Question

I want to convert the output of an OpenGL Program to h264 and stream the output. I got collected most of the code somewhere and I get an output file but I have no Idea what to do with it, or if it is valid. Currently the output is just saved in file.h264.

Edit: "Global" Variables

    x264_param_t param;
    x264_t* encoder;
    x264_picture_t pic_in;
    x264_picture_t pic_out;

    x264_nal_t *headers;
    int i_nal;
    FILE* pFile;

My init function:

initX264() {
    pFile = fopen("file.h264", "wb");

    x264_param_t param;
    x264_param_default_preset(&param, "veryfast", "zerolatency");
    param.i_threads = 1;
    param.i_width = 1024;
    param.i_height = 768;
    param.i_fps_num = 30;
    param.i_fps_den = 1;

    param.i_keyint_max = 30;
    param.b_intra_refresh = 1;

    param.rc.i_rc_method = X264_RC_CRF;
    param.rc.f_rf_constant = 25;
    param.rc.f_rf_constant_max = 35;

    param.b_annexb = 0;
    param.b_repeat_headers = 0;

    param.i_log_level = X264_LOG_DEBUG;

    x264_param_apply_profile(&param, "baseline");

    encoder = x264_encoder_open(&param);
    x264_picture_alloc(&pic_in, X264_CSP_I420, 1024, 768);

    x264_encoder_parameters( encoder, &param );

    x264_encoder_headers( encoder, &headers, &i_nal );

    int size = headers[0].i_payload + headers[1].i_payload + headers[2].i_payload;
    fwrite( headers[0].p_payload, 1, size, pFile);
}

This goes in the Render function and is executed about 30 times per second:

    GLubyte *data = new GLubyte[3 * 1024 * 768];
    GLubyte *PixelYUV = new GLubyte[3 * 1024 * 768];

    glReadPixels(0, 0, 1024, 768, GL_RGB, GL_UNSIGNED_BYTE, data);
    RGB2YUV(1024, 768, data, PixelYUV, PixelYUV + 1024 * 768, PixelYUV + 1024 * 768 + (1024 * 768) / 4, true);
    pic_in.img.plane[0] = PixelYUV;
    pic_in.img.plane[1] = PixelYUV + 1024 * 768;
    pic_in.img.plane[2] = PixelYUV + 1024 * 768 + (1024 * 768) / 4;

    x264_nal_t* nals;
    int i_nals;
    int frame_size = x264_encoder_encode(encoder, &nals, &i_nals, &pic_in, &pic_out);

    if( frame_size )
    {
        fwrite( (char*)nals[0].p_payload, frame_size, 1, pFile );

    }

I got the GRB2YUV funktion from http://svn.gnumonks.org/trunk/21c3-video/cutting_tagging/tools/mpeg4ip-1.2/server/util/rgb2yuv/rgb2yuv.c

The output looks like

x264 [debug]: frame=   0 QP=11.14 NAL=3 Slice:I Poc:0   I:3072 P:0    SKIP:0    size=21133 bytes
x264 [debug]: frame=   1 QP=20.08 NAL=2 Slice:P Poc:2   I:0    P:14   SKIP:3058 size=72 bytes
x264 [debug]: frame=   2 QP=18.66 NAL=2 Slice:P Poc:4   I:0    P:48   SKIP:3024 size=161 bytes
x264 [debug]: frame=   3 QP=18.23 NAL=2 Slice:P Poc:6   I:0    P:84   SKIP:2988 size=293 bytes

On Linux file file.h264 returns data.

Was it helpful?

Solution

Your file format is bad. Nothing will be able to read that. A .264 uses annexB with headers prepended to IDR frames.

change these parameters,

param.b_annexb = 1;
param.b_repeat_headers = 1;

And delete

x264_encoder_headers( encoder, &headers, &i_nal );
int size = headers[0].i_payload + headers[1].i_payload + headers[2].i_payload;
fwrite( headers[0].p_payload, 1, size, pFile);

Finally you should be able to take your output and convert it to an mp4.

ffmpeg -i file.264 -vcodec copy -an file.mp4

OTHER TIPS

 X264 expects YUV420P data (I guess some others also, but that's the common one). You can use libswscale (from ffmpeg) to convert images to the right format. Initializing this is like this (i assume RGB data with 24bpp).

That's most likely why you're not getting any usable video. You can find more about what to do here: How does one encode a series of images into H264 using the x264 C API?

To scale:

 sws_scale(convertCtx, &data, &srcstride, 0, h, pic_in.img.plane, pic_in.img.stride);
 //pic_in is your RGB data getting fed to scale. 
Licensed under: CC-BY-SA with attribution
Not affiliated with StackOverflow
scroll top