planet.linuxaudio.org

Internet radios have been a thing in some form for the last quarter century, and the advent of cheap networked microcontrollers has made them easier than ever to build. But just how small can one be made? [Milen] has made one that’s about as small as we’ve seen, and put it up on Instructables.

The hardware recipe is straightforward enough; take a microcontroller, connect it to the internet, and have your internet radio software squirt its output to a DAC. In this case the microcontroller is an ESP32-C£ on a very small dev board, and clipped to that is a custom PCB with a PCM5102A I2S DAC and an amplifier. The whole thing is tiny, small enough in fact that the two connectors are significant in size compared to it.

Whether or not it’s a practical device for listening remains to be seen, but it’s certainly tiny. But the question is, could it be made smaller? Perhaps eschewing the connectors would be an easy win, and in return for a loss in quality the ESP has a built-in DAC that’s not intended for audio but can be used. If you have any bright ideas, we’d be interested to hear them.

We’ve seen plenty of internet radios over the years, and for some reason this cassette-shaped one appeals to us.

Technology keeps asking us to speed up. Pd is happy to let you slow down: tinker, patch, make mistakes, obsess over a number or signal routing, and do it all visually. The free and open source software is getting a nice 0.57 update with new features, relevant to Pd and tools like libpd and plugdata that are based on it.

The post Pure Data (Pd) 0.57 inbound, enduring free software for tinkering with sound appeared first on CDM Create Digital Music.

 Believe it or not, I do still exist. And I do some projects still.
Sometimes.

Anyway, I've been starting to play more with hardware synths because I don't have  a dedicated studio space right now to set up softsynths and I just want to play around without messing with computers and wiring everything up. Regardless of the reason I wanted to be able to connect my KMI kboard up to my Korg Volca FM or my just arrived AMYboard but to do so requires a USB Midi Host. There are devices for sale that do this, but I realized that I have a device in a box that will do this already.

 So I dug out my raspberry pi zero W and a new project was born.

First thing I set up on the pi was a bash script that will parse the output of aconnect (an alsa utility that can list all midi devices/ports). It looks like this:


#!/bin/bash
LOGFILE="/var/log/midi-connector.log"
echo "$(date) - MIDI event triggered" >> $LOGFILE

# Function to connect available MIDI ports
connect_midi()
{
    # Find the client IDs of all connected mid devices
    # This grabs any non-system MIDI ports available
    PORTS=($(aconnect -l | awk '/^client/ { if (/card=/) { c=$2; sub(/:/,"",c); v=1 } else { v=0 } } v && /^ +[0-9]+/ { print c":"$1 }'))

    # Then loop through them:
    for port in "${PORTS[@]}"; do
        echo "Found hardware port: $port" >> $LOGFILE
    done
    listsize=${#PORTS[@]}

    for (( i=0; i < listsize; i++ )); do
        for (( j=i+1; j < listsize; j++ )); do
            
            port_a="${PORTS[i]}"
            port_b="${PORTS[j]}"
        
            # Extract just the client number (everything before the colon)
            client_a="${port_a%:*}"
            client_b="${port_b%:*}"
            
            # Only connect if they belong to different hardware clients
            if [ "$client_a" != "$client_b" ]; then
                
                # Connect Port A to Port B
                aconnect "$port_a" "$port_b"
                
                # Connect Port B to Port A
                aconnect "$port_b" "$port_a"
                
                echo "connecting $port_a $port_b" >> $LOGFILE
            fi
        done
    done
    
}

connect_midi

I saved that as /usr/local/bin/midi-connector.sh then did a chmod to make it executable.

sudo chmod 700 /usr/local/bin/midi-connector.sh

Next you need a udev rule that will run whenever a USB device is connected. Paste this one line into /etc/udev/rules.d/99-usb-midi.rules:

ACTION=="add", SUBSYSTEM=="usb", ENV{ID_USB_INTERFACES}=="*:0103*", TAG+="systemd", ENV{SYSTEMD_WANTS}+="midi-connect@$env{NAME}.service"

That line tells the USB manager that when a newly connected USB device is a midi device then run a systemd service. So we have to make that systemd service next.

