Why such limited MIDI control for Obsidian?

I'm a keyboard player. Obsidian is one of the few ios synths I would use in a live performance: It's never ever crashed on me, it's insanely versatile and sounds incredible, and it's an absolute joy to program.

Yet, when controlling it from a midi keyboard it's surprisingly limited.
As it stands, it's more or less on an island floating in the paradise of NS2.

Aside from the glaring lack of program change, which is a dealbreaker in its own right, the limit of 10 mappable parameters is sort of astonishing. Slate has over 20, which is certainly a bit better (though I'd love to be able to control filters, tuning for individual samples, etc from my launchcontrol, that would be good fun!)

Anyhow... Is there a reason for the paltry sum of assignable knobs in Obsidian? My midi keyboard alone has 113 midi knobs, sliders, and pads that are begging to control as many elements as possible. I realize that there are 8 knobs and an X/Y pad, thus the 10 assignment slots, but I can't help but wonder why midi can't be mapped to cc for using external controllers.

Comments

  • edited July 2019

    Reason is that this would trigger chain reaction mostly in UI area :) Adding possibility to map any of Obsidian parameters will need also wide rework of automation UI (obviously everybody will expect that what is mappable, is also automatable) - imagine that in current automation editor, instead of 10 automation lanes would be hundreds :) It would need complete change of automation editor UI, it would need to have it empy by default and just add manually lanes for parameters you want draw automations (or which already contains some automation) with some submenu system for easy search within plethora of Obsidian's parameters - instead of finding some particular parameter to add from that huge list would be pita.

    I agree that it would be great to automate directly any parameter (not just Obsidian, also in AU) and i was fighting for this for a while during developement, but then i understood all consequences and huge amount of work needed in UI area for making this possible..

    On other side, maybe allowing map HW controller directly to opObsidian knobs with some quick and easy mapping method based on same UI like quick modulation assigning, but without any changes in automation area - so it would serve JUST for realtime controlling of synth params - would be doable, hard to say.

    But i'm pretty sure in moment when this would be possible, inevitable request would arrive "i wamnt to record automations of all those mapped params" :)) So i'm not sure Matt would want to explore this rabbit hole :)

    Anywaty, we will see what priorities will choose mr. Borstel for roadmap after iPhone, audio tracks and aufx automation ... There is work for 5 people for next 10 years on todo list, so...

  • edited July 2019

    @dendy said:
    obviously everybody will expect that what is mappable, is also automatable

    I think it is entirely reasonable to limit the automation lanes to a manageable number (say, 10). But automation and mapping are two entirely different things. To hinder the expansion to midi gear for fear that people would expect these cc's to be automateable.. that's just sort of weird.

    Having parameters exposed but with 10 automation lanes for the macros would be a simple and elegant solution imho.

    I think unless you're a keyboardist, it's hard to see just how close yet so far it is from being a performance-ready instrument. With these fairly standard midi controls (program change, cc control), NS2 would replace my Nord for stage performance in a heartbeat. But without , it's stuck in the studio. And it's a great studio tool, but it's the fact that it's soo close. I've been hoping for a really killer, stable synth to come along that was a serious, roadworthy instrument.. and it's a hair's width away from stage ready.

  • yes i see your point.. as i said, without automation, just direct midi->parameter mapping for realtime control.. hm.. i can imagine that if this would be possible i would build customized DIY midi controller with knob layout exactly based on Obsidian edit page - just for sound design phase :)

  • @dendy said:
    i can imagine that if this would be possible i would build customized DIY midi controller with knob layout exactly based on Obsidian edit page - just for sound design phase :)

    Exactly.. =) I do this with all my favorite synths and it brings both programming and performing to a new level.

    Funny thing is that so many hardware heads bash on software because they "like the tactile control of knobs and faders".. yeah, so map it to a midi controller and there you have it. Best of both worlds.

    I love hardware and software, and I love it best when they get along well :-)

  • anyhow, please pass this along to the head honcho? I know there's a loong list of things, but man... this is just one of those essential things for those of us who are working with controllers. If people take note that NS2 could replace their reliance on a Nord or a macbook pro for a serious live performance instrument, it could open up a broader market than already exists.

    As I've said before, I love the attention to detail that goes into NS2. I can tell that it was made by someone who makes music and loves what they do. (i can also tell it was made by someone who doesn't play keys haha! ;))

    These seemingly small details can make such a huge difference. I simply wouldn't play any synth on stage where I had to dive into a menu to do something as simple as changing a patch.. so PC messages are absolutely essential. And to be out of controls after mapping filter and envelopes... well it's a bit too limited for my own needs. So much potential waiting to be tapped. It's all right there, but without cc's and PC's..

    I'll happily keep chipping away at it and working with it in-studio.. but for reasons of portability and sonic abilities, I'd truly love to take Obsidian on the road with me! Perhaps one day..

  • edited July 2019

    .. oops - double post, sorry!

  • nyhow, please pass this along to the head honcho?

    definitelly will do, as i said i already bothered matt with this during developement and will do it again :lol:

  • @palm said:
    .... If people take note that NS2 could replace their reliance on a Nord or a macbook pro for a serious live performance instrument, it could open up a broader market than already exists.
    So much potential waiting to be tapped. It's all right there, but without cc's and PC's..

    +1
    I wholeheartedly agree.
    The obstacles are not enough time & a very long To Do List.
    Let’s keep our fingers crossed. Except when we are playing...

