01 Start with the object
It's a studio microphone: the kind you'd find in front of a singer in a recording booth, not the one on a gaming headset.
Its two circuit boards were drawn for it, from a circuit published by Jules Ryckebusch, known as DJJules (True Condenser OPA Mics, 2021, CC BY 4.0 licence). The dual-diaphragm capsule and the metal body come from specialist makers. The plans, the parts list, the measurements and the bring-up journal are all public.
- Type
- condenser studio microphone, two channels
- Reference microphone
- RØDE NT1-A
- Assembly time
- 4 to 6 hours
- Status
- P3 prototype, being measured
- Target price
- €175, cable included · nothing is for sale today
02 What's inside
A studio microphone comes down to four things:
- the grille, on the left: it protects the capsule and breaks up puffs of air;
- the metal tube, at the bottom: the body;
- the rings and seals: the hardware that holds it all together;
- the long black board: the electronics, slid into the tube.
That cylinder is more than a casing. In the open, the electronics pick up the 50 Hz hum of the mains. Once they are inside the closed, firmly screwed body, the hum is gone: the body acts as a Faraday cage, shielding the inside from the electric fields present in any room.
03 The part that hears
At the top of the microphone sits the capsule. Inside it is an extremely thin diaphragm, stretched like a drum skin and coated in gold.
- You speak: your voice makes the air vibrate.
- The air makes the diaphragm vibrate, by a few millionths of a millimetre.
- That tiny movement becomes an electrical signal.
For that last step to work, the diaphragm has to stay electrically charged all the time, a bit as if it had a small built-in battery. That's the job of one of the two boards below.
The Talas One's capsule has two diaphragms, one on each face, and each has its own electronic channel. That is what lets you choose afterwards how the microphone hears: front only, all around, or front and back (stop 08).
The challenge: the signal that comes out is tiny and very fragile. Plugged straight into a speaker, you'd hear nothing. And it's so weak that the slightest disturbance drowns it. Hence everything that follows.
04 The two circuit boards
The first one amplifies. It takes the diaphragm's tiny signal and makes it strong enough to travel down a cable to a computer without degrading. This is the delicate part of the project: the signal is so weak that the components must let almost no current through — hence resistors of a billion ohms, a rare value even in electronics.
The second one produces a high voltage. The diaphragm has to be charged: this board makes that charge from the current supplied by the audio interface, stepping the voltage up stage by stage.
Why two boards, not one?
- The second one produces a very fast hum: 76,000 oscillations per second, inaudible but very real. Keeping it physically apart from the fragile circuit stops it from polluting the sound.
- If one board fails, or you want to change one, you replace only that one.
05 How it's all drawn
Before a single board is ordered, everything is drawn in design software, in two steps:
- The schematic: which component is connected to which. It's the logical plan, like the wiring plan of a house: you know the switch controls the lamp, without yet knowing where the wire runs.
- The layout: where each part sits on the board, and which path each copper track takes. It's the architect's plan.
The two screenshots above show the second step: each coloured line is a copper track, each round dot a pad where a component will be soldered.
The software then checks a long list of errors on its own: two tracks touching, a track too thin for the current it carries, a forgotten component. Only when every check passes do the files go to the manufacturer, who sends back the bare boards one to two weeks later. The circuit is also simulated before fabrication; whatever neither the checks nor the simulation can see, the workbench reveals: that's what prototypes are for.
06 Assembly
The boards arrive bare. The components then have to be soldered one by one — some are less than two millimetres across — before the two boards are wired to each other, to the capsule and to the socket at the back of the microphone.
Allow four to six hours for a full build, and a good magnifier.
To find your way, the repository provides an interactive parts map: preamp · power supply. The parts list on one side, the board on the other: click a component and it shows where it goes. It's the tool that makes the build practical without printing a plan.
07 The test bench
This is what the work really looks like: boards laid flat, crocodile clips, an oscilloscope probe, a soldering iron, and the capsule in the middle of it all.
This is where each prototype is brought up before its first sound: the voltages are checked one by one, each channel is listened to, then the body is closed for the measurements.
08 Does it actually work?
Yes. The Talas One has worked on both of its channels since 5 November 2026 (P3 revision of the boards, dual-diaphragm capsule).
But "it works" isn't enough: you need to know how well. On 7 November 2026 we compared it with a very widespread studio microphone, the RØDE NT1-A: the two microphones were placed one after the other at the same spot, in front of the same loudspeaker. The electronics were then measured on their own, by injecting an electrical signal in place of sound.
- ✓The same tone as the reference, from the lows to the upper mids. From 70 Hz to 4 kHz, its response stays within ±1.5 dB of the NT1-A's.
- ✓A strong signal. About 55 mV/Pa, more than the NT1-A: it needs less gain from the audio interface.
- ✓Three pickup patterns from a single take. By mixing the two channels afterwards, you choose how the microphone hears: front only (cardioid), all around (omnidirectional), or front and back (figure-of-eight).
- ✓Very clean electronics. Distortion of 0.0006 to 0.0041 % from 47 Hz to 10 kHz, and plenty of headroom: clipping only sets in around 127 dB SPL (1 % distortion).
- ✓Quiet. The mains hum is gone, and the electronics' self-noise is around 5 dB(A), measured on channel A with a dummy capsule.
These figures come from one unit, measured in an ordinary room: the shape of the results is reliable, the exact values will be confirmed under controlled conditions. Method and data: P3 measurements.
09 Hear the difference
So far, everything has been something to look at. Here's what matters most: the sound.