In a new file called /etc/systemd/system/midi-connect.service put the following:

[Unit]
Description=Auto-connect USB MIDI for %I
After=sound.target alsa-restore.service local-fs.target

[Service]
Type=oneshot
# Give ALSA 2 seconds to fully spawn the client nodes in the background safely
ExecStartPre=/bin/sleep 2
ExecStart=/usr/local/bin/midi-connector.sh

[Install]
WantedBy=multi-user.target

Alsa takes a second after the udev rule fires to get the midi port ready so we have a little delay before we run the connector script, then the script goes and takes care of it. We also have this set up to run once at boot in addition to when udev launches it in case I power it on with everything connected (or in case the system browns out which happened a couple times).

That's all it takes. If you don't want to reboot just run:

sudo udevadm control --reload-rules
sudo systemctl daemon-reload

to reload the pieces then you should be able to plug in any usb midi devices and they'll all start sending messages to each other.

So overall how well does it achieve the goal? I don't love how long it takes to boot (a couple minutes). And it seems like the system is fickle. Not sure I'm ready to rely on it if I were doing real gigs or something. The USB hat POGO pins aren't a supre great connection so I found better luck powering through the Pi USB port directly rather than through the USB hat power input. If I were to do this properly I'd use a rp2040 based chip and make a custom firmware which should boot much faster than a full OS on a single core that I have now. But this was quicker.

 Anyway. I did a thing. You can now too. Enjoy!

In 2024 we opened the MOD OS images, so anyone could change the software and share the result. One Dwarf owner did exactly that. Seventeen releases later, Starless has its own community, and it changed how we develop MOD OS. Here is his story, in his own words, and what it led to.

Why I started Starless

By Andrea Del Signore

I first heard about MOD in its early days through the open source world, which I’ve always been passionate about. Then I lost track of it, until I came across an ad for a used Dwarf and bought it.

At the time my main unit was a Line 6 Helix LT, and I immediately saw the Dwarf’s potential: a compact device with creative routing options, a perfect fit for my acoustic and classical guitar rig. But a few software limitations made me doubt it could work live for me. They were small details, but vital ones. I couldn’t change a pedalboard’s parameters on the fly without a computer, not even to turn down a reverb or adjust an EQ. The plugins and the creative possibilities were there, but those few gaps kept me dragging the big, heavy Helix to every gig.

Then I noticed that almost all the sources were available on GitHub and that, in theory, I could modify the software and patch the official images. I spent the next three or four evenings on the couch reading code and setting my first goals. The plan was simple. First, a minimal change to get familiar with the codebase. Then, the few features I needed to play live without worrying, so I could finally replace the Helix.

After a few weeks of trial and error I built my first image. It introduced port groups in the UI and two complementary ways to edit pedalboard parameters: new firmware and a new mobile-friendly page, now known as the Performance View.

From there, the community stepped in. Abotte, Rom, Fer and others adopted the image, tested it and gave me feedback, and that is how Starless was born. The name comes from what I was working on at the time: a classical guitar arrangement of King Crimson’s song of the same name.

A few months later I sold my Helix and kept adding features that help both while editing and during performances. Today other talented developers, such as Stefan and Andy, are helping me refine Starless for use cases beyond my own.

I want to thank Abotte, Toferyu, Remy, Rom, Fer, Zwabo, UnvisualStudio and many others, along with the MOD team for the help in the very first days. And finally, Gianfranco and MOD Audio, for so warmly welcoming the development of this project.

