Room correction - what device works best?

Simplistically speaking above 250Hz (room's transition frequency - look that term up) the frequency response at the listening position is dominated by the speaker. Below 250Hz it is dominated by the room.

The idea behind speaker correction is that you are measuring and correcting for the on axis speaker anomalies only. The correction applied can be higher resolution than room correction, which necessarily must use multi-point measurements and time domain smoothing to prevent over correction.

Room correction you cannot separate speaker from room, so if you apply "full range room correction" like Dirac then you are correcting the combined response. The best full range algorithms are very gentle in how they correct above the transition frequency. I only like Trinnov and Dirac at the moment.

Like Kal from Stereophile found in his review, the best results would be combining DEQX speaker correction with Dirac room correction. However parametric EQ, when applied in the bass as DEQX intended will get you just as good results as Dirac, it just takes a lot more time and expertise to get it dialed in.

There's a subtle difference between room and speaker correction.

Room correction you put the mic at the listening seat and you measure speaker and room combined.

Dirac only offers room correction (though a very good algorithm, and one I really like).

The DEQX separates out speaker and room correction into separate processes.

The speaker correction is done by measuring the speaker nearfield and windowing out the room response (or measuring outside). Then you take the mic and put it in the listening position like you would do with Dirac.