The two excerpts below were recorded on 7 November 2026, one after the other, at the same spot and at the same setting: pink noise played through a loudspeaker, in an untreated room. Since the Talas One is more sensitive, its level was matched to the NT1-A's: what remains is the tone.
Pink noise
Pink noise is a hiss that carries the same energy in every octave, from low to high. With no melody or words to catch the ear, it lets you hear each microphone's tone.
What to listen for: from the lows to the upper mids the two microphones are very close (±1.5 dB from 70 Hz to 4 kHz), and the top end above 10 kHz is the same. The main difference sits around 5.5 kHz, where the NT1-A lifts the presence. The two players aren't synchronised: play them one after the other, on headphones if you can.
10 The measurements, as pictures
Reading these curves, in two sentences. From left to right: the lows, then the mids, then the highs. Where a curve rises, that range stands out more; where it falls, it sits further back.
Against the RØDE NT1-A
Pickup pattern
Distortion
Method, raw data and further curves: measurement documentation.
11 Milestones
- March 2026First sources published: the project goes public.
- 12 April 2026First recording from the first prototype (P1).
- June 2026First measurement campaign against the RØDE NT1-A.
- Summer 2026The P3 revision of the boards is drawn, simulated and verified.
- 18 October 2026First P3 unit assembled.
- 5 November 2026The P3 works on both channels.
- 7 November 2026Acoustic measurements against the NT1-A: response, sensitivity, distortion, pickup pattern.
- 8 November 2026The electronics measured on their own: distortion and headroom.
- 13 November 2026Mains hum eliminated; the electronics' self-noise measured.
12 Why the plans are public
The plans, the measurements and the parts list are published under an open licence. Anyone can build this microphone, modify it or draw on it, including to sell it — provided they publish their own changes in turn.
The bring-up journal is published too, problems and fixes included: anyone rebuilding the microphone knows in advance where to look.
13 Where it stands, and what's next
- Done
- the P3 on both channels, measured against the NT1-A and electronics alone
- Still to measure
- channel B with a dummy capsule, the complete microphone with its real capsule, an absolute sensitivity, a second unit
What comes next, in three steps:
- Finish measuring the P3 and build a second unit, to compare it with the first.
- Choose the production capsule and body, after measuring several capsules from several makers; the raw data will be published.
- Complete the hardware release: open hardware certification (OSHWA), frozen fabrication files, then units built from the documentation alone.
The detail, with target dates: project status.
Today, the Talas One is a self-funded research project.
14 Frequently asked questions
- "Why not just buy a microphone?"
- Because the point isn't only to own a microphone, but to understand how it works and to be able to change it. Here the schematics are published and the two boards can be replaced separately: you can change one element and hear what it does.
- "How does it sound next to a shop-bought microphone?"
- Against a RØDE NT1-A, it follows it within ±1.5 dB from 70 Hz to 4 kHz; the NT1-A lifts the presence a little more around 5.5 kHz. Listen to the pair at stop 09.
- "How long did it take?"
- About a year, from the first schematic to today.
- "Can I buy one?"
- Not yet. talas.fr keeps a waiting list, with no commitment and no payment, so you can be told. In the meantime, the plans and the parts list are published: if you can already solder, you can build your own in a day.
- "What's the hardest part?"
- Handling such a weak signal without polluting it. Everything else — soldering, ordering the boards, assembling — is within reach of a patient person.
15 Going further
- Technical documentation: plans, assembly guide, parts list, measurements and project status.
- Interactive parts map: preamp · power supply.
And Talas
The Talas One is Talas's first object. The rule is the same everywhere: publish the plans and the measurements, and build nothing you can't take back control of.
Applied to a service rather than an object, it gives Veza: a platform where musicians share their tracks, work on them together and sell their work. Self-hosted, no data resale, commission shown upfront.
talas.fr — the mic · veza.fr — the platform
A short glossary
| Word | What it means |
|---|---|
| Capsule | The part that hears: a thin diaphragm that vibrates with the air. |
| Preamp | The circuit that makes a too-weak signal usable. |
| Board / PCB | The small plate the components are soldered onto. |
| Schematic | The plan of what connects to what. |
| Layout | The plan of where the parts and tracks actually sit. |
| P1, P3 | The first prototype (April 2026) and the current revision of the boards, working since November 2026. |
| Pickup pattern | How a microphone hears depending on direction: front only (cardioid), all around (omni), front and back (figure-of-eight). |
| Open hardware | The plans of a physical object, freely published — like a recipe. |
| dB (decibel) | A unit of comparison. +6 dB: "clearly louder". −12 dB: "clearly quieter". |
| dB SPL, dB(A) | The level of a sound in the air. 5 dB(A) is far quieter than a silent room; 127 dB SPL is louder than a rock concert right by the stage. |
| Faraday cage | A metal enclosure that shields against electrical interference. |
| Distortion | What a device adds to the original sound. The less there is, the more faithful it is. |
| Phantom power | The current the audio interface sends down the cable to power the microphone. |
16 The audio interface: Talas Delfin (revision B)
The microphone needs an audio interface able to supply clean phantom power and digitise both channels. Talas Delfin is the open-hardware USB-C audio interface (10 inputs / 6 outputs) developed alongside it in KiCad 9 and OpenSCAD.
Its revision B architecture packs six stacked boards into a compact
enclosure: the 6-layer audio mainboard, the 6-layer XMOS XU216
usb module, the 4-layer power board, and three front/rear I/O
mezzanine boards.
audio mainboard (6 layers, 235 × 150 mm): preamps, converters and LDOs.
audio board.
usb board (6 layers): XMOS XU216 and USB-C.
usb board.
power board (4 layers).
power board.