Uploaded October 2026 | Updated November 2026, 4 days ago
Results will be posted on X https://x.com/GregKocis/status/2110842412876660736
Watch Next ➤ Official Bootlegs | Evolving Songs • Human-Written • AI Re-Creations: youtube.com/live/JjJbUF68WMw?si=w2y_rsq6lxvMiIxG
Google Flow Music generated a CLA Vocals-style audio plugin that processes vocals with EQ, compression, reverb, delay, and stereo widening, and this video compares its sound with the original Waves CLA Vocals plugin.
The experiment tests whether listeners can distinguish the Google Flow Music-generated plugin from the original CLA Vocals using vocal audio samples from one of my own songs. The video compares the processed vocals, examines the generated plugin’s controls and limitations, and demonstrates how its sound and behavior differ from the original. It also explores whether AI-generated audio plugins can be customized with additional features, such as auto gain and saturation, simply by describing the desired changes. The goal is to evaluate the practical results of generating a working audio plugin with AI rather than assume that it is an exact replica of an existing commercial product.
What's covered in this video:
- Generating a CLA Vocals-style plugin with Google Flow Music: The video begins with an experiment in which Google Flow Music generates an audio plugin inspired by Waves CLA Vocals, producing a working interface that can process audio.
- Comparing the original and AI-generated vocal plugins: Two versions of a vocal recording are played for listeners to compare, with the challenge of identifying which sample uses the original CLA Vocals plugin and which uses the Google Flow Music-generated version.
- Testing dry vocals before adding effects: The initial listening comparison focuses on dry vocal processing without reverb, delay, or pitch-based stereo widening, providing a starting point for evaluating the plugins.
- Hearing the results on an original song: The vocal samples feature my own song, “No More,” allowing listeners to hear the processing in a real songwriting context rather than relying only on synthetic test audio.
- Examining the generated plugin's controls: The video demonstrates the plugin interface and explains differences in its bass, treble, and compressor sections, including the limited number of operating modes available compared with the original CLA Vocals.
- Understanding why the settings do not match directly: The controls and parameter ranges differ between the two plugins, making it difficult to recreate identical settings and meaning that the comparison focuses on the resulting sound rather than identical knob positions.
- Investigating compression and gain reduction: The generated plugin's compressor requires substantial adjustment to avoid excessive processing, and its interface includes a gain reduction control that is not present in the same form on the original CLA Vocals interface.
- Comparing delay and stereo widening features: The generated plugin offers limited delay timing options, including eighth-note and quarter-note settings, alongside reverb, delay, and pitch-based stereo widening functions.
- Considering additional audio plugin features: The experiment raises the possibility of adding auto gain, saturation, or other processing options, as well as removing individual effects and using separate bus plugins for more flexible vocal processing.
- Discussing AI-generated plugin customization: Because the plugin was created by describing the desired functionality, the video considers how natural-language instructions might be used to modify an audio tool, while acknowledging that requested features may or may not work as expected.
- Inviting listeners to identify the original plugin: Viewers are encouraged to share their guesses about which vocal sample uses the original Waves CLA Vocals plugin, with the answer to be revealed on X.
Mentioned in this video: Google Flow Music, Waves, Waves CLA Vocals, CLA Vocals, audio plugins, AI-generated plugins, EQ, equalization, compression, compressor, gain reduction, reverb, delay, pitch processing, stereo widening, auto gain, saturation, bus processing, dry vocals, vocal processing, vocal recording, No More, Greg Kocis, Songs by Greg, YouTube, X, eighth-note delay, quarter-note delay, natural-language plugin generation, music production, audio processing, plugin customization, AI-generated audio tools
0:00 Intro: Can AI Recreate CLA Vocals?
0:47 Comparing the Vocal Samples
1:56 The Original Song and Plugin
2:04 CLA Vocals Controls and Differences
2:22 Is It an Exact Plugin Clone?
