Explanation to do with image bitrate(bpp) allocation

I would like someone to help me with an explanation of how bit allocation is done in the DCT image compression technique. I have tried to read a few papers but i haven't found one which clearly explains how the bit allocation is done, other than just stating that a given image has a certain bitrate,say 1.25bpp. I have also read where people say that bit per pixel is given by : Bpp = numbers of bits/number of pixels.

but the above equation does still not explain how the bit allocation, during the whole DCT compression process, is done. Any helpful links or explanations here are welcome. Thank you.