D3D11 Vs DXVA2

what is the difference between them?

They are very similar. DXVA2 uses D3D9 video functionality, while D3D11VA obviously uses D3D11. There is no difference in the formats that they can decode.

D3D11 works the same as DXVA2 copy-back. There is one exception, and this is where D3D11 has a benefit, is when you use it together with MadVR. In that case it can automatically switch to native mode for improved performance.

If I select D3D11VA in LAV Decoder, then which type of decoding it will do (native or copy-back)?

pyumredkar wrote:If I select D3D11VA in LAV Decoder, then which type of decoding it will do (native or copy-back)?

copy-back unless you use MadVR where it can automatically switch to native if needed for improved performance. You need copy-back for most post-processing though(if you use them), since post processing can't modify the frames stored in VRAM, and only main RAM. D3D11VA uses the same DXVA 2.0 data structure for decoding so there's no difference there.\

Also i think LAV Video automatically uses Native(when using Automatic) unless you force a specific accelerator, then it will force copy-back. If you have no performance issues, i recommend copy-back to prevent quality reduction. D3D11VA only works on Windows 8+ by the way, because it's part of the DirectX 11.1 update which was introduced in Windows 8 and never ported back to 7.

got it. thanks