What Starless is today

Starless is an alternative OS for the MOD Dwarf, made possible by the open images we published with MOD OS 1.13.5 in October 2024. It is now on version 17, and its latest release added, among other things:

• Complete pedalboard editing from the Dwarf itself, no computer needed
• Tone3000 support in the NAM, cabinet and IR loaders
• Stereo cables and cable teleporting, for tidier pedalboards
• A per-plugin CPU monitor, mouse-wheel zoom and a canvas that resizes itself as you add plugins

AndyCap has also been working on mapping the same control to the Dwarf’s own controls, MIDI, CC and CV at the same time, now in beta. See it in action on the forum.

Try Starless

Starless is Andrea’s project, with its own name, schedule and decisions. It is an unofficial image, so MOD does not provide support for it. The community does, on the forum.

1. Back up first. Use the Backup & Restore function in the web interface (see the MOD wiki).
2. Install the latest Starless image as a manual upgrade. Your pedalboards and downloaded plugins stay where they are.
3. Ask on the forum if anything is unclear. Here is one Dwarf owner’s first install, step by step.

What Starless changed: the new MOD OS release flow

Starless showed us something important. When the software is open, the people who use it every day will improve it, and their work deserves a real way into the product. So we changed how official MOD OS releases are made.

MOD OS is now released through three channels, a model inspired by how long-running open source projects like Debian work:

The first release candidate under this flow is MOD OS 1.14, and it is already built from community work. It brings Tone3000 into the web interface, so you can browse and load captures directly on your device, and a “Report an Issue” button, so testing is one click away. The announcement of the new flow publicly thanked five community contributors: Andrea Del Signore, Andy Cap, Remy D, Stefan D and Mark Melvin.

In other words: what happened with Starless is now how MOD OS moves forward. Community work flows in, gets tested in the open, and ships.

Join in

All of this happens on the forum, through people testing, reporting problems and asking for features. Try a release candidate and tell us what breaks. Try Starless and tell Andrea what you would like next. And if you have your own idea for MOD, the forum is where it starts.

Get the next story by email

We will be sharing more of what the community builds. Sign up to get it in your inbox.

We're awash with emulations at the moment, but this one is on another level. The chip that powered everything from Yamaha’s little-known CX5M computer to the best 80s FM synths and video game masterpieces gets a long-awaited (unofficial) rendition on your computer, and it's just about perfect -- both as a instrument plug-in and an ultra-authentic player for classic games. And where this emulation started was physical, not virtual: a workbench, vintage gear and documentation, research, and physical labor.

The post Mortal FM Kombat: Plogue’s meticulous OPMZ Yamaha emulation totally slaps appeared first on CDM Create Digital Music.

Artists Tina Mariane Krogh Madsen and Malte Steiner have been artists in residency in the Life on a Leaf house for the month of September. On Sunday 27.9, they will host an event at the Leaf house presenting some of the artistic processes they have worked with, as solo concerts, scores and software art. The event will start at 16.00 until 19.00, and there will be time to hang out and have conversations with the artists.
https://www.lifeonaleaf.fi/en/events

The GStreamer Conference team is pleased to announce that the preliminary list of talks, abstracts, and speaker biographies is now available for this year's lineup of talks and speakers, covering again an exciting range of topics!

The GStreamer Conference 2026 will take place on 10-11 October 2026 in Prague, Czech Republic, followed by a hackfest.

Details about the conference, hackfest and how to register can be found on the conference website.

This year's topics and speakers:

