Question to CalculateDefaultBitrate in sample_utils.cpp

[boukhalfa__zakaria]

I have a question regarding the function: 'CalculateDefaultBitrate' in sample_utils.cpp

mfxU16 CalculateDefaultBitrate(mfxU32 nCodecId, mfxU32 nTargetUsage, mfxU32 nWidth, mfxU32 nHeight, mfxF64 dFrameRate)
{
    PartiallyLinearFNC fnc;
    mfxF64 bitrate = 0;

    switch (nCodecId)
    {
    case MFX_CODEC_AVC :
        {
            fnc.AddPair(0, 0);
            fnc.AddPair(25344, 225);
            fnc.AddPair(101376, 1000);
            fnc.AddPair(414720, 4000);
            fnc.AddPair(2058240, 5000);
            break;
        }
    case MFX_CODEC_MPEG2:
        {
            fnc.AddPair(0, 0);
            fnc.AddPair(414720, 12000);
            break;
        }
    default:
        {
            fnc.AddPair(0, 0);
            fnc.AddPair(414720, 12000);
            break;
        }
    }

    mfxF64 at = nWidth * nHeight * dFrameRate / 30.0;

//....
}

1. What happes when my input framerate is higher then 30? We want to use this calculation function for framerates higher then 30.
2. What are the X and Y values of the used PartiallyLinearFNC?

[Mark_L_Intel1]

Hi Zack,

For #1, I did a simple experiment and have the following code:

    unsigned long nWidth, nHeight;

    double dFrameRate, at; 

    nWidth = 640;

    nHeight = 480;

    dFrameRate = 30.0; 

    at = nWidth * nHeight * dFrameRate / 30.0;

the results: dFrameRate = 30.0, output = 307200; dFrameRate = 31.0, output = 317440

So the final result should be the linear interpolation of the at after you applied PartiallyLinearFNC::at() function

For #2, are you asking how the function PartiallyLinearFNC::at() works?

Mark