I wouldn't say that the Dirac was "brighter" because it actually sounded smoother. I've said a couple times that the Datasat/Dirac had a very "class A" sound to it- I think this is evidenced in the trace above 4k; there the variations in amplitude are significantly smaller than with Trinnov or Raw/untreated. The lower amplitudes in the midbass region was noticeable and I think contributed to my desire for more bass.
Trinnov made more corrections to the midrange area from 500-4k, but both Dirac and Trinnov seem to treat that passband equally in magnitude- just in different areas.
The Datasat/Dirac trace can be adjusted by bumping the midbass passband in the target curve.
Overall, the Dirac optimization felt more mellow, less grating at high volume- more like listening to a live concert (amplified) that uses tube gear at the source (guitar amps, etc.) The Trinnov wasn't fatiguing, but certainly had more edge, sounds more solid state. My only explanation is that I'm more sensitive to the fact that the trace is not as smooth as with the Dirac, but then, that's exactly what my training tells me as well. Gerry Lemay talks about smooth transitions being more important that overall flat-line when it comes to frequency response, and I believe that Floyd Toole says something similar in one of his white papers: that the average listener likes a gentle slope down as frequencies increase- and that transitions between passbands are what's ultimately important.
hopefully, i'm not grossly misquoting.
Here's the Dirac and Trinnov traces overlaid to show their differences.