• GStreamer State of the Union
  Tim-Philipp Müller, Centricular
• Fulfilling a Childhood Dream: Video Streaming for Remotely Controlling a Train
  Rainhard Driessler & Simon Himmelbauer, A&R Tech
• A new MPEG-TS demuxer in Rust
  Edward Hervey, Centricular
• That F#$%ing AI Talk - Debugging GStreamer With AI Tools
  Ray Tiley, Tightrope Media Systems
• GstVA and GstVulkan updates
  Victor Manuel Jáquez Leal, Igalia
• Building a Ringbuffer - Going in circles with V4L2 and PipeWire
  Simon Himmelbauer, A&R Tech
• GStreamer 20 years of fun
  Stéphane Cerveau, Igalia
• Permissions, Isolation and Streams: Redesigning WirePlumber's Core
  Julian Bouzas, Collabora
• GStreamer highlights in Servo
  Xabier Rodríguez Calvar, Igalia
• Write once, run anywhere inference with GStreamer
  Sebastian Dröge, Centricular
• Every Camera Is a Snowflake: Building a VMS That Works With Thousands of Device Models
  James Nickerson, IP Configure
• The Quest for Python Analytics
  Aaron Boxer, Collabora
• WebRTC in the Wild: It Works on My Network
  Jake Foytik, IP Configure
• Rock, paper, scissors
  Xabier Rodríguez Calvar, Igalia
• Exploring GStreamer as a media backend for game engines
  Dominique Leroux
• Embedded Reality: GStreamer Versioning strategy on SoC Products
  Kye Yeon Ko, LGE
• WebRTC take 2, a decade later
  Matthew Waters, Centricular
• Gst-Audit: A graphical way to instrument your pipeline
  Jorge Zapata, Fluendo
• A GStreamer-Based Performance Evaluation Tool for End-to-End Visual AI Pipelines
  Jan Gorecki, Intel
• A new Media Source Extensions implementation in GStreamer for cross-platform apps
  Sebastian Dröge, Centricular
• GStreamer as a Foundation for Vision AI Edge Solutions
  Tomasz Janczak, Intel
• Local transcription and translation with GStreamer: whisper.cpp + llama.cpp
  Sebastian Dröge, Centricular
• How GStreamer is used in WebKit
  Enrique Ocaña González, Igalia
• librice: challenges with sans-io
  Matthew Waters, Centricular
• Ingesting thousands of streams over SRT on the same endpoint
  Jonas Danielsson, Spiideo
• Multithreading in GStreamer: how do flushes work
  Alicia Boya García, Igalia
• GStreamer 2.x without breaking API ?
  Edward Hervey, Centricular
• Outsourcing the boring work: setting up LLM agents for fixing bugs in legacy software
  Kim Schmider, A&R Tech
• Real-time offshore crane teleoperations powered by GStreamer
  Adam Blonsky, Palfinger
• Dolby/AC-4: The plan to support the Dolby ecosystem in GStreamer
  Pablo García & Maciej Sabiniok, Fluendo
• Color, tracing and WebAssembly: a year of GStreamer for video editing
  Thibault Saunier, Igalia
• Fluster over the course of 2026
  Rubén Gonzalez, Fluendo
• SIL 2 Video Streaming - Functional Safety from Lens to Glass
  Rainhard Driessler, A&R Tech
• Secure Video Path Done the Right Way
  Nicolas Dufresne, Collabora
• Making GStreamer Feel Native in Zig
  Per Andersen, TV2
• Restream and GStreamer, a success story.
  Loïc Le Page, Igalia
• When the robots came for our parsers: GStreamer security in the age of LLMs
  Sebastian Dröge, Centricular
• Digitally Signed Content & C2PA
  Diego Nieto Munoz, Fluendo
• Unifying GstAnalytics Metadata: Achieving Cross-Vendor Interoperability in Video Analytics
  Oskar Fiedot, Intel; Pep Cots, Milestone Systems; Daniel Morin, Collabora
• Whither GStreamer 2.0 ? - Overview, Presentations and Panel Discussion
  Various speakers, tbd
• Whisper transcriber
  Mathieu Duponchelle, Centricular
• GstVulkan in Production: Lessons from the Real World
  Anders Hellerup Madsen, Veo Technologies
• Per pipeline logs tracer
  Mathieu Duponchelle, Centricular
• Why is MPEG-TS so hard?
  Jan Schmidt, Centricular
• Let's solve our long standing issues: time for GStreamer 2.0 ?
  Thibault Saunier, Igalia
• Latency is bad, isn't it?
  Jan Schmidt, Centricular
• Let's renegotiate!
  Mathieu Duponchelle, Centricular
• GStreamer in and out of the box
  Jan Schmidt, Centricular
• ...and more to come
• Submit your talk or lightning talk now!

