Hardware version of mlr

  • Lately, I have really wanted to get away from using the computer to make music and get back to some straight hardware sequencer/sampler roots because I generally find hardware a lot more fun, however I just can't get away from certain apps, like mlr.

    I know there has been discussion of this before, but has anyone thought of making a hardware only version of mlr?

    I am looking at a ton of samplers and grooveboxes out there, but none of them quite match the hypersampling instrument.

    has anyone thought of porting mlr to some small mobile computer or something where it ran as a complete sampling solution without a computer? I would love to be able to take mlr anywhere without a laptop, or a cables or any other crap that gets in the way.

    btw,

    what is the lowest spec'd machine that can run mlr effectively?

  • a portable computer is still a computer though right?

  • >...ton of samplers and grooveboxes out there, but none of them quite match the hypersampling instrument.

    yes, time to wake up !
    I guess you have to reinvent the wheel first...

  • If a monome and an MPC had a baby............

  • Or a monome and an Octatrack.

  • Oooh, this needs to materialize!

    How big a step up from the bliptronic5000 would this be? Can't be THAT hard for the people in the business can it?

    Line in, sample/preset storage on mini SD, little speaker, perhaps some cheap delay/reverb chip..Ahh, my life would be complete.

    EDIT: Anyone tried running MLRV (2.2) on one of those Asus EEPC's?

  • tried renoise on a lenovo netbook. no joy. just a short google of 'max msp netbook' shows difficulties... sampling is fairly load intensive in software.

    likewise most tablets, unless running something like i5 processor.

  • chipPad gets you part of the way there...

    http://earsmack.com/

  • Mmmyeah, chipPad is very nice and all, I just can't getting used to tapping on a glass plate..Also music performance on an iphone/pad just looks retarded.

    But hey, that's just my opinion.

  • the only way i can think of doing this with hardware that im aware of would be to do an OS hack on MPC1000, but i have no idea on how to go about it. the only custom OS i know of is the JJOS series but to my knoweledge it doesnt do MLR follow and jump to actions, but it might be a good starting point to open up one of those and take a look. But honestly seems like way more trouble than its worth

  • exactly. just hide the computer under the table. how would you even set up mlr without a screen?

  • well see if it was hardware, it would be easy to setup. Make the app have two pages, the first default page on startup. You might have to memorize the locations of a few buttons (no labels) but still we do that anyways. Someone needs to write mlr in a different, more resource friendly language like C++ or assembly :), so it could even run on a super cheapo linux tablet or something. I mean sampling is that processor intensive, but resource hogs like max/msp are. Think of the samplers of the old days, many of those ran on motorola 68ks and I had a fairly good sampling program for my atari falcon and that was a 68020.

    It's all about careful use of resources.

    who wants to start coding?

  • actually, does anyone know of any samplers that have functions like mlr? or is mlr a unique beast with live sample slicing?

  • I guess if you were really good at embedded systems, you could make an arduino, or even more extreme, straight atmel version of mlr, which could fit inside an arduinome. You'd need to make some sort of interface for it as well, but I see you've already thought about that

  • my project [[https://github.com/wrl/rove|rove]] is a rewrite of MLR in C. it's very light on resources and doesn't need serialosc/monomeserial to run (i.e. it can interact directly with the serial device).

    it should run comfortably on a beagleboard or pandaboard.

  • i've been wanting to do exactly this for quite some time now.

    the only problem is, i spend all day coding at work, and get home, and don't want to code anymore.

    that may change soon though..


    they way i'm envisioning the design in my head is something like a korg electribe or mpc, but with a usb host that speaks monome. bare-bones OS that just handles sample management (no linux) running on a modest microcontroller.

    stereo out, samples on sd card.

    is a kickstarter in order?

  • i was going to mention rove, too. the nice thing about it is that it's written in C, so it should work as-is or be easily portable to ARM-based devices. it also doesn't have a gui the way max-based patches like mlr do, so it's even easier to script. i don't think you'd need to modify the code at all. just add some startup scripts to launch your rove sessions from an sd card.

    bonus: you don't need serialosc, and JACK is already ported and working on ARM; just look at all the stuff that ubuntu packages.

    i've been kicking around ideas on running rove from a purpose-built tiny, portable machine, either an android/arm tablet, or something similar. make it small enough, stick a battery in it, add a case/chassis for your monome, and you have a self-contained unit. at some point, though, depending on what you want to do with it, it's probably easier to just use a laptop. if you don't need onboard dsp/fx, and don't need to run more than one app, then arm, arduino and similar microcontrollers are good places to start.

    @soundcyst:

    you'd have better luck running some kind of tiny embedded linux on your arm board, at least if you want to run existing apps like rove. openembedded, angstrom, gentoo, and a host of other distros are all excellent starting points for providing tiny, tiny minimal environments. especially if you're just going to take the scriptable startup approach.

