I forgot to inform you that when you perform the measurements with the two different mode you will try a difference of 3 dB (the method 1 is higher of 3 dB than the method 2). This is because the method 2 is a measurement that, for how is performed, don't consider the image frequency, because you perform two noise measurements with the image placed over, and, since the Y method is a ratio of two noise the image is eliminated. In the method 1 usually you consider the gain sending a singol tone without its image, instead you will measure the noise with its image. So you will try a results 3 dB higher.