WolfTalk: Podcast About Audio Programming (People, Careers, Learning)

Professor Julius Orion Smith III needs no introduction. He’s been at the renowned CCRMA at Stanford University since the beginning, first as a PhD student and then as a professor. He worked with Steve Jobs on the NeXT computer. He developed digital waveguide synthesis used in some of Yamaha’s synthesizers. He worked on the Shazam audio fingerprinting algorithm. He’s written four free DSP textbooks. He’s authored or co-authored hundreds of publications. And on, and on, and on.Needless to say, interviewing him has been my dream since starting this podcast. Listen to the interview to find out how you can profit from my joy 😉From this episode, you will learn:How Julius Smith became one of the world’s most well-known and respected DSP professorsThe story of CCRMA at Stanford UniversityBehind-the-scenes of the music features of the NeXT computerThe story of Yamaha synthesizer development (FM and physical modeling)The story of Shazam and GeoShred music appsHow to learn DSP wellHow to do excellent DSP researchHow Julius uses AI for research, learning & programmingThis episode was recorded on September 23, 2026.Audio post-production, mixing & mastering by Sathira Tennakoon. Video edited by Vadzim Vezhnavets.

Phil Burk has had an amazing career as an audio developer: from writing DSP code on Z80, through creating a music language, writing code for mobile phones, PlayStation audio support, and Android, up to MIDI 2.0 contributions. He’s also a co-creator of the PortAudio library, which is one of the most popular OS-agnostic audio libraries (and it’s used not just from C/C++ but from Python as well).He’s been there from the 80s up until today; he’s seen it all!What I love about Phil is his purely interest-driven approach. He was able to make his hobby his work and thus live a life of passion. Even today, as a retiree, he still codes 8 hours a day just for fun.After listening to this episode, you will not only learn a ton of useful audio programming knowledge and feel inspired, but you will also feel thankful that the world has audio developers such as Phil; they’re a real blessing, making our lives easier and more pleasant to the ear!Episode Contents From this episode, you will learn:how Phil created his first analog synth and started programming on Z80what challenges did early music programming facehow HSML music programming language came to bethe challenges of programming digital signal processors (DSPs) and designing audio hardwarehow mobile phone audio worked in the late 90s/early 2000sPhil’s awesome audio projects: PortAudio, JSyn, WebDrum, and more (see below)the story of MIDI 2.0how the Android team fixed the latency problemwhich languages Phil has used throughout these 4 decades of audio programmingwhat are his work habits for maximum programmer productivityThis episode was recorded on February 4, 2026.00:00:00 Podcast Intro00:00:36 Introducing Phil Burk00:02:13 Early Music & Homebrew Electronics00:04:08 Building a Shoebox Synthesizer00:07:18 Z80 Programming via Hex Keypad00:10:15 Emulating a 68000 CPU on a Z8000:14:08 Sponsor: JUCE00:15:55 Phase-Locked Loops & the Commodore 6400:18:41 From Biophysics to Programming00:22:11 Meeting Larry Polansky at Mills College00:26:02 HMSL: A Music Language Built in Forth00:32:43 Motorola 56000: Real-Time Synthesis00:35:30 3DO: First Software Synth Console00:43:38 Designing a Custom DSP in Verilog00:54:38 JSyn: Interactive Music in the Browser01:02:00 Building PortAudio with Ross Bencina01:05:52 Max Neuhaus Sound Installations01:14:33 A 14KB MIDI Synth for Mobile Phones01:27:02 Designing the MIDI 2.0 Standard01:40:41 Sony PlayStation 3 Audio Libraries01:44:44 AAudio: Fixing Android Audio Latency01:53:16 Why Android Audio Lagged Behind iOS02:03:08 Reviving HMSL with JUCE & PForth02:07:07 Retirement: Kotlin, AI & New Projects02:12:17 Every Language Phil Has Programmed In02:18:42 Are Dedicated DSP Chips Still Needed?02:20:43 Developer Setup & Favorite Tools02:23:23 The Music-DSP List & Learning DSP02:30:08 Just Enough, Just in Time Learning02:33:05 Audio Is Harder Than People Think02:36:26 Shower Debugging & Work Habits02:41:40 How to Reach Phil Burk02:42:50 Outro

