It is not that easy to do a fair comparison between an unmodded Sega Master System 2 and a modded Sega Master System 2 in terms of the direct audio out from the VDP, because I guess any RF demodulator / VCR / TV will add its own character to such a process and therefore skew any tests.
So instead, I am comparing the audio line output on the Sega Master System 1 to the direct audio line from the VDP on the Sega Master System 2. The audio line output on the SMS1 is found on an unmodified console in the form as part of a 5 pin din switch, while the SMS2 direct output from the VDP must be done manually as a modification.
Here is an audio example of the SMS1 and the SMS2 playing the exact same sequenced material. This audio file has not been normalised, and both consoles have been recorded with the same gain setting. The silence at the start is recorded silence.
The peak level during musical playback for the SMS1 is at -5.4 dBfs.
The peak level during musical playback for the SMS2 is at -4.7 dBfs.
The peak level during recorded silence for the SMS1 is at -53.2 dBfs.
The peak level during recorded silence for the SMS2 is at -69.0 dbfs.
A normalised portion of recorded silence for the SMS1 when analysed as a sonogram looks like so:
A normalised portion of recorded silence for the SMS2 when analysed as a sonogram looks like so:
Both consoles share noise in the same spectrum areas - predominantly up to about 1KHz and also that very obvious spike at about 15.6KHz. Since both are normalised, it is fair to say that the actual character of the noise is similar, but it tends to cover a larger band for the SMS1.
However, since there is a difference of around 15.8 dB between the two, the practical difference is much larger in terms of audio quality and musical output.
Although the SMS2 Prosound modification is not that much louder than the SMS1 line level output, the difference in the noise floor of around 15.8 dB makes the mod worth doing.