Originally Posted by alluringreality
Just as a background for anyone following along with Manni01's measurements, the HDMV and Patched versions use identical video encodes. Assuming that there are not any problems in the download or burn to disc, the main variable from my perspective is the player. I can't reproduce or explain the "large reddish rectangles within larger greenish rectangles", but I don't have any reason to at all question the color measurements.
Since relative Y for the colors was the major difference in the prior measurements, I copied the Y values for color and pasted them in a spreadsheet. They're input by rows in the order measured and split into sets based on common display settings. Columns A-G are Y for Red, Green, Blue, Yellow, Cyan, Magenta, Gray. Columns I-N are relative Y for RGBYCM. Columns P-W just compare relative Y, where P-U represent RGBYCM. For simplicity the line in bold is primarily what I'm looking at to quickly compare relative Y between the different discs.
I would tend to throw out a few of the measurements where 100% Y varies between runs, for example the first and fourth runs vary by quite a bit for the Y measured. In the good duplicate measurements there were only a couple instances where relative Y ever differed by more than about 0.2%. There does appear to be a difference in measurements between the disks that is greater than that 0.2% difference from common measurements. The most consistent difference appears to be the relative Y measured for red, which measures consistently lower on the Patched version than the HDMV.
To try to put things into perspective I included some comparison runs:
A) I measured HDMV and Patched on my electronics. Because I know my display varies with time, I only measured color to make the runs quickly. In my measurements relative Y between the discs is similar to the difference between measurements of the same disc. The main point here is just that the difference in Manni01's measures isn't necessarily directly related to the disc version.
B) I included a comparison between a correct and incorrect decode matrix I had previously taken. The difference in Manni01's measurements are not at all similar to what would likely happen if one of the discs was being decoded with an incorrect matrix.
C) I included a previous comparison between my display receiving an SD input and the display receiving an HD input. This is generally in line with the sort of difference being shown in Manni01's measures.
Thanks a lot for your time and for the information.
I agree the difference in Y percentage is small, however the difference between the two discs is consistantly of around 3 to 4 dE in gamut measures for some colors.
I don't think the difference is that bad as a 3dE error on a color is probably not visible per se.
I did try to compare the discs on my HDMI laptop, but I have recently upgraded to Win7 rtm and my video drivers are not compatible with PowerDVD, so I couldn't play the AVSHD discs. I have therefore no way to check if this behaviour is specific to my BD player or not.
However I recalibrated my PJ yesterday using an HDMV disc for the first time, (back to HDMI enhanced, brightness=-6 contrast=0), using the Display2 trained to the i1pro for all the work, gamut and greyscale) and the results are extremely interesting. The i1pro was facing the screen to take the reference measures, and the display2 was facing the PJ (diffuser on).
When I measured my existing calibration (the last "good" one, that has been posted here), my color temp was all wrong. Red was way too high. So I pushed green back up (a lot) and blue (not as much but still), aiming for a 100% white as perfect as possible as per Le Dahu's procedure.
The first result was that my RGB got instantly flat for all three colors from 50IRE to 100IRE, which I had never seen before (I had a dE for white of 0.8). 5 to 50 was not as flat, but much closer that what I had before (pre gamma corrections). I also incidentally got a substantial amount of brightness back thanks to green not being pushed down that far.
Then I revised my gamut, and realised that a few colors, especially those that showed the biggest difference between the two discs, like red and yellow, where pushed two much especially as far as hue is concerned.
So calibrating with the HDMV version allowed me to "push back the controls" and use less hue and less saturation per color, which to me feels like a good thing. I still have a dE of around 8 for blue (and 6 for magenta), but apart from cyan that I intentionally moved back on the line between blue and yellow giving a dE 2-3, all other colors have a dE under 1.
I then did a pass on gamma with Le Dahu's excellent procedure (white only), and after a single pass my gamma was almost a flat 2.22 all across. For some reason my second pass was worse than the first one, but I was getting tired so it may be human error.
I have no idea if this is because I had completely botched my former calibrations (possibly because I was using the patched version, possibly because I'm totally incompetent
), or if the small varirations between patched and hdmv are amplified when something as essential as a tweeked color temp is wrong (as you may recall there was no substantial difference between the two discs with THX and when when I used the standard 6500K color temp with my last user calibration).
The resulting picture is even better than what I had (and I was pretty happy with my last settings). Much more details in dark scenes, a more dimensional picture overall.
Until someone else measures a difference between the two discs, I'm happy to believe this is a one off due to my specific setup (possibly my player, possibly my AVR a Denon 1909?), made worse by my relative lack of experience.
But I would still encourage those who have calibrated a 750/rs20 with a patched disc (especially from a Panasonic BD playerand/or a Denon AVR) to check their calibration with an HDMV version. If they find a repeatable, significant difference on some colors, I would then encourage them to recalibrate using the HDMV version, from scratch (ie not only the gamut, but also the color temp and gamma adjustments).
I'm going to do more tests when I calibrate to SMPTE-C, especially to measure the new calibration with the patched disk. It would be intersting to see if the difference between the two discs is gone with my new calibration.
If I'm fully happy with it, I'll double check everything and post all the settings (before resuming watching films