2:53 Compressor Behavior and Gain Reduction
3:14 Delay Settings and Final Thoughts
#howtomakemusicnow
Results will be posted on X https://x.com/GregKocis/status/2110842412876660736
Watch Next ➤ Official Bootlegs | Evolving Songs • Human-Written • AI Re-Creations: youtube.com/live/JjJbUF68WMw?si=w2y_rsq6lxvMiIxG
Google Flow Music generated a CLA Vocals-style audio plugin that processes vocals with EQ, compression, reverb, delay, and stereo widening, and this video compares its sound with the original Waves CLA Vocals plugin.
The experiment tests whether listeners can distinguish the Google Flow Music-generated plugin from the original CLA Vocals using vocal audio samples from one of my own songs. The video compares the processed vocals, examines the generated plugin’s controls and limitations, and demonstrates how its sound and behavior differ from the original. It also explores whether AI-generated audio plugins can be customized with additional features, such as auto gain and saturation, simply by describing the desired changes. The goal is to evaluate the practical results of generating a working audio plugin with AI rather than assume that it is an exact replica of an existing commercial product.
What's covered in this video:
- Generating a CLA Vocals-style plugin with Google Flow Music: The video begins with an experiment in which Google Flow Music generates an audio plugin inspired by Waves CLA Vocals, producing a working interface that can process audio.
- Comparing the original and AI-generated vocal plugins: Two versions of a vocal recording are played for listeners to compare, with the challenge of identifying which sample uses the original CLA Vocals plugin and which uses the Google Flow Music-generated version.
- Testing dry vocals before adding effects: The initial listening comparison focuses on dry vocal processing without reverb, delay, or pitch-based stereo widening, providing a starting point for evaluating the plugins.
- Hearing the results on an original song: The vocal samples feature my own song, “No More,” allowing listeners to hear the processing in a real songwriting context rather than relying only on synthetic test audio.
- Examining the generated plugin's controls: The video demonstrates the plugin interface and explains differences in its bass, treble, and compressor sections, including the limited number of operating modes available compared with the original CLA Vocals.
- Understanding why the settings do not match directly: The controls and parameter ranges differ between the two plugins, making it difficult to recreate identical settings and meaning that the comparison focuses on the resulting sound rather than identical knob positions.
- Investigating compression and gain reduction: The generated plugin's compressor requires substantial adjustment to avoid excessive processing, and its interface includes a gain reduction control that is not present in the same form on the original CLA Vocals interface.
- Comparing delay and stereo widening features: The generated plugin offers limited delay timing options, including eighth-note and quarter-note settings, alongside reverb, delay, and pitch-based stereo widening functions.
- Considering additional audio plugin features: The experiment raises the possibility of adding auto gain, saturation, or other processing options, as well as removing individual effects and using separate bus plugins for more flexible vocal processing.
- Discussing AI-generated plugin customization: Because the plugin was created by describing the desired functionality, the video considers how natural-language instructions might be used to modify an audio tool, while acknowledging that requested features may or may not work as expected.
- Inviting listeners to identify the original plugin: Viewers are encouraged to share their guesses about which vocal sample uses the original Waves CLA Vocals plugin, with the answer to be revealed on X.
Mentioned in this video: Google Flow Music, Waves, Waves CLA Vocals, CLA Vocals, audio plugins, AI-generated plugins, EQ, equalization, compression, compressor, gain reduction, reverb, delay, pitch processing, stereo widening, auto gain, saturation, bus processing, dry vocals, vocal processing, vocal recording, No More, Greg Kocis, Songs by Greg, YouTube, X, eighth-note delay, quarter-note delay, natural-language plugin generation, music production, audio processing, plugin customization, AI-generated audio tools
0:00 Intro: Can AI Recreate CLA Vocals?
0:47 Comparing the Vocal Samples
1:56 The Original Song and Plugin
2:04 CLA Vocals Controls and Differences
2:22 Is It an Exact Plugin Clone?
2:53 Compressor Behavior and Gain Reduction
3:14 Delay Settings and Final Thoughts
#howtomakemusicnow










