Norns 2.0 is excellent so far. I finally have a working wifi setup along with a lot more storage space.

I think a major point of friction on Norns at the moment is sample management, especially for the sketches with more than one sample file. Here’s the process:

  • Exit the PLAY view and go to the EDIT view
  • Dial over to the parameters page
  • Dial down to the first sample file parameter
  • Transition to the file dialog, find your sample.
  • After selecting the file, scroll down to the next sample slot in a fairly fiddly manner (depending on how the params page is organized)
  • Transition back to the file dialog, where you are reset to the top (!) of the tree.
  • repeat, repeat, repeat

While you can then save this as a pset file, you then have to do this over and over again for each sketch.

Here are some generalized ideas based on how other gear handles this.

Sample Pool
Elektron boxes, ER-301, and Assimil8or to an extent have this, with ER-301 being an especially relevant example about how to pull this off with a minimal interface. The Assimil8or is the simplest: pick a sample folder and the whole thing gets loaded (unless there’s not enough RAM, in which case it just loads whatever it can into the pool). On Elektron boxes and the ER-301, you can either load one file at a time (like Norns), or select multiple files to load into the pool. In the latter case, while scrolling through the file list, you can press an alternate button to mark the file for loading. You then press the other buttons to commit the files to the pool.
With an active pool, the parameter page on Norns could avoid the file loading dialog altogether. Instead, you could scroll to a sample slot, and then rotate quickly through files 1, 2, 3… etc. This is even better for those of us who would want to map this to MIDI so that we can change samples mid-performance.

(EDIT: I imagine the sample pool would be another tab in the EDIT mode set. It would persist between sketches but would be flushed on Sleep. Additionally, there would be options to save/load/clear the entire pool)

Sample Kits
This is another option on Elektron boxes. You can pick a pre-mapped kit that will assign 8 samples (on the Digitakt), one onto each specific track. I’m imagining a way where we could create Norns kits, then loading a kit will add the kit’s samples to a pool in sequential order (Perhaps a kit could be saved from a pool…?).
Using this method, I could create an 8-sample kit in Glut, go over to Playfair, and load the same kit into the active pool, thus auto-filling the 8 param slots on Playfair.

15 Likes

I’ll look into this a bit more when I have a touch more time.

I did some sample preset things awhile back with the x0x samples and had the intent to write a “load folder of samples” script. I gotta did that stuff up and see what I can do with it again.

1 Like

I don’t have the knowledge to implement this… but I hope someone takes a crack at it. I’d love to just plug in my Blue and record samples into MLR live…

1 Like

Agree folder loading is very useful, I have this implemented in an (unfinished) script:

27 Likes

would you be cool with posting the folder loading bits to a gist or something? I’d love to take a look at that.

2 Likes

will probably wait until it’s more finished/tested

2 Likes

Thank you for sharing this playlist

I sat down 2 hours ago to begin ‘Albers’ and instead have the beginnings of ‘Smalltalk’, based roughly on David Behrman’s computer programs (ala Leapday Night). very interested in exploring systems based on musical input/interaction and all of the work put into norns makes it very easy to get results.

7 Likes

dreaming of a nice-sounding solina-ish string synth engine.

(truth be told i’m still barely capable at supercollider…)

7 Likes

An idea I have (I thought about it as a long-term project for me but if someone likes it and wants to tackle it I’d be grateful) is an Euclidean drum machine like playfair or foulplay, but with pinged filters. I know there’s good filters to ping in sc. I’d love to have an array of tunable filters pinged rhythmically.

6 Likes

you can (sortof) actually do this now— i did it yesterday. loaded a 606 hihat sample (included in dust/audio) and turn filter Q way up. sounds great.

edit: works well with Q in the 90’s, speed turned up to 500%, cutoff at desired freq

5 Likes

First thing I’ll do tomorrow morning for sure.

function-alike, sure, even PolySub is not so far already

circuit emulation, no.

if there’s one thing norns is not, it’s an analog synthesizer.

13 Likes

/offtopic but this overview has made me think about a Deckard’s Dream pretty much daily since I watched it. It’s a crazy amount of money but it sounds sooooo nice :drooling_face:

1 Like

Same here. :star_struck:
I may end up figuring out the DIY cost and see about building one.

1 Like

BTW: in softcut 2.0, record + playback at <1.0 rate is effectively equivalent to sample rate reduction with cubic resampling.

there is an option for linear resampling in the code (and of course zero-hold resampling is pretty trivial.) lower-order options are not currently enabled for runtime selection. maybe in a future update. (is this appealing?)

bitcrushing is another option. linear remapping is trivial; but i suppose mu-law mapping is important for getting that “12-bit sound.” (is static companding an important component of that?)

an immediate update i’d like to add for that subsystem is post-resampling SVF (in addition to the existing pre-record SVF, which allows a different set of possibilities.) post-filtering makes distorting processes more useful i think. that’s not a big deal to add, it’s just that there are already a lot of new parameters in 2.0. so currentlty, we’re leaving additional baked-in FX options for, like, 2.1, simply for usability reasons.

8 Likes

re: sample/bit rate reduction, something I find incredibly useful is to output the residual (f(x)-x)
and then give the user control over the wet/dry mix. then you can control the amount of grit, or send the residual to a separate EQ before mixing it back in. In terms of commercial products I had an old TC electronics processor (I think Fireworx) which allowed for this, otherwise I haven’t seen it. But it is very useful.

2 Likes

I got my Norns last week and the first thing I’ve done is have a play with Cranes (guitar straight into Norns). Beautiful, intuitive script (thanks @dan_derks)

My wishlist and ideas would be:

  • Integrate some form of ‘record/overdub’ enable for either buffer (so independent material can be recorded to either buffer).
  • Independent shift/wobble (Key 3) values, and enable, per buffer.
  • Simple forward/reverse control per buffer (with speed as independent parameter).
  • Panning per buffer (so both buffers could be routed independently via the L+R outputs for further processing).

These are perhaps things that would be a good idea for my first Norns coding endeavor (I have basic/rusty grasp of C++)?

Or better to wait until Norns 2 is released before delving into the coding side?

3 Likes

excellent ideas, @Ithacus, definitely all on the roadmap :slight_smile:

cranes has not yet been updated to norns 2.0, but i’d say upgrading for the new softcut is 100% worth it, so it might be a really cool exercise to create your own looper and hack out the stuff from cranes you do like — at least until i can get some time set aside.

2 Likes

Hey thanks Dan. Sounds like a good idea!

The stuff I’ve listed is pretty simple, so could be doable for me using your existing code (only one way to find out :grin:).

Quick question: for Midi CC mapping to parameters, is this a system level function (so doesn’t need configuring in script code)?

1 Like