PS Audio vs NAD M12

The M12 uses what was previously the Zetex ZXCZM800 chip, which was purchased by CSR CSRA6601, and is now part of Qualcomm.

The chip has some interesting features such as a 35 bit digital volume control and a use of PWM in its processing. There is a high degree of IMD distortions that was tested in earlier products, but it was fairly low in level so audibility didn't make it a major issue. 

The Stellar DAC uses the ESS9018, which also has an IMD issue, but very typically a little lower in measured compared to the chips NAD uses. 

The PS Audio DSD line is entirely different and uses FPGA chips in a similar fashion to what Chord does. Many use FPGA to implement custom controls or filtering and processing, but few use them as a complete DAC solution as you must carefully construct an analog output since the FPGA chip contains none. Some of the latest Chord designs have achieved some rather excellent performance, so its feasible to use this route to get the job done, but some of the top shelf chips out there can also get right up there as well.

One thing of note is that with FPGA based designs, it is possible to improve performance with firmware upgrades. This is something that PS Audio has successfully accomplished with the DSD over the years.

The NAD does upsample to a very high sample rate and I believe the rate is to coincide with the PWM encoding scheme used.

I couldn't say what the NAD M12 uses as an output stage, but I don't believe its discrete implementation. Don't take this as negative as I have seen both good and bad implementations from the very same company on differing product lines. I would expect the M12 to have been held up to the efforts of their Master Series. 

PS Audio in the past has offered a variety of filtering options, upsampling, and the ability to defeat oversampling, or NOS mode. Download the manual for the DSD and you can read the options provided as they are some of the more extensive offered. Even though some will measure poorly, people have hoot experimenting with it and figuring if they like the distortions. Always keep in mind that not all preferences are for dead on accuracy and its why certain products exist. PS Audio tends to cater to this affection as well as the camp that prefers accuracy.

PS Audio uses a mosfet based Gain Cell in their Stellar DAC and use it to adjust the gain to avoid the use of the in built ESS digital volume control. The reason for this is to avoid truncation when attenuating 24 bit digital signals. With 16 bit, there is plenty of headroom to digitally attenuate without any real significant loss information. When I've seen this implemented in other products with measurements, it appeared to be of sound design.