Many thanks to our amazing sponsors ‒ Platinum sponsors Centricular, Pexip, and Igalia, Gold sponsors Axis Communications and Collabora, Fluendo, and Silver sponsor Tightrope Media Systems, and our Media Partner UbiCast without whom the conference would not be possible in this form.

We hope to see you all in Prague soon!

Don't forget to register as soon as possible if you're planning on joining us, so we can order enough food and drinks!

There's also a hackfest after the conference.

In the cross-over between the era of tapes into that of MP3s, you’d see quite a few of those special cassette tapes that were actually digital music players inside. Some simply provided a 3.5 mm input, while others were complete MP3 players or Bluetooth receivers that just happened to also output to the magnetic read head of a cassette player. Recently [Jonathan Rowny] decided to make his own version of the latter.

Although getting the actual audio signal into the read head is easy enough – requiring little more than its equivalent being used as a write head on the cassette side – actually interfacing with the player’s mechanisms like auto-stop, reverse and so on requires the use of some gearing that detect motion on what would be the tape spools, as well as transfer the motion from the take-up spool to the other spool so that features like the auto-stop mechanism don’t get triggered.

A lot of inspiration here can be found in e.g. the videos made by [Clint] of [Lazy Game Reviews] who looked at a number of examples – including their internals – over the years, with various levels of functionality. For this particular implementation an ESP32-S3 module is used for the brains, along with a microSD card reader for music and a PCM5102 I2S audio codec to create the analog audio signal.

The gears were printed using an SLA printer and seem to work all right. Unfortunately he didn’t realize the importance of the capstan as the mechanism that actually transports the tape, so its motion was not measured as is done in the better cassette adapters. This will likely be corrected in a future iteration, however.

David García Goñi released a new version of Elektroid, a sample and MIDI device manager for Elektron, Arturia, Eventide, Moog, Novation, and now Korg devices.

Device support

Elektroid 3.4 introduced support for several devices:

• KORG prologue, minilogue xd and NTS-1 (units only);
• Elektron Monomachine SFX-60 + MKII.

Additionally, there is now support for uploading Elektron Syntakt samples.

The next version will be coming with MFB Tanzmaus support, the code is already in the development tree.

Fixes

David also added a Preferences option to disable safety questions and fixed model cycles renaming and preset uploading to Microfreak units and.

Elektroid is available in source code and as a flatpak build. The program also builds on macOS, but there is no DMG yet.

Malte Steiner's generative projection piece ‘retexturizing: recontextualizing‘ for the open house Konstrundan event at Life on a Leaf 6. September in Turku (FI). A computer generates and animates randomly collages out of 1892 image fragments of processed real estate ads from Steiner's project The Big Crash, Art for the pending burst of the real estate bubble . Programmed in Pure Data / Gem, running on a Linux computer.

The GStreamer team is pleased to announce another release of liborc, the Optimized Inner Loop Runtime Compiler, which is used for SIMD acceleration in GStreamer plugins such as audioconvert, audiomixer, compositor, videoscale, and videoconvert, to name just a few.

This release mostly contains bug fixes for some i386-related regressions.

Highlights:

• Fix various i386-related regressions
• avx512: Disable target on x86
• avx512: Implement the missing memoff-reg moves
• Fix Debian gcc miscompiling the backup float conversion functions

Direct tarball download: orc-0.4.44.tar.xz.

As should be obvious from browsing our website for any amount of time, Ubuntu Studio releases a Long-Term Support (LTS) release every two years. The latest LTS release happens to be our latest release: 26.04. What’s more is we just released 26.04.1, a roll-up of bugfixes and security updates, on August 27th.

