Soundaktor editor released. Tutorial on how to use inside.

First of all: What is a Soundaktor?
Modern Volkswagen Group cars (VW, Audi, SEAT, Škoda, CUPRA) use a system called the Soundaktor to shape what the engine sounds like inside the cabin. Rather than relying purely on the natural sound coming through the firewall, a small actuator (essentially a speaker bolted to the firewall or dashboard) plays back a carefully crafted engine sound on top of the real one.

There are three variants:

— Internal Soundaktor: a speaker inside the cabin, producing the in-car sound
— External Soundbooster: a speaker in the engine bay, amplifying the sound outside the car
— AVAS (Acoustic Vehicle Alerting System): found on electric and hybrid cars, legally required to emit a sound at low speeds so pedestrians can hear the vehicle approaching

The Soundaktor is controlled by a small dedicated ECU. Like any ECU, its behaviour is defined by firmware that contains audio samples and a set of tables telling the unit how loud to play them, depending on what the engine and driver are doing.

Some people hate the soundaktor, because it's fake. But. instead of hating it I decided it's a cool thing to customize, so I created a tool that allows you to edit and finetune the Soundaktor in your car, so you can make it sound like you want it to sound.
You can find it at mqbtools.nl/soundaktor/

What does this app do?
This app lets you read, inspect, preview and modify that firmware file. You can load an original firmware from your car (or pick one of the preset example files), adjust the sound maps, and write the modified firmware back. No special hardware or software licenses needed, just a browser.

Loading a firmware
You can either:

— Upload your own file: a raw .bin binary or an .odx file (the standard diagnostic container format used by VW Group tools). FRF files are not supported at this moment.
— Pick an example: the app ships with a collection of known-good firmwares from various cars, organised by generation and type (Internal / External / AVAS)
The app automatically detects which generation of firmware you loaded. To avoid compatibility issues, make sure you select a firmware that fits with your car.

Found on photo: 5F0907159

GEN2 vs GEN2.5
VW Group has released two major hardware generations:

GEN2 is the older format (with a 57 kB data file), found in most MQB-era cars up to roughly 2019. It is fully editable in this app.

GEN2.5 is the newer format (with a 96 kB data file), found in newer MLB Evo and late MQB Evo platforms (Audi A6 C8, Q7, TT Mk3, CUPRA etc). It has a more complex structure with additional gain axes. The app currently shows GEN2.5 firmwares in read-only mode: you can inspect everything, but saving changes is not yet supported until I figure out some of the values that are currently not fully known yet. DO NOT TRY TO FLASH GEN2.5 ONTO GEN2 HARDWARE AND VICE VERSA. YOU WILL BRICK YOUR HARDWARE.

What's inside the firmware?

The VIN table

This table maps your specific car variant (identified by a code stored in the ECU) to a driving mode profile. Different trim levels, markets, or equipment lines can point to completely different sound profiles, which is why a Golf GTI and a Golf GTE share ECU hardware but sound different. In the car table, you can find a button to unlock an entry to be suitable for all VINs and engine codes, which disables the VIN- and engine lock. This makes it possible to use a VW Soundaktor in an Audi, which is pretty cool because some brands have cheaper modules than others :-)

Profiles and layers

A profile represents a driving mode: think Comfort, Normal, Sport, etc. Each car can have multiple profiles assigned per car variant and gear selector position, defined in the VIN table.

Within each profile there are multiple layers. Think of these as independent audio channels playing simultaneously, mixed together. Each layer has its own sample assigned and its own set of gain curve references. This is how the system creates a richer, more complex sound than a single sample could provide.

Each layer also has individual settings for Gain, Pitch Shift, Phase, and Delay, allowing fine control over how the layers blend together.

Gain curves

The Soundaktor doesn't just play the sound at a fixed volume. It constantly adjusts the playback level based on what the engine is doing, using a set of lookup tables called gain curves:

M-Gain (torque/load): how loud to play based on engine load. Under light throttle the sound stays subtle; under full load it gets louder.
N-Gain (RPM): adjusts volume based on engine speed. Typically the sound builds as revs rise.
V-Gain (speed): adjusts based on vehicle speed. Often used to fade the sound out at motorway speeds.
P-Gain (pedal, GEN2.5 only): adjusts based on accelerator pedal position, independent of the actual load.
Each curve is visualised as an interactive graph, which you can edit, export, and import. If you want to modify your firmware to work with a differen RPM range, or to work at higher speeds, you can change the X-axis scale here.

Audio samples

The firmware embeds one or more short audio clips: raw PCM audio, the same format as a .wav file. These are the actual engine sounds and the waveshape determines the timbre of the sound. You can listen to each sample in the browser, download them as .wav files, and even replace them with your own recordings. Note: these samples are VERY short and have a fairly low sample, so don't expect to play Daft Punk on your Soundaktor.

Sound simulator

Here you can preview the sound. Select a car from the VIN table, select a driving mode, and press "Preview". The simulator will go throught the gears like a normal car would. If you want, you can change the shift point, top speed and how long the car will stay in a certain gear.

Sometimes it might sound pitched too high. This is because I don't yet know the initial pitch of the samples, which might cause some weird sounds in the preview. You can modify the "tuning" level so it actually sounds like your car. This will be improved in future versions.

Underneath the graphical equaliser, you can unfold the various gain curves, and see how this curve influences the sound.

Note: this simulator is just an educated guess on how the hardware interprets the sound settings.

And when you're done… just press "Save changes to new file". You can export to ODX(only when using an example or uploading your own ODX) or BIN file (which will give you a raw binary file). If you like the software, please consider placing a donation, buying me a coffee or so. This project took a lot of time and effort to finish, and the webserver isn't running for free either.

Enjoy editing the firmware! I am sure some of you will make some awesome creations with this!