Albeit there are utilities to measure performance for each of the three, I’d
rather voice a few subjective impressions/opinions from my experiences with
the GeForce 6800.
For 2D NVIDIA has in my opinion the “snappiest” driver of them
all; differences with the competition might be labeled as miniscule, yet I’m
still able to see a difference. This particular monitor can run at 1600*1200
with a quite acceptable 85Hz refresh rate, while it’s highest resolution/refresh
rate is at 2048*1536@75Hz. I run the desktop at 1280*960*32@100Hz because I
can still read effortless text from a reasonable distance. At 1600 I already
need to move closer. Anything above that would require potentially closer contact
between my nose and the screen and performance isn’t what I’d expect
from a normal Windows XP environment. For others running at resolutions past
1600 it might be perfectly enough, but I’m a tad demanding obviously.
The control panel and especially nView are packed with useful and interesting
features; Digital Vibrance for instance caters for quicker monitor calibration
if one is a fan of a bit more vivid colours in either 2D or 3D. I’ve
still to see on the other hand the real use of Image Sharpening; once you enable
it even with a very conservative setting it adds aliasing to images or motion
pictures, similar to the way some image editing programs do.
One thing I really like are the added transparencies the user can enable for
dragging windows or the taskbar, the windows repositioning and dialog box repositioning
features in nView.
Judging video/DVD playback seems even harder. Video playback is highly relevant
to the quality and compression ratio of the file used. In general I couldn’t
notice anything annoying or different here.
For DVD playback I normally use PowerDVD XP and I had the time watching a
couple of movies while writing this review. For that I used the Lite-On DVD-RW
LDW-811S. Firing up the task manager during playback showed a CPU usage that
fluctuated between 15-22% with HW acceleration disabled, and 11-15% with it
enabled (here the drive isn’t irrelevant; CPU usage is a notch higher
on the Hitachi GD7500 DVD-rom).
PowerDVD also comes with a DirectDraw test; I couldn’t detect any problems,
yet performance is obviously capped by the refresh rate of the currently used
display resolution. It wasn’t any different with past accelerators I’ve
used from NVIDIA; I can’t know though why it happens or if it is directly
related to playback performance or not.
Overall I’ve no serious complaints about DVD playback quality, yet I
wouldn’t say either that there aren’t competing solutions out there
that do just a bit better in that department. I would provide screenshots,
yet the resulting images don’t do real time image quality any justice.
Granted I tried only PowerDVD’s image capture function (which saves in
bitmaps), but the screenshots turned out way blurrier than in real time.
Please note that it may very well be that the application cannot make proper
use of the accelerators true capabilities and needs either an according update
or an entirely new version. From what I’ve read so far there are quite
a few tricks one could theoretically pull with SM3.0 capable accelerators and
mostly in the adaptive de-interlacing department (amongst others), meaning
that both future display drivers as newer versions of playback applications
might get enhanced with added functionalities.