I tried using 2 slots in the matrix to double the mod amount but that didn’t seem to work.

  • > imagine pressing one of the Trigger keys and making a slaved drum machine suddenly speed up and gradually slow down

    First, it will require a big overhaul of the sequencer / clocking code. At the moment it is working at 4 ppqn so it cannot send a MIDI clock (24 ppqn).

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

In this Discussion