Well, here's the SINAD graphs of my RIAA from both MM and MC inputs and lowest gain setting. I use 5 mV input signal for MM and 0.5 mV for MC (as used in the reviews here).
I find < 40 dB of gain something that would not be normally needed with MM and that's why the four selectable gain steps are 40 dB, 43 dB, 46 dB and 48 dB. MC signal path just adds another 20 dB on top. In the case of MM input, the 1st stage is TI OPA1656 which has very low distortion and current noise, whereas voltage noise is not the lowest possible. However, with audio analyzer, source resistance is very different from a typical MM cartridge which favors input stages that have low voltage noise but may have poor current noise. In the real world situation, low current noise is very important since otherwise MM cartridge inductance and resistance boost noise floor up quite a bit due to current noise of the input stage. With SINAD reading of ~72 dB, my design would be right in the middle of the chart presented here.
As you can see, the SINAD figure is noise dominated and in this case, particularly noise < 300 Hz. This is what OPA1656 datasheet also shows for voltage noise. A-weighting increases the SINAD figure to 86 dB at 5 mV input.
Measuring MC input noise with an analyzer is a whole different story. Since both typical analyzer and MC cartridges are low impedance sources, your measurement results are very much in line with the real world performance. The challenges with MC lie totally elsewhere and that is external noise. When input signal is only 0.5 mV or 500 uV, noise components could be 80 dB below measurement source generator (or even almost 90 dB below like in the example here) and would still show clearly above noise floor. That is 0.5 mV / 10000 = 50 nV! In the case of my measurement setup with QA403, I could not yet get it perfected in my hobby room. Clearly the system picks up weak noise signals from nearby consumers of electricity that appear as 150 Hz noise spikes and other higher harmonics of 50 Hz AC (I'm located in Finland where AC is 230 V / 50 Hz). These noise signals change size if I move the setup around but fortunately they don't affect the SINAD reading much. In this case, the MC input SINAD is ~70 dB which I believe is pretty good.
I'm pretty happy with the result and performance, especially distortion figures and RIAA EQ accuracy are good. Additionally there's no electrolytic caps that would eventually wear down. Note that the left channel shows a bit elevated distortion but this is not real, it is caused by the QA403 output buffer which for some reason has slightly better performance in the right channel. Swapping the QA403 outputs also swaps the distortion figures so it is actually the analyzer that limits distortion performance slightly.
SINAD from MM input:
SINAD from MC input:
