I’ve been working on my W5000 for the last few days, trying various approaches to calibration and finally reached a conclusion on what produces the best picture for me. It may work for you too, so I thought I’d share it.
Equipment used for or the purpose, was an i1 pro, software was Calman 4 and source was a WDTV Live, outputting 1080/24p YCbCr 4:4:4 through HDMI. AVS forum and Spears & Munsil patterns were used.
I tried both Whisper/Normal lamp and Brilliant Colours On/Off settings and found Normal lamp, BC off gives the overall best picture on my 100”, 1,3 gain, Stewart Studiotek ST130 screen.
I went through various workflows to find what works best and I ended up with the following:
1) First of all, I went through adjusting the basic parameters: I picked 7,5 IRE for video levels, and deactivated Dynamic Black (as calibration is almost impossible with that on), to reactivate it again when everything finished. I also deactivated Brilliant colours, set Overscan to 0, lamp to Normal, Luma and Chroma transmission to 0 (Luma can be set to 1 btw, when projected material is too soft).
2) Afterwards, I entered ISF menu and adjusted Brightness and Contrast. I ended up with B=58 and C=44, which produced a usable video range of 17-244. I left Colour/Tint at 50/0 for the moment, and set Sharpness=6 and CT=Normal. I also set gamma to 2,4, as I have a dedicated room with good light and reflections control.
3) Then, I entered Factory menu, deactivated CCA and measured CCA off values with Normal CT. I chose Normal because I didn’t want to ruin another set of Warm CT Biases/Gains I had created, to use with lamp at Whisper; so you may as well start with Warm -or anything else, as a matter of fact. The colourspace my W5000 can produce, is shown bellow:
This is important, not only because it provides the necessary values to enter into the respective CCA fields (which is the next thing to do), but also defines the boundaries within which we can move x,y during calibration; moving outside them, would cause disturbing side-effects, like distorted low-saturation points. So, looking at my data, we can tell right away that Blue’s y is a little off on this W5000, in an unfixable way: we can only shrink, not expand colourspace. As we’ll see later, there will be some post calibration deltaH and deltaC errors for Blue; not too grave though –and besides Blue is the colour with the least perceptible errors to the human eye, so not a real problem here.
The actual CCA off x,y values I came with (as an average of three measuring runs), is this:
W R G B
x 293 657 304 145
y 321 319 620 63
4) After entering the Measured values (just x,y -left Y at default values of 1000, 80, 436, 56), I started tweaking gamut through the Desired CCA values, so that all colours had correct x,y, leaving Y CCA values at their default, for the moment. When I finished that, I looked at the colours luminances and at their deltaL errors. The one that had the largest deltaL and a negative luminance value at the same time (meaning it’s the weakest of all), was Red.
So, next step was to go and adjust Colour, so that Red’s relative luminance increased and moved just a little on the positive side of the chart (meaning that Red’s deltaL minimized too). The reason behind this, is that we can’t surpass the point to which one colour clips, so the only option is to lower the stronger colours, to match the weakest one -not vice versa. The correct setting was Colour=54.
5) Afterwards, I started tweaking Y for all six colours, until they matched White and had their deltaL minimized too. A second run followed, with special attention to minimize deltaL for primaries (RGB) and deltaH for secondaries (CMY), as these are the most important factors to each colour-category. I left deltaC were they may be. The final values are:
R G B C M Y W
x 633 295 149 218 305 414 301
y 333 603 64 307 149 506 315
Y_ 985 665 570 671 1019 973 1000
and the produced gamut is this:
6) Grayscale was next. I entered Factory menu and used RGB gain/bias to fix Normal CT RGB and overall balance at 30% and 80% levels respectively. That produced a smooth RGB and overall gamma of 2,4 (except for a fairly small hiccup at 50%) and 6500 K colour-temperature across the range. The values are:
CT Normal modified Gain/Bias 511 511 514 / 502 502 493
and here are the diagrams:
(I rechecked gamut and grayscale, as changes to one affect the other, with the minor tweaks needed to smooth everything out, already been updated in the two tables above.)
7) Finally, I set manual Iris at 2 (I had it at 19 while calibrating, for more accurate readings), reactivated Dynamic Black and started watching some of my reference videos.
What I saw, was the most pleasing picture I’ve seen so far from my W5000: sharp yet artifact-free, colour-balanced, gamma-correct, contrasty, punchy and bright enough for my application (12,15 fL with iris at 2), truly inky blacks and great shadow detail. So, it was time and effort well spent!
In time I’ll try to further refine the BC feature, as I think there’s great potential in that too; potential that can be used with ambient light or very large screens, as it outputs about 50% more white light than BC off does. Any ideas and suggestions on that, are welcome!
I’d like to seize the opportunity and thank all the people here, who share their knowledge and experience on the W5000 -some more so- and keep this thread alive throughout all these years. Kudos guys!
Edited by Max Headroom - 8/31/12 at 7:55am