  • @dendy said:

    nyhow, please pass this along to the head honcho?

    definitelly will do, as i said i already bothered matt with this during developement and will do it again :lol:

    thanks @dendy!

    @SlapHappy said:

    @palm said:
    .... If people take note that NS2 could replace their reliance on a Nord or a macbook pro for a serious live performance instrument, it could open up a broader market than already exists.
    So much potential waiting to be tapped. It's all right there, but without cc's and PC's..

    +1
    I wholeheartedly agree.
    The obstacles are not enough time & a very long To Do List.
    Let’s keep our fingers crossed. Except when we are playing...

    Ha, yeah! let's hope this isn't so low on the list that it takes years to get there :#

  • edited July 2019

    Shh. You’re not supposed to say that out loud. 😳
    ...and definitly don’t say it three time in a row. 😄

    I share that hope as well. I would guess not to see better MIDI implementation this year, since there are already so many stated big goals, but I have my hopes that next year could see some development on this. I better make some offerings to the MIDI gods... 🗿👺👹

  • I really value being able to assign any CC to any knob when using virtual instruments, and would love to see this implemented in Obsidian, but doesn't the modular nature of obsidian (with different params for different modules) make universal Obsidian mapping next to impossible?

  • edited July 2019

    Possible compromise briefly discussed during beta: use a set of fixed MIDI CCs to control Obsidian parameters. Then you could 1) map a hardware controller to any OBS parameters (by programming the unit itself) and 2) use additional MIDI Instrument tracks if you wanted to automate any parameters above 10.

    That is, nothing in the UI changes. If you want to automate Obsidian, you'd have two options.
    1) the current macro set up
    2) one or more MIDI instrument tracks with the correct CCs set in its macro screen, pointed at the obsidian track (10 CCs per track)

  • Maybe worth a mention... I was probably the most vocal about wanting a 1:1 for Obsidian controls during beta but the more I used Obsidian, the more I realized how cumbersome this would be in practice. Obsidian just has a metric shit ton of parameters!

    Certainly more parameters than 120 available MIDI CCs would cover. For instance, each LFO has 10 parameters, not including the 'amount' set per mod matrix lane (13 if tempo based time adjustments should be available... one to toggle it on, two for the new controls). That's 50 CCs for the LFO section alone. Or 65 if tempo based controls are included. Or 90 if not ignoring the mod-matrix amount (which can be for up to 5 destinations per LFO!)

    image

    Similar for the Envelopes. If you were to include the Scaling screen parameters and segment curve adjustments, that's 14 per ENV. Not including mod-matrix amounts.

    So that's 120 CCs (on the light end), without even getting to the Oscillators (holy FM controls), Filters or Effects.

    Suddenly, we're in NRPN/Sysex territory and, well, hurrah for simple mobile music making apps.

    I still want more CC control over Obsidian but yeah, as Dendy alluded to, it's not a trivial ask.

  • edited July 2019

    yeah total amount all parameters is incredible so make them ALL CC assigned automatically is impossible ..

    but i think most uses would need simply map visible knobs same way like modulations are mapped (tap on knob, tweak kbob on controller, mapped) - because also hw controllers have just limiter number of physical knobs :)

  • yeah exactly.. I think there’s no need to over complicate this. no need for a cc for every single parameter in the entire universe.

    simple midi learn would do the trick. and could work just like assignments do - from the main panel, tap midi learn, twiddle a controller knob.. and bam.
    if you change the oscillator to something with different parameters, the previous learned parameter disappears. mapping is hopefully stored per patch, just as other assignments are.

  • @palm
    no need for a cc for every single parameter in the entire universe.

    Actually, this would be f*cking awesome, it would make possible to affect anything will happen in future :lol:

  • @dendy said:

    @palm
    no need for a cc for every single parameter in the entire universe.

    Actually, this would be f*cking awesome, it would make possible to affect anything will happen in future :lol:

    and past!

  • haha.. oh you guys...

    @WinConway said:
    Note also that the AUfx automation MUST be recordable and editable in the part automation and automation parts, being track based is terrible for sound design, if you have ever made a complex Neuro you will know why ;)

    I agree with everything you said, but I can't help but wonder.. what's a complex Neuro?

  • edited August 2019

    @palm
    for briefidea check all "neuro" and "reese" Obsidian factory patches.. there is also demo track "Bite Me" using one of those patches

    There are various alternatives of this type of bass.. for example in patch "Oh Oh Bass" i was trying to reproduce bass from track "Noisia - Oh Oh" - of course it's far away from perfect match :)))

    Neuro bass is one of most complex synth sound design challenges...

  • @WinConway said:
    I can't agree with any of this, every other host does this with zero issues.
    You speak about the problem with the UI, there is no problem, just implement last touched ala Reaper, so similar to the current method of touching a dial to assign it to an LFO etc, use the exact same UI to select a knob and have that send automation data for recording and/or be in the automation editor for drawing automation.

    If only it were as simple as "just implement last touched ala Reaper." That's saying "just do this thing that a fairly large group of developers working on a desktop DAW for well over a decade have done." Considering the number (and quality) of features already in NS2, it's probably fair to presume that if this was a 'just implement...' sort of situation, it'd already be in there, right?

    I'd love to see 1:1 for automation too. I'd guess Matt would be stoked to wake up and see it there as well! There are only so many hours in a day though.

  • So what we need is a new app that increases the number of hours in a day. I’d buy that!

  • edited August 2019

    So what we need is a new app that increases the number of hours in a day

    I would prefer app which increases number of hours in the night, so i can sleep more...

Sign In or Register to comment.