AudioCortex is a new innovative sound browser app for macOS that magically categories your samples & loops using artificial intelligence.
Making music with samples is very popular, especially in beat-oriented productions. Select the sample, drag it to the desired drum pad and you are off you go. This way of working has two main disadvantages: firstly, you need a lot of storage space. High-resolution samples take up a lot of memory. The second follows the first.
Big sample collection also requires proper management. Otherwise, you lose track of the files, and you quickly become a sample messy. Ok, always better than a messy at home. In addition, you often don’t know which sounds are hidden behind these files especially if they bear names like new sound 1, 2, 3… But there is new hope for all sample collectors.
AudioCortex
AudioCortex is a new mac-only audio app with which you can automatically categorize your sample content. It uses state-of-the-art machine learning (AI) to analyse your audio. You can import your files simply by drag and drop. It supports “aif”, “caf”, “wav”, “mp3”, “mp4”, “aac” and “Apple Lossless” formats.
Once analyzed, the app magically arranges the samples in 443 trained /pre-defined categories. You can find here categories like atmospheres, bass, loops, and more. According to the developer, this list will continue to evolve over time. However, AudioCortex is not a categorizer for complete tracks. Music production, sound design, soundtrack work and general sample management is the focus here.
After the process, you can find the sounds in a simple browser that tells you all important information: the category (sounds like), filename, duration, frequency, sample rate, channels, date modified, and more.
On top, you have two super handy audio visualizers available in spectrogram and waveform view. Here you can also preview the samples in one-shot or in loop mode. Just like you know it from a classic sound browser, only more intelligent and smarter thanks to the AI-powered engine. Furthermore, you can save your sounds into collections aka lists, and drag & drop them into your projects.
First Impressions
Richard Plom, the developer sent me a code to check out the app. I did it immediately and installed it on my Intel MacBook Pro with Mojave. The app is, however, optimized for the new M1 Neural Engine, but also runs smoothly on older machines. The first results are impressive. I loaded an older sample library with a size of 1.4 GB into AudioCortex. The analysis process took a few minutes and ran cleanly without crashing. The app classified most of the samples/loops perfectly.
There were also some mistakes where wrong categorizations occurred. Like a piano sound placed in the trumpet section. The same applies to specific sounds. It can’t tell whether it’s an 808 or 909 kick or a Prophet 5 or Jupiter 8 pad sound. It’s an atmospheric or pad sounds for the app. However, this not a bug. Richard states that the app currently has a goal of 80% accuracy within the specified categories. It’s not a marketing slogan, I can confirm this. The process works and very well.
AI or in Apple’s words the Neural Engine has arrived in music production. This app is a perfect model of how we musicians can benefit from this ground-breaking technology. It brings sound browsing to the next level. Think how cool it would be if Logic Pro or Ableton Live had such a feature built-in. We can only hope that it will happen in the future.
For the current price, it is well worth a try and highly recommendable for anyone who wants to organize and browse through huge sound libraries.
AudioCortex is available now for an introductory price of $9.95 USD on the Apple Mac AppStore. It’s not available for Windows. The price will be raised in early 2021.
More information here: Iced Audio
The accuracy rate is no where near 80%, more like 30-50% depending on category. you’ll waste so much time going through bad matches.. As of 3/2022 this is a waste of money.