Okay, I've spent a couple of hours watching patterns...
I've decided to post here and not in the AVS HD thread because it looks like the issue is linked to (amplified by) the settings of the PJ.
First of all I checked a bit further and my BD-50 does detect the AVCHD format on the DVD-R, so that's not the issue.
I used the i1pro, HDMI standard, brightness and contrast at zero, iris at -5.
I recalibrated the i1pro before each run to avoid any thermal drift (bringing the plate to it, careful not to move the tripod/meter). I usually just press "hide" in my bat cave, but I thought I'd do things as they should be done to rule out any variable (one could say my method is actually better as there is no contact with the meter/tripod, but hey, I'm not going to do it again
...). The meter was facing the screen. I didn't use the display2 as I didn't want to have the meter training factor in the equation. So take values below 20IRE for the greyscale with a pinch of salt.
I did two runs per disc for each serie (except for the last two) so that we have a control read. I did 10 steps greyscale plus gamut (100% stim) for each run, again except the last two (just gamut, and only one run per disc).
The whole session took place over a couple of hours, and at no point was the tripod/meter moved. I gave the PJ 30mn to warm up before I started.
Here are the results:
I started with THX OOTB (with untweaked 6500K preset). There was NO significant difference between the patched and HDMV version (around 1dE, max 2dE, which is what one could expect between two runs of the same measurements).
I then tried my User1 preset (last calibration, tweaked 6500K, tweaked gamma as per my last posted measurements). Sure enough, the difference between the two versions was back, up to 3-4 dE, depending on the color (EDIT: I only checked the gamut data between all the runs, as this is where I first noticed the difference. I took the greyscale info in case it can help those who know better to understand what's going on).
I then tried with a normal gamma, and the difference was still there.
I then tried with the 6500 untweaked color temp, and most of the difference was gone...
I then checked my last calibration using the internal ramps, and I realised it was quite badly clipping on green
I thought maybe when we get it wrong the differences between the two versions are amplified, so I typed in my last "good" calibration. The clipping on green was gone, but the difference between the versions was still there. For these last two runs I measured only the gamut and did only one run (I was getting tired).
So I'm posting here because it is pretty inconclusive. There is a repeatable difference between the two discs, and it's pretty significant (IMHO), but not on THX, and not on the 6500K standard color temp preset, so YMMV, and it could be a 750 specific issue.
I'll try to check with a pattern generator at some point, but for now I'm going to use the HDMV disc, as if one of the two is right it's most likely to be HDMV.
I'm getting a bit tired to watch patterns, so whoever wants to take over from here is welcome
I had to split the archive it two parts. Unzip both parts in the same folder and sort by date/time, that way you'll see them in the order I measured them.
AVS HD Variation Part1.zip 243.759765625k . file
AVS HD Variation Part2.zip 418.2119140625k . file