Also announced is the “Butllet-Proof” KDE Initiative, where Kubuntu Focus, Techpaladin Software, and KDE e.V. are working together to make Plasma 6.6, along with KDE Frameworks, an LTS release as well. This means that Techpaladin, Kubuntu, and Kubuntu Focus will be working to fix bugs in the version of the desktop environment and back-end libraries released in Ubuntu Studio 26.04 LTS. This means that, while Kubuntu will be primarily working on the Stable Release Updates as provided by KDE, Ubuntu Studio uses the same packages that get these updates.

Our project leader, Erich Eickmeyer, has been doing what he can to help support Kubuntu with this task from his side, while the Ubuntu Studio team (including Erich) focuses on bug fixes to the multimedia applications included with their own Stable Release Updates for the life of 26.04, which is three years total until April 2029.

That, combined with Ubuntu Pro (free for every user up to 5 simultaneous machines) means you will not only receive bugfixes better than in the past, which was done at best effort, but also benefit from the security updates provided by Canonical. Yes, even Ubuntu Studio can use Ubuntu Pro. Do note that Ubuntu Pro will not extend the support time period for Ubuntu Studio, but give the added benefit of sponsored security updates.

If you would like to help with everything Ubuntu Studio has going on, from supporting the LTS to supporting the upcoming 26.10 release due in October, please feel free to join us!

Additionally, while there was an initial jump in monetary support following the 26.04 LTS release, it has dropped-off considerably. Please consider becoming a monthly monetary contributor to help support this project. Remember, our website now has a monthly cost along with the usual cost of packaging and development. If you cannot commit to a monthly monetary contribution, please consider at least a one-time contribution, even if you’ve given before. Even if it’s just a tank of petrol! For those already giving, thank you so much for your continued support. It means the world to us, and allows us to continue giving you and others this operating system for free! You’ve helped countless people throughout the world!

A new release of gst123, my commandline media player based on GStreamer is available. The main new “feature” is probably that it now can play videos without crashing on Wayland, although it still relies on Xwayland to do so. See release link for full list of changes.

About

Prerequisites

Installation

dd if=/path/to/downloads/a3vtl_VERSION.img of=/dev/sdb

Persistence

Getting Started

Advanced -- Shell script to create USB drive

#!/bin/bash
# the disk-device to use: 
DEVICE=/dev/sdb # or /dev/mmcblk0 or /dev/hdb ,...

URL=http://robin.linuxaudio.org/a3vtl_latest.img
URL=http://rg42.org/d/a3vtl_latest.img

# Download the image and write it to the disk:
curl http://rg42.org/d/a3vtl_latest.img | dd of=$DEVICE bs=64k

### OPTIONAL: ###  
# re-read the partition-table after the image has been written
partprobe $DEVICE
# create a 2nd partition using the rest of the space on the device..
echo -e "n\np\n2\n\n\np\nw\n" | fdisk $DEVICE
# ..and create an ext4 filesystem on it, labeled live-rw'
mkfs.ext4 -L live-rw ${DEVICE}2

MDA.lv2 1.2.12 has been released. This is a port of the MDA VST plugins to LV2.

Changes:

• Avoid over-use of yielding meson options
• Fix build with MSVC 2022
• Fix strict aliasing violation
• Fix use of uninitialized values in TalkBox
• Improve const-correctness
• Remove development symbols from modules by default
• Use a deterministic and more uniform RNG

Jalv 1.10.0 has been released. Jalv (JAck LV2) is a simple host for LV2 plugins. It runs a plugin, and exposes the plugin ports to the system, essentially making the plugin an application. For more information, see http://drobilla.net/software/jalv.

Changes:

• Add block length command-line parameter for PortAudio
• Fix PortAudio backend
• Fix crash with plugins that have non-event inputs
• Fix stale parameter descriptions in man pages
• Fully implement fixedBlockLength and powerOf2BlockLength
• Gracefully handle failure to open audio backend in Gtk interface
• Update meson fallback subproject wrap files