[WiP] madVR support in MP1 (11 Viewers)

[Kotik]

Ok 124 regarding 3D is same. No change

Good 3D MadVr detection. No failure . Tv Goes 3D, madvr goes 3D and image 2D.
After moving mouse, image 3D. I'm going mad with this problem. I don't know if that's only my Palit gt730 GDDR5... I'll remove it and test tomorrow with Intel hd4000.

Can u try with this setting CHECKED?

 

[kenwonders]

Yep 124 should go on top of 123
I think i will recreate a full installer, the test this night and this morning was OK with Nvidia.
On HTPC, first 3D failed before but not now it seems but i have see sometimes even with MPC-HC that the first start wasn't always OK maybe a bug with Nvidia lol.

I'll try the full 124 too then. Sorry if I've missed Seb but what are your settings? Are you set to return back to non-3d after playback?

 

[Sebastiii]

Yep 124 should go on top of 123
I think i will recreate a full installer, the test this night and this morning was OK with Nvidia.
On HTPC, first 3D failed before but not now it seems but i have see sometimes even with MPC-HC that the first start wasn't always OK maybe a bug with Nvidia lol.

I'll try the full 124 too then. Sorry if I've missed Seb but what are your settings? Are you set to return back to non-3d after playback?

I have try with and without and both works :

or

But we can run into a timing issue, it was need to add 3 seconds between starting on new madinit so maybe it's too short for you.

 

[kenwonders]

If you could pinpoint where (and the branch is up to date) then I could play about with those 3 seconds? Or if you're feeling really generous throw up a 10 second super-safe-but-tedious version?

 

[Sebastiii]

It's the MP1-4781 branch and the code is in directshow filter
In \DirectShowFilters\DirectShowHelper\source\dshowhelper.cpp (Sleep(3000); line code) 3000 ms

BOOL MadInit(IVMR9Callback* callback, DWORD width, DWORD height, DWORD dwD3DDevice, OAHWND parent, IBaseFilter** madFilter)
{
  // Delay for 3 seconds on init to clear all pending garbage from C#
  Sleep(3000);
  Log("MPMadDshow::MadInit 1");
  m_RenderPrefix = _T("mad");

  m_pDevice = reinterpret_cast<LPDIRECT3DDEVICE9>(dwD3DDevice);
  Log("MPMadDshow::MadInit 2()");

  m_madPresenter = new MPMadPresenter(callback, width, height, parent, m_pDevice);
  Log("MPMadDshow::MadInit 3()");
  m_pVMR9Filter = m_madPresenter->Initialize();
  Log("MPMadDshow::MadInit 4()");
  if (m_pVMR9Filter)
  {
    Log("MPMadDshow::MadInit 5()");
    *madFilter = m_pVMR9Filter;
    Log("MPMadDshow::MadInit 6()");
  }

  if (!madFilter)
    return FALSE;

  return TRUE;
}

 

[Sebastiii]

If you could pinpoint where (and the branch is up to date) then I could play about with those 3 seconds? Or if you're feeling really generous throw up a 10 second super-safe-but-tedious version?

I can make a 10 secs version ofc

 

[Sebastiii]

Here we go

 

[kenwonders]

I can't get 3d working in 123+, it's always 2d even though mode is changing correctly. I'll try to find the last working..

 

[Sebastiii]

I can't get 3d working in 123+, it's always 2d even though mode is changing correctly. I'll try to find the last working..

Weird

 

[Sebastiii]

Do you have log ? @kenwonders