Julian “Jules” Storer is the creator of the JUCE C++ framework and the Cmajor programming language dedicated to audio.He created JUCE in the late 90s, and it grew to become the most popular audio plugin development framework in the world. Apart from audio capabilities, it is a general-purpose cross-platform application development framework (Windows, macOS, Linux, Android, iOS, and embedded platforms). Most plugin companies use JUCE; whether you like it or not, it has become a de facto industry standard.You know that I love JUCE; I created the official JUCE audio plugin development course with them, and they are the sponsor of the podcast. So naturally, I was super excited to be able to interview Jules!His next big thing is the Cmajor programming language. It is a C-like, LLVM-backed programming language dedicated solely to audio.He has also given many talks at the Audio Developer Conference, so I encourage you to check them out as a way to relax and get inspired.Jules is known for his strong opinions and dry humor, so I guarantee you’ll find yourself chuckling every few minutes 😉Note: If you like the podcast so far, please go to Apple Podcasts and leave me a review there. You can do so on Spotify as well. It will benefit both sides: more reviews mean a broader reach on Apple Podcasts, and feedback can help me improve the show and deliver better-quality content to you. You can also subscribe and give a like on YouTube. Thank you for doing this 🙏Episode contents From this episode, you will learn:How Jules created the JUCE framework and distributed it initiallyHow to maintain such huge codebases as JUCEJulian’s coding principles that will make you (and me) a better devWhat problem does CMajor solve, and howWhich tools is Jules using when coding, especially when it comes to AIHis exact everyday work routines and relaxation strategies, andDoes Jules really hate CMake?This episode was recorded on January 30, 2026.00:00:00 Intro00:00:33 Introducing Jules Storer, Creator of JUCE00:02:37 Meet Jules00:03:11 How Jules Got Into Audio Programming00:04:41 University: Computing & Psychology00:05:28 Early Career00:07:46 Why Jules Built the Tracktion DAW00:11:04 Sponsor: JUCE Audio Plugin Course00:12:52 The Mackie Deal & the Birth of JUCE00:16:35 10 Years Solo: Bootstrapping JUCE00:18:59 ROLI Acquires JUCE00:20:56 How the Audio Developer Conference Began00:21:38 From ROLI to C-Major00:23:33 Cmajor: DSP Language for Any Hardware00:28:58 Cmajor Licensing & Commercial Plugins00:32:57 Cmajor Stability & AI-Written Code00:36:04 What Building a Compiler Taught Jules00:39:31 Jules' Dev Setup00:41:59 Jules' AI Coding Workflow with Claude00:47:23 Reviewing AI-Generated Code00:50:18 AI Coding: Where Jules Draws the Line00:52:22 Cmajor's Language Choice: C++, Go, and Rust00:54:36 Why Jules Hates CMake00:59:01 Clean Code: Structure, Files & Naming01:04:25 API Design Other Developers Won't Hate01:07:11 Code Review: Excess State & Verbosity01:11:35 JUCE's Inspiration from Java's JDK01:14:24 Why JUCE Dominated the Market01:15:48 Challenges with JUCE, Waveform & C-Major01:19:56 TDD and the Pain of UI Testing01:23:06 How to Start Learning Audio Programming01:26:42 Jules' Daily Routine01:31:52 How to Contact Jules01:32:42 Outro

How do you become a C++ Standards Committee member?Why is C++ prevalent in audio?Should you still use it for audio software?Honestly, Timur Doumler is someone I have looked up to ever since I saw his “C++ in the audio industry” talk at CppCon 2015.He has a rich development history with C++ and/or audio:developer at Native Instrumentsdeveloper of the JUCE C++ framework (podcast sponsor ❤️)C++ linter developer and developer advocate at JetBrains (who make the CLion IDE)founder of Cradle, an audio plugin startupC++ Standards Committee memberCppCast podcast hostnotorious Audio Developer Conference and CppCon speakerI have probably missed a ton of stuff here, but that should already give you a flavor of what Timur is up to 😉I especially enjoy his technical talks on synchronization with the (real-time) audio thread; For example, at the Audio Developer Conference 2025, he made me finally understand memory ordering in C++ (or so I believe 😅)In the podcast interview, we discuss his story, tactics, and tips, which I hope will inspire you to follow his footsteps (as they sure did me).It also turns out we are both fond of the same music band… Listen to the interview to find out which one 😁All in all, this one of my favorite podcast episodes (audio & C++, what more do you need?), so don’t miss it!Note: If you like the podcast so far, please, go to Apple Podcasts and leave me a review there. You can do so on Spotify as well. It will benefit both sides: more reviews mean a broader reach on Apple Podcasts and feedback can help me to improve the show and provide better quality content to you. You can also subscribe and give a like on YouTube. Thank you for doing this 🙏Episode ContentsFrom this episode, you will learn:How Timur’s early interests and experiences led him toward audio programmingHis early engineering work at Native Instruments, including contributions to Kontakt and NI’s internal C++ librariesHow he joined ROLI and worked on the development of JUCEThe story behind founding CradleHow he became involved in the ISO C++ Standards Committee including his audio-related contributionsWhat he’s most excited for in the upcoming C++26 standardHow he approaches software development and maintaining a healthy work-life balance (that was a great one to learn for me personally, too)This episode was recorded on September 26, 2025.TIMESTAMPS00:00:00 Podcast Intro00:00:37 Timur Doumler00:01:55 Timur's Introduction and Background00:02:46 Early Interest in Music00:03:35 Linking Music to Software Development00:04:42 Studying Physics & Astrophysics (Berlin, Lyon, Potsdam)00:06:08 Learning C & Fortran through Cosmology Simulations00:07:46 Deciding to Become a Developer00:10:48 Getting Hired at Native Instruments00:12:16 Working on NI’s Internal C++ Libraries & Kontakt00:18:05 Sponsor: JUCE00:19:37 Moving to ROLI and Working on JUCE00:22:08 Lessons Learned from Jules & Fabian00:24:27 Joining JetBrains00:30:13 Becoming a Developer Advocate00:34:40 Founding Cradle & Building the First Products00:37:17 Life as CTO & Startup Reflections00:39:09 Bloomberg & C++ Contracts00:41:29 The 2015 Talk: C++ in the Audio Industry00:46:58 Representing the Audio Industry00:48:18 Joining the ISO C++ Committee00:50:19 Timur's Work in the C++ Committee00:55:24 How the C++ Committee Works01:02:20 How to Learn C++ Today01:06:48 Real-Time Audio Programming & Tools01:08:02 Personal Productivity: Meditation, Exercise, Time Management01:13:46 Tech Stack and Developer Tools01:14:13 Music Timur Codes To (Meshuggah, Animals as Leaders)01:17:15 How to Contact Timur01:18:59 Outro

