{"id":77,"title":"Pro Audio Source Separation vs GPU Audio SDK Module","bg_image":{"url":"https://eap-spaces.fra1.cdn.digitaloceanspaces.com/storage/newsfeed/article/bg_image/77/b45c-image.jpg","collage":{"url":"https://eap-spaces.fra1.cdn.digitaloceanspaces.com/storage/newsfeed/article/bg_image/77/collage_b45c-image.jpg"}},"type":"Newsfeed::Article","preview":"The best solutions on the market vs our new SDK Module ","views":0,"content":"\u003ciframe class=\"ql-video\" frameborder=\"0\" allowfullscreen=\"true\" src=\"https://www.youtube.com/embed/QMZBlEc9Vhc?showinfo=0\"\u003e\u003c/iframe\u003e\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\u003cp\u003eGPU Audio have a new SDK Module to share with you! Our Source Separation Module is available now as part of our SDK, find all the details here: \u003ca href=\"https://gpu.audio/sdk\" rel=\"noopener noreferrer\" target=\"_blank\"\u003ehttps://gpu.audio/sdk\u003c/a\u003e\u003c/p\u003e\u003cp\u003eIn this video we take a look at what Source Separation is, then compare it to the best non-realtime platforms, then to the leading Real-time Source Separation software. We also cover some use cases for the module - which is not limited to any one application. The machine used for this demo is an NVIDIA 4090 Desktop.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eOverview of the Technology\u003c/strong\u003e\u003c/p\u003e\u003cp\u003eStart by reviewing the motivation and core principles behind our technology here: \u003ca href=\"https://github.com/gpuaudio/gpuaudio-sdk?tab=readme-ov-file#motivation-and-introduction\" rel=\"noopener noreferrer\" target=\"_blank\"\u003ehttps://github.com/gpuaudio/gpuaudio-sdk?tab=readme-ov-file#motivation-and-introduction\u003c/a\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e1. SDK Installation\u003c/strong\u003e\u003c/p\u003e\u003cp\u003eDownload and install the SDK by following the instructions here: \u003ca href=\"https://github.com/gpuaudio/gpuaudio-sdk/blob/main/installation/main.md\" rel=\"noopener noreferrer\" target=\"_blank\"\u003ehttps://github.com/gpuaudio/gpuaudio-sdk/blob/main/installation/main.md\u003c/a\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e2. Build Initial Example Projects\u003c/strong\u003e\u003c/p\u003e\u003cp\u003eWe recommend beginning with simple examples such as the Gain and FIR processors to verify your setup: \u003ca href=\"https://github.com/gpuaudio/gpuaudio-sdk?tab=readme-ov-file#example-projects\" rel=\"noopener noreferrer\" target=\"_blank\"\u003ehttps://github.com/gpuaudio/gpuaudio-sdk?tab=readme-ov-file#example-projects\u003c/a\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e3. Evaluate a Target Module (Realtime Source Separation – HSTasNet)\u003c/strong\u003e\u003c/p\u003e\u003cp\u003eFor example, you can test realtime source separation based on the HSTasNet neural network:\u003c/p\u003e\u003cp\u003e\u003ca href=\"https://github.com/gpuaudio/gpuaudio-sdk?tab=readme-ov-file#real-time-sound-source-separation-rt3s\" rel=\"noopener noreferrer\" target=\"_blank\"\u003ehttps://github.com/gpuaudio/gpuaudio-sdk?tab=readme-ov-file#real-time-sound-source-separation-rt3s\u003c/a\u003e\u003c/p\u003e\u003cp\u003ePlease make sure to test this module within your DAW environment by building it as a plugin (details here): \u003ca href=\"https://github.com/gpuaudio/gpuaudio-sdk/blob/main/installation/main.md#use-plugins-with-gpu-acceleration-in-the-daw-1\" rel=\"noopener noreferrer\" target=\"_blank\"\u003ehttps://github.com/gpuaudio/gpuaudio-sdk/blob/main/installation/main.md#use-plugins-with-gpu-acceleration-in-the-daw-1\u003c/a\u003e\u003c/p\u003e\u003cp\u003eFor a detailed technical explanation of how we re-implemented HSTasNet using the GPU Audio Platform, please refer to: \u003ca href=\"https://github.com/gpuaudio/gpuaudiosdk/blob/main/Guides/RealtimeSourceSeparation.pdf\" rel=\"noopener noreferrer\" target=\"_blank\"\u003ehttps://github.com/gpuaudio/gpuaudiosdk/blob/main/Guides/RealtimeSourceSeparation.pdf\u003c/a\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003ePerformance Results - Using the following configuration\u003c/strong\u003e:\u003c/p\u003e\u003cul\u003e\u003cli\u003eSTFT size: 1024\u003c/li\u003e\u003cli\u003eOverlap: 512 samples\u003c/li\u003e\u003cli\u003eSample rate: 44.1 kHz\u003c/li\u003e\u003cli\u003eBuffer size: 512 samples.\u003c/li\u003e\u003c/ul\u003e\u003cp\u003eWe observed the following realtime ratios (RTR - how many times faster than real time the system processes one audio buffer):\u003c/p\u003e\u003cul\u003e\u003cli\u003eCPU version (Eigen): 0.18× RTR = ~5x times slower than realtime\u003c/li\u003e\u003cli\u003eMeta PyTorch (GPU): 0.58× RTR = can't process one buffer on time still\u003c/li\u003e\u003cli\u003eGPU Audio SDK: 6.95× RTR 🏆.\u003c/li\u003e\u003c/ul\u003e\u003cp\u003eThis represents approximately:\u003c/p\u003e\u003cp\u003e~12× performance improvement over standard GPU (PyTorch) 🚀\u003c/p\u003e\u003cp\u003e~36× performance improvement compared to CPU implementation 🔥.\u003c/p\u003e\u003cp\u003eThe GPU Audio SDK is free to download here: \u003ca href=\"https://gpu.audio/sdk\" rel=\"noopener noreferrer\" target=\"_blank\"\u003ehttps://gpu.audio/sdk\u003c/a\u003e\u003c/p\u003e","pathname":"pro-audio-source-separation-vs-gpu-audio-sdk-module-77","human_date":"10 Mar 2026","read_time":"5 Minute Read","category":"","related_items":[{"id":1,"title":"ADOBE MAX LOS ANGELES 2022","bg_image":{"url":"https://eap-spaces.fra1.cdn.digitaloceanspaces.com/storage/newsfeed/event/bg_image/1/efa3-image.jpg","collage":{"url":"https://eap-spaces.fra1.cdn.digitaloceanspaces.com/storage/newsfeed/event/bg_image/1/collage_efa3-image.jpg"}},"type":"Newsfeed::Event","preview":"GPU Audio joins our partners, AMD, to demonstrate how GPU Audio plugins can be used to enhance workflows of post production workstations. Find us at the AMD booth, nearby Meta and other great compa...","views":0,"pathname":"adobe-max-los-angeles-2022-1","human_date":"18 Oct 2022","human_time":null,"event_type":null,"registration_link":""},{"id":2,"title":"AES NYC 2022","bg_image":{"url":"https://eap-spaces.fra1.cdn.digitaloceanspaces.com/storage/newsfeed/event/bg_image/2/6528-image.jpg","collage":{"url":"https://eap-spaces.fra1.cdn.digitaloceanspaces.com/storage/newsfeed/event/bg_image/2/collage_6528-image.jpg"}},"type":"Newsfeed::Event","preview":"October 19-20, Time TBA","views":0,"pathname":"aes-nyc-2022-2","human_date":"19 Oct 2022","human_time":null,"event_type":null,"registration_link":""}]}