randy's Recent Posts

by Randy Jones
photos: glia

Glia makes a very refined music that explores what looping and synthesis can do in a jazz context. Though Fela Kuti's funk is one of many influences he cites, there are no freight-train sized grooves here—think instead of an intimate Blue Note or ECM trio with acoustic players and friendly AIs.

A drummer first, glia has shifted more and more creative effort towards electronic music over time. His latest album, OUANEM, ties together a variety of fruitful collaborations recorded over the last two years into a cohesive whole. It's very deliberate, but far from sterile. It's enigmatic and runs counter to so many commercial trends of this time. To me it sounds like the jazz that future hiphop will sample.

I got to ask glia some questions.

How long have you been making music as glia?

I adopted the name back in 2008 as I more fully embraced electronic music and improvisation. Prior to that I spent a few years learning the basics of production, beatmaking, and recording but considered myself a drummer first and foremost.

One thing that strikes me about OUANEM is how it's short for an album, but very condensed. Like each track was distilled down into its essence. Is "album" a meaningful category for you or are you reaching for something new, here?

You are very observant...I can agree that finding the essence of a musical idea intrigues me. It's no accident that many of the tracks feel condensed but I'm not sure it was totally deliberate. I guess part of the reason they turned out this way is that I've always liked playing with scale/duration/concision in my art and find that, as a listener, I'm often drawn to repeat-listens of the same song. Making recordings brief is one way to encourage my audience to do the same & savor or study a tune more than once.

Albums do hold significance for me and I consider this to be a proper collection in my loose definition of the term (though I know ONM might seem too short...more like an EP for others).

The short nature of the tracks also serve as a personal invitation to revisit, extend and reinterpret them on future albums or live performances.

There are a number of other performers listed. Can you talk a little about how the collaboration works?

Each collaboration was unique. The common thread linking them all was remote file exchange...all the recordings and edits took place virtually with plenty of trust and communication to ensure we were happy with the completed songs.

I usually tried to build a rhythmic foundation before getting other folks involved and just followed an intuitive flow to decide what to do next or how to adjust the parts depending on what they contributed. There were, of course, a few exceptions (Chris Hill and Logan Kane sent me things to build around or rearrange freely)

What does building one of these tracks look like?

I recorded acoustic percussion or drums as a core rhythm. Then looped those parts and layered takes with more percussion, digital synths and woodwinds in ableton. I tend to use many fx but the arrangement or structural workflow is based heavily on the simplicity of "tape" recording.

How do you know when a track is done?

When I sit with a track for a few weeks or months and "hear" nothing else. Sometimes I know as soon as I finish recording but most of the songs on Ouanem were works-in-progress for all 1- 2 years. If I could listen to a piece after that period of reflection and still see beauty in it then I know for certain it's done...otherwise it might need things slightly adjusted (or might not be worth releasing)

What do you do when you’re not making music?

Think about music! My answer is surely cliché but I spend very little time actually making music. Studying language and religion is my main diversion outside of music. I also enjoy cooking, sports and watching films.

I've met you through the lines forum- (llllllll.co) --are there other online or physical communities that sustain your musical world?

The pandemic has definitely made me more cautious about connecting in person but I really love seeing what High Zero/Sonic Circuits and Rhizome in DC have done over the years to involve the community. NPR Tiny Desk Concerts are another local favorite of mine and it's crazy to see how their popularity has grown and spread worldwide!

The people behind Experimental Sound Studio in Chicago also come to mind, especially for promoting really incredible shows across the full spectrum of improvised music during lockdown. I'm a longtime fan of Dublab and felt immediate kinship with ESS heads cause they seem to cover a similarly diverse range of musical interests.

I feel a special connection to all the artists on past Smartbomb compilations and the creative team who worked on Video Home System the past few years. Also hard to ignore the positive impact certain labels have had on my music...along with Minaret I'm indebted to friends at Leaving, Phinery and the many artists associated with Dolfin, Eglo & Touching Bass. They've found unique ways to fill virtual spaces with the same beauty and intention their in-person events are known for...so I love checking archived shows and music videos for inspiration.

https://ifizu.bandcamp.com

heard!