Kurt James Werner, PhD, is a senior research scientist at Soundtoys and one of my favorite researchers in audio DSP (virtual analog modeling in particular).While you may not have read his papers, it’s very likely that you used one of the plugins he worked on:NeoverbVinylRXGuitar Rig (Hammond vibrato/chorus)OzoneNeutronTrashIf you start peeking into the research of modeling analog audio effects in plugins, you stumble across his name right away.I consider him not only brilliant, but also insanely productive. Definitely a role model for me!In the research world, he’s mostly known for his work on Wave Digital Filters (WDFs). It’s a technique for creating a mathematical model of an analog circuit that once done is easy to implement (provided you use a WDF library).WDFs are great for modeling analog audio effects to put them inside plugins.But as a true researcher, Kurt is involved many other audio subfields that we discuss in the podcast!Did I mention that he’s a graduate from Stanford’s CCRMA?In this episode, you’ll learn about Kurt’s transition from academia to industry, his work on products at iZotope, Native Instruments, and Soundtoys, and the realities of being a research scientist in an audio plugin company.You’ll also learn all about WDFs, which is a powerful tool to master.Trust me, you don’t want to miss this one 😉Note: If you like the podcast so far, please, go to Apple Podcasts and leave me a review there. You can do so on Spotify as well. It will benefit both sides: more reviews mean a broader reach on Apple Podcasts and feedback can help me to improve the show and provide better quality content to you. You can also subscribe and give a like on YouTube. Thank you for doing this 🙏Episode Contents From this episode, you will learn:How Kurt’s early musical and DIY electronics experiments led him toward audio DSPHow was it like to study at CCRMA under Julius Smith, Jonathan Abel, and Ge WangA clear, high-level explanation of Wave Digital Filters, their origins, and why they matter in virtual analog modelingInsights from his work at iZotopeHis current work at SoundtoysCreative music practices like circuit bending and 1-bit musicWhether you need a PhD to work at an audio plugin companyHow audio research translates into pluginsTIMESTAMPS00:00:00 Podcast Intro00:00:37 Introducing Kurt Werner00:04:11 Kurt's Intro and Background00:07:20 Studying Engineering & Music at UIUC00:10:06 PhD Studies at CCRMA (Stanford)00:13:57 Structure of the PhD Program00:18:23 Sponsor: JUCE00:20:09 PhD Thesis: Wave Digital Filters00:21:26 The Culture at CCRMA00:31:44 Assistant Professor at SARC (Queen's University Belfast)00:35:59 How Teaching Influences Research00:44:19 Transition to Industry: iZotope00:49:54 The Research Team at iZotope00:52:54 Work on Neoverb, Vinyl, and Trash00:58:43 Publishing Reserach Papers01:01:35 Moving to Soundtoys01:05:12 Role and Soundtoys01:09:17 Soundtoys SuperPlate01:11:04 What are Wave Digital Filters (WDFs)?01:19:28 WDFs Before Kurt's Research01:20:54 Handling Complex Topologies01:26:56 Can Circuit Modeling be Automated?01:33:37 Importance of "Warming Up" the Simulation01:35:16 Strengths of Wave Digital Filters01:41:52 Best & Worst Circuits for WDF Modeling01:44:47 How to Approach Modeling a New Circuit01:46:47 Modeling the Motown EQ (Altec 9062A)01:52:28 Circuit Bending & Hardware Hacking01:56:58 1-Bit Music & Velvet Noise02:05:35 Deep Learning in Virtual Analog02:07:29 Do You Need a PhD for Audio Research?02:10:46 PhD for Software Engineering Roles?02:12:12 How to Learn Virtual Analog Modeling02:16:04 Productivity & Writing Research Papers02:20:04 Tools & Tech Stack02:21:29 Listening to Music While Working02:23:16 Outro