Smart Mixing and Mastering

As we build out AudioStack, we're interested in making it seamless for anyone to produce professional quality audio. Based on an extended research and development project in digital signal processing and machine learning, AudioStack is now able to offer an auto mixing service to improve your audio productions by standardising and enhancing the quality of generated and uploaded audio.

Our Mastering service delivers audio content by mixing together speech, fx, music and other assets.

Our Speech (Voices) and Sound Design (Sound Templates) are very diverse in quality, frequency ranges and dynamics furthermore, we are constantly working on expanding the sources for these.
We wanted to improve and standardize (master) the quality of our mixes when:

• Our voices are from several providers and vary in tonalities, dynamics and frequency ranges.
• Music and sound effects (sfx) come at different loudnesses (RMS) and have diverse tonal qualities.

The Auto Mixing Service analyses each component in each mix and applies an individualized mastering signal chain to enhance the output quality. These signal chains will be different based on the music, sfx and voice characteristics.

Fundamentally this is an out of the box solution that improves sound quality by applying an audio treatment parametrised to enhance each individual mix.

Code Example

If you tried out the code example from the Sound Design page, you'll have noticed that we added a few different mastering presets. If you haven't already, try saving the following code as a Python file and running it from your Terminal using by typing Python3 followed by the file name, and pressing enter.

from uuid import uuid4
import audiostack
import os

audiostack.api_key = "APIKEY" ## fill me in!
SCRIPT_TEXT = """
<as:section name="main" soundsegment="main">
Audio mixing and mastering enhance the overall quality of a music or audio production by balancing and refining its elements, resulting in a polished and professional sound that captivates and engages the audience.
</as:section>
"""

names = ["Elaine","Edward"]# "narration_gia", "yolanda", "jeff", "narration_terra", "narration_jeremy", "promo_joe", "Wren",  "jollie", "aspen", "renata", "christopher", "jenny", "Bronson", "Adele", "sara", "kellie", "hailey", "michelle", "geena", "monica"]
presets = ["musicenhanced", "balanced", "voiceenhanced"]
templates = ["friendly_electronica","eighties_chill","funky_summer"]

script = audiostack.Content.Script.create(scriptText=SCRIPT_TEXT, scriptName="test", projectName="ams_tests_2")        

for name in names:
    # Creates text to speech
    speech = audiostack.Speech.TTS.create(
            scriptItem=script,
            voice=name,
    )
    for template in templates:

        for preset in presets:

            mix = audiostack.Production.Mix.create(
                speechItem=speech,
                soundTemplate=template,
                masteringPreset=preset,
            )
            print(mix)
            uuid = uuid4()

            mix.download(fileName=f"V4_{name}_{template}_{preset}")

            print(mix)

Notice how musicenhanced, balanced and voiceenhanced sound different? If you listen carefully, you might hear that "musicenhanced" focuses on the music in the sound, "balanced" is somewhere in between, and "voiceenhanced" focuses on the voice. This allows you to mix using our AI engine a range of voices, and have studio quality audio produced at scale

📘

If you are uploading a sound template that has already been mixed and mastered (such as an existing advert), you can choose to apply mastering only to generated speech...

To do this, simply set the masteringPreset to be "voiceonly".