OpenGL FreeType: weird texture - c++

Question

After I have initialized the library and loaded the texture I get http://postimg.org/image/4tzkq4uhl.
But when I added this line to the texture code:
std::vector<unsigned char> buffer(w * h, 0);
I get http://postimg.org/image/kqycmumvt.
Why is this happening when I add that specific code, and why does it seems like the letter is multiplied? I have searched examples and tutorials about FreeType and I saw that in some of them they change the buffer array, but I didn't really understand that, so if you can explain that to me, I may handle this better.
Texture Load:
Texture::Texture(FT_GlyphSlot slot) {
glPixelStorei(GL_UNPACK_ALIGNMENT, 1);
glGenTextures(1, &textureID);
glBindTexture(GL_TEXTURE_2D, textureID);
int w = slot->bitmap.width;
int h = slot->bitmap.rows;
// When I remove this line, the black rectangle below the letter reappears.
std::vector<unsigned char> buffer(w * h, 0);
glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA, slot->bitmap.width, slot->bitmap.rows, 0, GL_LUMINANCE_ALPHA, GL_UNSIGNED_BYTE, slot->bitmap.buffer);
glGenerateMipmap(GL_TEXTURE_2D);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_REPEAT);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_REPEAT);
}
Fragment Shader:
#version 330
in vec2 uv;
in vec4 tColor;
uniform sampler2D tex;
out vec4 color;
void main () {
color = vec4(tColor.rgb, texture(tex, uv).a);
}

Answer 1

You're specifying GL_LUMINANCE_ALPHA for the format of the data you pass to glTexImage2D(). Based on the corresponding FreeType documentation I found here:
http://www.freetype.org/freetype2/docs/reference/ft2-basic_types.html#FT_Pixel_Mode
There is no FT_Pixel_Mode value specifying that the data in slot->bitmap.buffer is in fact luminance-alpha. GL_LUMINANCE_ALPHA is a format with 2 bytes per pixel, where the first byte is used for R, G, and B when the data is used to specify a RGBA image, and the second byte is used for A.
Based on the data you're showing, slot->bitmap.pixel_mode is most likely FT_PIXEL_MODE_GRAY, which means that the bitmap data is 1 byte per pixel. In this case, you need to use GL_ALPHA for the format:
glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA, slot->bitmap.width, slot->bitmap.rows, 0,
GL_ALPHA, GL_UNSIGNED_BYTE, slot->bitmap.buffer);
If the pixel_mode is something other than FT_PIXEL_MODE_GRAY, you'll have to adjust the format accordingly, or potentially create a copy of the data if it's a format that is not supported by glTexImage2D().
The reason you get garbage if you specify GL_LUMINANCE_ALPHA instead of GL_ALPHA is that it reads twice as much data as is contained in the data you pass in. The content of the data that is read beyond the allocated bitmap data is undefined, and may well change depending on what other variables you declare/allocate.
If you want to use texture formats that are still supported in the core profile instead of the deprecated GL_LUMINANCE_ALPHA or GL_ALPHA, you can use GL_R8 instead. Since this format has only one component, instead of the four in GL_RGBA, this will also use 75% less texture memory:
glTexImage2D(GL_TEXTURE_2D, 0, GL_R8, slot->bitmap.width, slot->bitmap.rows, 0,
GL_RED, GL_UNSIGNED_BYTE, slot->bitmap.buffer);
This will also require a slight change in the shader to read the r component instead of the a component:
color = vec4(tColor.rgb, texture(tex, uv).r);

Answer 2

Solved it. I added the following to my code and it works good.
GLubyte * data = new GLubyte[2 * w * h];
for( int y = 0; y < slot->bitmap.rows; y++ )
{
for( int x = 0; x < slot->bitmap.width; x++ )
{
data[2 * ( x + y * w )] = 255;
data[2 * ( x + y * w ) + 1] = slot->bitmap.buffer[x + slot->bitmap.width * y];
}
}
I don't know what happened with that particular line I added but now it works.