Pages: [1]   Go Down

Author Topic: A few odd things (and questions) about Spectraview II on PA  (Read 3561 times)

WombatHorror

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 299

1. I tried to use the Broadcast video target and well it works fine and lets you chose all sorts of gamma options sRGB, 2.2, HDTV, etc. BUT....
for some bizarre reason it seems to have the primaries locked into native gamut! AFAIK, Adobe After Effects is just about the only video program
that is actually color-managed so this seems very, very bizarre. Watch HDTV, blu-ray, etc and it all looks radioactive (if you calibrate using the sRGB mode they look fine though since that does use the sRGB/REC 709 primaries).

Is this a bug? Or is broadcast video target something other than what I think it is?


2. sRGB emulation mode target is locked into gamma 2.2 (and yet the default factory program for sRGB mode as revealed by Multiprofiler locks in a sRGB tone response).
Why don't they allow any choice for gamma for sRGB emulation mode in SVII? And if they say that is what broadcast video target it for, well that also is oddly locked into
native gamut not sRGB/REC 709 gamut. And then why is sRGB mode not locked into sRGB tonal response curve?

a second bug that they use 2.2 instead of sRGB? or that they don't allow a choice of whatever gamma you want (since sometimes you might prefer 2.2 instead of sRGB and for movies in a dark room maybe even gamma 2.3) if broadcast video mode really is supposed to be native gamut.


3. as far as I can tell SVII sRGB mode locks in factory coordinates for primaries but doesn't use the probe to adjust them and yet it does appear to use the probe to measure them and to store that into the ICC profile. That seems like a weird mix of deciding the factory presets are more accurate than the probe can measure (since it locks them in) and yet not since for color-managed programs it does write what the probe finds into the ICC profile. So why not an option to have the probe adjust them internally so non-managed stuff also works ideally? Or if the probe doesn't measure them as well as the factory measurements then why write what it measures into the ICC by default for some probes?

Oddly multiprofiler lets you adjust the primaries (at leas the xy if not Y coords) but doesn't directly work with the probe and SVII works directly with the probe but gives no option for the probe to internally shift the primaries.

4. how come multiprofiler lets you adjust primary xy but not Y coordinate if most other profilers say the primary luminances are very important (the nec docs appears to downplay the importance of primary luminance)


It's very interesting to use a monitor with a 14bit 3D LUT since it seems like it just measures the WB of white and sets it and then doesn't need to check any other shades along the gradient and it just relies on the internal engine to set the rest, same for gammas. It doesn't appear to use the probe to measure and adjust and build a big table but it seems to base things off of the factory preset information and just uses the measurements to make the report on how well things worked out. The presets+color engine do seem to do a very good job at creating various tonal reponse curves. If you pick gamma 2.2 and measure with an external program the lines does lie exactly long 2.2 and the sRGB response curve also seem to be right on track as well. It also lets it maintain smooth gradients when switching down to sRGB mode. A monitor with a poor color-engine or only external calibration that is wide gamut would dump off shades in sRGB simulation mode.
OTOH, there is a bit of a WB disagreement between what the factory sRGB mode's preset white point and what the NEC puck calibration using SVII gives though. I haven't tried to check yet which is more accurate. They are not radically different, but they are different. So either the factory preset white point for D65 isn't really quite D65 or the NEC puck is a little bit off.
« Last Edit: July 04, 2010, 03:55:33 pm by LarryBaum »
Logged

WombatHorror

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 299
A few odd things (and questions) about Spectraview II on PA
« Reply #1 on: July 04, 2010, 04:21:27 am »

Not sure if this last part necessarily has to do with the NEC PA profiles or not but:

I also noticed an odd thing. 8bit gradients when viewed in Photoshop have a sudden drop off after about gray 16 and then the last 15 shades are suddenly all much darker (although many are distinguishable) but when viewed with say Irfanview with color-management turned off everything smoothly goes down to 0 with no sudden drop at all. In either case there is no banding so long as the monitor is set to the same tonal response curve as the format you tag the gradient as.  


Logged
Pages: [1]   Go Up