Ok here's one more:
https://discord.gg/4VzcjxPS

the links expire after 7 days. here's a new one!
https://discord.gg/p5XXD9D3

ok! https://discord.gg/p5XXD9D3

I'm sending out a newsletter soon with more info, as soon as I can finish this sub-project (Utu) I'm working on. If you're interested in learning more about that in the meantime, there's more info on the Discord.

https://discord.gg/peZN5V4D

yes!

Oh I see. I'll try to figure out some reasonable parameters for the upgrade discount but I'm not sure exactly what those will be.

There will be a discounted upgrade price for Aalto (1) owners. Or if you bought Aalto very recently, Aalto 2 will be free.

Aalto has been out for over ten years with only free upgrades...

here you go:
https://discord.gg/RPCJBzEf

When Sumu is beta I'll start working also on the other updates, starting with Aalto. I'll feel way behind if they're not out before the end of this year. But honestly there's only one of me and I've managed to get way behind my plans on Sumu already, so take this with a grain of salt.

All the major updates will move from VST2 to VST3 as well. Aaltoverb is VST3 + AU right now.

When you put Aalto presets into the “Aalto/MIDI Programs” folder you can load them by sending a MIDI Program Change message to Aalto. As installed there are a few example presets “test1”–“test4”. They will be assigned numbers in alphabetical order so Program Change 1: Test1, 2: test2 and so on. This should work in every host because the host does nothing but send the MIDI message.

Here you go!
https://discord.gg/g8e8Quxy

I got your emails also. I'll copy my response here for reference:

Hi Davide,

I never was able to use the VST2 format to store the patch bay data because I could not find a good way to do this. VST just stores values for single parameters like dials and this does not map well to a patch input, that can have any number of outputs. Bidule only accesses Aalto through the VST2 API and so can not access the patch bay.

I am somewhat in touch with the Bidule folks now. I will see if there’s a way for more complete automation of the patch bay using VST3 or another format that Bidule supports.

Thanks for the report. I just tried to reproduce, finally with success, I think. I could not get this to happen until I changed the display resolution for the external monitor to a higher one. With the laptop at "default for display" and the external monitor at "More Space" (max resolution) I get weird behavior like you describe.

I'll look for a fix ASAP!

Thanks for the encouragement! It is truly helpful.

You can sign up to the beta Discord if you like:
https://discord.gg/KFwJa5Dp
(link good for 7 days)

It's possible to compile utu (Sumu analysis tool) from the instructions in the github repo, however, it's only usable from the command line right now.

There's nothing you need to do. On any Apple Silicon Mac, just install the plugin and the system will use the Rosetta 2 emulator to run it.

the current versions work in Apple Silicon under emulation. After inatalling, can you find the plugins under Audio Units? If not, restarting may be needed.

If restart does not work, try clearing your Audio Units cache:
https://madronalabs.com/topics/8791-how-to-clear-your-audio-unit-cache

version 1.8.3
April 23, 2019
link: Soundplane 1.8.3.dmg, 6MB

release notes:

To install this version: move the "Soundplane" application to your Applications folder, then move the directory "ZonePresets" to ~/Music/Madrona Labs/Soundplane.

When you launch the Soundplane application, the menu at the top of the "Zones" page should contain three presets: chromatic, rows in fourths and rows in octaves, followed by a list of all the JSON files in the ZonePresets folder. If the JSON files do not appear, the ZonePresets directory is probably not in the right place.

The format of the zone .json files is human-readable and relatively self-explaining. Take a look to see how to make your own zone maps.

The Soundplane Max/MSP examples require the CNMAT Max externals to run. These can be found at http://cnmat.berkeley.edu/downloads .

