Re: [PATCH 0/3] MR713: Draft: mfplat: Partially implement some format description conversion helpers.

[Nikolay Sivov (＠nsivov)]

On Thu Aug 25 17:28:23 2022 +0000, **** wrote:
> Zebediah Figura (she/her) replied on the mailing list:
> On 8/25/22 12:13, Nikolay Sivov wrote:
> > +            mediatype_set_guid(media_type, &MF_MT_MAJOR_TYPE,
> &MFMediaType_Video, &hr);
> > +            mediatype_set_guid(media_type, &MF_MT_SUBTYPE, subtype, &hr);
> > +            mediatype_set_uint64(media_type,
> &MF_MT_PIXEL_ASPECT_RATIO, 1, 1, &hr);
> > +            mediatype_set_uint32(media_type, &MF_MT_INTERLACE_MODE,
> MFVideoInterlace_Progressive, &hr);
> > +            mediatype_set_uint64(media_type, &MF_MT_FRAME_SIZE,
> vih->bmiHeader.biWidth, vih->bmiHeader.biHeight, &hr);
> > +            mediatype_set_uint32(media_type,
> &MF_MT_ALL_SAMPLES_INDEPENDENT, 1, &hr);
> > +
> > +            if
> (SUCCEEDED(MFGetStrideForBitmapInfoHeader(subtype->Data1,
> vih->bmiHeader.biWidth, &stride)))
> > +            {
> > +                mediatype_set_uint32(media_type,
> &MF_MT_DEFAULT_STRIDE, stride, &hr);
> > +                mediatype_set_uint32(media_type, &MF_MT_SAMPLE_SIZE,
> abs(stride) * vih->bmiHeader.biHeight, &hr);
> > +                mediatype_set_uint32(media_type,
> &MF_MT_FIXED_SIZE_SAMPLES, 1, &hr);
> Note that it's legal for the height to be negative, indicating that the 
> video is flipped. I don't know if that means the same thing in 
> MF_MT_FRAME_SIZE (the documentation doesn't say so, at least) but I 
> think it does mean that MF_MT_SAMPLE_SIZE is broken.

That's a good point, I'll add some tests.