  • wow, I am glad that some people have been thinking about this too.

    The idea of an mpc with usb port is cool, but I really dig the portability of the monome, just not the cables. It would be really nice if it was more tenori/bliptronic geared where it could be battery powered. I mean the shruthi 1 is a perfect example of the pure power of an atmega. Imagine using something like a dsPIC or arm processor and you would be killing it.

    It probably wouldn't be a bad idea for some dsp effects, but I was thinking far simpler than that. No sample storage, no effects (maybe multiple outs because I use a mixer) just a grid, a mic or stereo input and a headphone and stereo output. Basically just for live usage or jamming. I would totally start a kickstarter, but I have no coding experience, just hardware design and I can make pretty enclosures and solder well.

    has anyone compiled rove for macos? I would like to try it out as a standalone.

  • i like this thread and cannot wait to see where this all leads.

  • okay, here is where i ask the community for their input,

    what features do you want in a hardware mlr?

    what layout do you think would work (2 row LCD, simple grid interface, more complex stuff like multitouch displays etc)

    do we want effects, and what kinds?

    what are some good embedded platforms or cheap older computers that we could run it off of?

    I am really hoping this will take off into something!

    let's come together monome community!

  • one thing to keep in mind: as you add complexity, you get closer and closer to the point where you might as well just hook up your monome to a laptop, as usual.

    i've been contemplating the idea of a minimal rove/mlr controller, on android/arduino devices, for a few days now, even before i saw the news about google's usb host mode code drop for android.

    simplicity is a primary goal, as is portability. even to the point of just chucking a dedicated microcontroller board with an sd slot into the same enclosure as the monome, along with a battery pack. add some custom logic for startup and set loading, and you have a completely self-contained device.

    as soon as you start adding onboard fx or additional controls, you stray closer and closer to needing a real computer with more substantial capabilities. you'll need more horsepower, thicker software stacks, and more ways to tweak it live . . . meaning a full-fledged computer, no matter how small it actually is. for example, if you're after onboard dsp, then you have to write your own sound stack, if you're not going with an existing embedded solution, such as linux.

    also, the more controls you add, the more you take away from the simplicity of the thing. yeah, all-in-one can be nice, so that you don't have the usual tangle of cables from four offboard controllers, but you also clutter your interface. at this point, you have a livid block or ohm; at the very least one of these: https://filib.io/posts/2012-03-03-arduinome.html

    the fewer additions to the existing grid UI, the better. otherwise you'll end up with a weird apc/mpd/block/manta/ribbon hybrid. it may suit your needs as an all-in-one device. it may even be battery powered. but the same things can be accomplished with off-the-shelf hardware and existing software. back to square one.

    i encourage you to think about what, exactly, your need is for a dedicated mlr device. some things to ponder:

    - what do you want it to do?
    - does it need to run more than one app?
    - how can it be configured?
    - what advantages and disadvantages does it have over the usual laptop + monome combination? what can't you do with the usual hardware vs. what would this allow you to do, and vice versa.
    - how much would it cost to build?
    - how long would it take to build?
    - how would you interact with it, and in what settings? studio, live performance, etc.

  • I am thinking along the lines you are,
    simplicity is a must. I would rather memorize button presses than have any LCD, or any of that crap. The whole point of the monome is simplicity; not letting all the other bs get in the way of making actual music.

    I was also journaling some things about this last night as I do all my ridiculous ideas and I kinda put together a paged interface, with the paging operated by the top row of buttons much like on the computer. It would allow paging through to volume parameters, a primitive sample loader (kinda like in loopy), pattern recording, resampling and cutting parameters. Real simple. I will try to make a mock up and see what everyone thinks.

    It might actually be more limited than mlr, but I have another question:

    What features do you use in mlr? What features could be cut or simplified?

  • anything ever come of this?

  • don't think so. i wonder would the new raspberry "pc on a keychain" be of use for something like this:

    http://techcrunch.com/2011/05/05/raspberry-pi-a-25-bare-bones-pc-that-fits-on-your-keychain/

  • @loflow - excellent to read and feel how you think and how you answer - I mean it's really appreciable to see how the knowledge stimulates - I don't understsand really the sens of this but on one hand I feel the interest to understand and on other side it's sure....to minimize the hardware allows you to travel more easely and not only on the stage - Good luck n8bit!

  • http://ohpeewon.com/discussion/293/monomep-1
    I brought this up on an OP-1 Forum. The OP-1 is a breeze to use and almost all of the clips I load into MLR come from my OP-1. The color coded knob system is amazing intuitive... so I photoshopped my dream...