The Soundplane application is open-source, available under a permissive license. For source code and more technical information see http://github.com/madronalabs.

changes:

1.8.3:

• fixed issue with excessive CPU use / repainting since 1.8.0

1.8.0:

• cmake build rewritten for compatibility with new madronalib and soundplanelib.
• fixed graphics incompatibilities with OS X Mojave
• fixed controller Zone outputs and reduced redundant data output

1.7.0:

• fixed possible crash starting for the first time with Kyma connected
• "destination port" menu now allows connecting to other hosts on the local network
• added wait on startup to resolve ports on other hosts
• added clear all notes when switching OSC destinations and on shutdown
• added test pattern mode
• allow running app hidden / in background without added latency
• fix redundant calibrate on startup
• fix a stuck notes issue
• add quick recovery after unplug / replug instrument

1.6:

• improved isochronous USB driver: lower latency, allows restart after pausing app
• fixed an issue selecting cleanest carriers after "select carriers"
• fixed possible crashes on startup/shutdown
• now distributing as signed .dmg

1.5:

• new touch tracker algorithm:
  • consumes much less CPU
  • improved latency
  • improved pressure sensitivity
  • improved pressure uniformity
  • improved position accuracy
  • allows better tracking into corners
  • lengthy normalization step no longer required
  • fixed hanging touches
• fix Kyma connection
• improve selection of lowest-noise carrier set
• fix a possible crash when switching zone presets
• fixed a window-related crash on shutdown
• fixed latency issue when in background for an extended time

1.4:

• fixed a problem with MIDI output preventing slide between rows
• tweaked pre-touch filtering for lower noise
• (dev) new cmake-based build system
• (dev) experimental Linux build
• (dev) added HelloSoundplane command line app for low-level testing

1.3:

• MPE MIDI support. Requires Aalto 1.7 / Kaivo 1.2.
• Implemented note splits to multiple ports over OSC.
• Improved stability and sensitivity of MIDI velocity
• changed "z max" control to more intuitive "z scale."
• fix crash with uninitialized driver on shutdown
• allow 1-500Hz MIDI data rate
• fix bug where pressure wasn't getting set w/o a MIDI connection
• send quantized pitch bend on MIDI note off
• fix touch tracker issues including zone-switch bug
• fix to allow MIDI note 0
• fix for double note off problem
• fixes for MIDI glissando

1.2.5:

• [1.2.5.1] Fixed a crash on startup if no preferences folder was present.
• Kyma listener off by default to fix collisions on port 3124. Use 'kyma' toggle on Expert page to turn on.
• fixed automatic connection to selected OSC service on startup.
• restored some values from 1.1.2 to improve touch tracking.
• add automatic saving of window dimensions. This is saved in /Application Support/SoundplaneViewState.txt.
• fixed a problem resolving OSC services
• fixed wrong MIDI note offsets in default Zone setups

1.2.4:

• rendering fixes for Retina display
• make touches easier to get into top and bottom rows
• code signing application
• turned Kyma polling over MIDI OFF by default
• fixed some state issues on startup that required reselecting zone to refresh
• clamp zone outputs to [0, 1] as documented
• fix touch rotate bug
• fix OSC browser
• made normalizing easier and mre accurate
• sending out x, y, and z from zones via MIDI.

1.1.2:

• more complete fix to the note-off problem
• restored the note lock feature for new zones.
• fixed an odd font-related bug

1.1:

• PLEASE NOTE: moved support files to ~/Music/Madrona Labs.
• fixed a problem where the wrong note value was sent on note-off
• fixed a potential crash sending MIDI if a MIDI device was not set
• made benign errors less alarming

1.0a2:

• two new kinds of zones: z (pressure only) and a toggle switch.
• fixed an error where inactive touches were continuously sending their data
• shortened some JSON zone names (just remove "controller_" to fix your presets)
• updated Max/MSP examples
• moved matrix message into t3d OSC bundle with touches
• restored some debug printing in the in-app console
• added error info for JSON parsing
• fixed possible bug with zone parsing
• clarified the T3D format in docs
• fixed an error reading calibration files
• fix view issues for Retina display

1.0a1:
-new Zone features allow mapping notes and controllers to key grid
-changed t3d format for wider OSC compatibility
-OpenGL accelerated graphics
-fixed graphics for Retina display
-fixed a bug where SoundplaneController was initialized twice
-fixed OpenGL errors on quit
-fixed a possible crash in adjustPeak()
-fixed note release when quantized
-new raw matrix output

Very glad to hear it. Enjoy!

