GeoffPark
commented
3 years ago [AMD Official Use Only - Internal Distribution Only]
The beta-cross-platform branch of TAN https://github.com/GPUOpen-LibrariesAndSDKs/TAN/tree/beta-cross-platform/tan Compiles and runs on Windows, Linux, and MacOS using CMake.
From: Kristian notifications@github.com Sent: February 5, 2021 5:50 PM To: GPUOpen-LibrariesAndSDKs/TAN TAN@noreply.github.com Cc: Subscribed subscribed@noreply.github.com Subject: [GPUOpen-LibrariesAndSDKs/TAN] Possible changes and updates? (#12)
[CAUTION: External Email]
With new GPU's coming out with raytracing hardware it might be worthwhile to look into using that for audio. Using Vulkan compute could be another thing to look into, depending where OpenCL is headed.
Not sure if it's possible looking at samples, but cut the audio output or atleast let us choose where the output goes to, such as OpenAL-Soft, FMOD or such.
Most of the samples use C/C++ instead of "proper" C++ along with the mess of Win32 API, it would be much easier to understand and read if they were rewritten without mixing C and C++. Also keeping Win32 things in their own source and header files would prevent a drankwholebottleofrumhangover levels of headache when trying to identify which parts are relevant to TAN.
Let's not forget CMake as a cherry on top, though that can be done afterwards as a contribution.
- You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2FGPUOpen-LibrariesAndSDKs%2FTAN%2Fissues%2F12&data=04%7C01%7Cgeoffrey.park%40amd.com%7C42e73912df444567a1c408d8ca285b4a%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637481621963905874%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=LJr5JmK8oZkG9kk4FUYdkaexaKRUK8Cq6FcJZgPcdvA%3D&reserved=0, or unsubscribehttps://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FAD7CD7BWQGV7RICL4XW3HVDS5RYZFANCNFSM4XFRZN6Q&data=04%7C01%7Cgeoffrey.park%40amd.com%7C42e73912df444567a1c408d8ca285b4a%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637481621963915831%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=tjG65qcRRtwKlPbD64I9VHgHce9Wex148fBy5fdQGLU%3D&reserved=0.
DatCaptainHorse
With new GPU's coming out with raytracing hardware it might be worthwhile to look into using that for audio. Using Vulkan compute could be another thing to look into, depending where OpenCL is headed.
Not sure if it's possible looking at samples, but cut the audio output or atleast let us choose where the output goes to, such as OpenAL-Soft, FMOD or such.
Most of the samples use C/C++ instead of "proper" C++ along with the mess of Win32 API, it would be much easier to understand and read if they were rewritten without mixing C and C++. Also keeping Win32 things in their own source and header files would prevent a drankwholebottleofrumhangover levels of headache when trying to identify which parts are relevant to TAN.
Let's not forget CMake as a cherry on top, though that can be done afterwards as a contribution.