Sounds like you fixed the patch? If so I'll leave it at that—if not let me know.

I'm sorry to hear it's not registering for you. I've heard this happens with some Windows systems but can't duplicate on mine. Do you get the demo noise every time you launch Aalto, or just sometimes?

Can you check to see if the license file Aalto.txt is there and contains your license key? It should be in C:/AppData/Roaming/Madrona Labs/Licenses.

Hi New Aalto user, I can definitely check oiut your patch if you send it to me. You can post it here or email to support@madronalabs.com.

Also please tell me what kind of computer CPU / OS you are using.

Another idea is to reduce the signal vector / process buffer size in your DAW and hear if that helps. But most things in a sequencer patch like you describe should be at full audio rate anyway.

It sounds like you are selecting a patch via a MIDI Program Change message. By putting patches in the ".../Madrona Labs/Kaivo/MIDI Programs" folder you make an alphabetical list of patches that the plugin switches to when it gets that message. So: a.mlpreset for Program 1, b for 2 and so on. It doesn't have to be "a" and "b", whatever the patches are in the in alphabetical order will be loaded.

So, Kaivo is getting a Program Change. Maybe you are in Ableton and have a Program Change message set for the clip? More info here: https://help.ableton.com/hc/en-us/articles/209070489-Sending-Program-Change-messages-from-Live

We don't have an email notification feature here. Maybe we should—will look into it.

Thanks for the info re: resolutions. I hope this helps someone! I'm not sure why other JUCE-based plugins do not have the same issue. I will look into it in case there's a quick fix. Mostly I'm focused on version 2 going forward.

A smaller delay would be very useful! The difficulty is that the pitch shifter is in the loop, and pitch shifting in real time can't be done instantly. I tried a lot of things to get the pitch shifting time down but as you do it starts to sound worse. So I probably had it at 20ms at one point then needed to go up to 50.

I could imagine maybe a mode where the non-shifted input is delayed instead if you do want a quick delay time. In other words the shifter would be smoothly removed from the loop somehow.

Hi, that is a good question. Without looking at the code: I'm 90% sure that the preset was simply made before the delay time changed to a minimum of 50ms. So it will clip to 50 ms in reality. Of course, we could measure this and verify.

I'm just about to release an update for Aaltoverb that will address this problem. It makes version 2 into a separate plugin—like most plugin makers seem to do this. Incredibly, I've never released a major version update before Aaltoverb 2, so thanks for your understanding.

I'll add more explanation when I officially release, probably on Monday. But it should be obvious and meanwhile here are the links:

Mac: http://madronalabs.com/media/aaltoverb/Aaltoverb2.0.2.pkg
Windows: http://madronalabs.com/media/aaltoverb/AaltoverbInstaller2.0.2.exe

Aaltoverb 1 is still available here:
Mac: http://madronalabs.com/media/aaltoverb/Aaltoverb1.1.0.pkg
Windows: http://madronalabs.com/media/aaltoverb/AaltoverbInstaller1.0.0.exe

Really neat! The first clip sounds almost synthetic because of the perfect timing. I think I can hear the stronger servos in action. These new pieces are exciting and I hope I get to hear longer ones when you make them. For what it's worth, it feels like complete music I would either work or bliss out to and I don't think it needs anything but the guitar.

I have been playing Egberto Gismonti and some related ECM-universe records I have on my morning "radio" show. Sundays at 10am at twitch.tv.joenszfamilyradio if you want to tune in!