I've got a JVC HD350 and I'm using a Lumagen HDQ to mitigate the colour gamut (not a full CMS by any means, but it does reduce the errors for the primaries at least). What I've noticed is that apart from the lower resolution, DVDs upscaled by the Lumagen seem to have more 'depth' than BluRays. I'm wondering if this is the effect of colour gamut on the two sources? I've checked and double checked the brightness and contrast settings and adjusted the greyscale to a delta of less than 3 from 20 to 100IRE. I should point out that I use a Sony DVD player for DVDs (output at 576i upscaled to 1080/50p by the HDQ) and a separate Sony BDP-S350 player for BluRays (output at 1080/60i for reverse telecine by the HDQ to 1080/24p).

I'm sometimes slightly dissapointed with the PQ on BluRays apart from the higher resolution and wondered whether it is to do with my CIE diagram being closer to (PAL) DVDs at rec 601 compared to rec709? The image often seems a bit 'flat' and lacking in shadow detail compared to DVD. I have attached the two charts, both the same response, but one is referenced to rec601 and the other rec709.

All other things being equal (correctly set brightness/contrast/greyscale) could the colour gamut be making this difference? I intend to check the gamma curve incase there is a difference between the two sources, if so I wonder what I should be looking for? Although my setup is calibrated using the 709 disc I figured I could recheck using my DVD and BD copies of DVE as it's any differences I'm interested in between sources (I much prefer the 709 disc for ease of use).

Not in my experience. I would suspect other differences like gamma, or the content itself.

Thanks Chris. I think I should concentrate on the gamma curve between the two players. I have a DVD and BD of Terminator 2, I think I'll pause them at the same scene and flick between inputs, though they could be mastered differently I suppose it might be interesting to compare them. Ironically I used the BluRay player to calibrate the PJ and just used my Lumagen's input controls to match the DVD player's brightness and contrast, so if anything BDs should be more accurately setup.

EDIT: Just some very quick checks using my Lumagen's 11 point greyscale menu (it displays the incoming IRE level for the middle of the screen). I found that my DVE DVD at 20 IRE is correct, but my BluRay player shows 19IRE. This is consistent down to 5 IRE, but corrects itself above 50 IRE. This would seem to back up the poor shadow detail of BluRays compared to DVDs that I'm seeing.

Quick fix is to try the BD player's alternative gamma setting 'Bright room' which gets the 5,10,15 & 20 spot on and the mid range ends up 1 IRE high, back to spot on for 80 and above. I'll obviously have to create two custom gamma settings in the HD350 and save them as DVD and BluRay (User 1 and User 2 presets), or try to compensate in the HDQ perhaps using it's gamma controls.

Looking at my gamma response (I'd quickly tweaked the greyscale on Friday as it was last done at 30 hours and I'm at 190 hours now on the HD350). I seem to have the opposite curve to what would be desired: High gamma (2.4) at low IREs and low (less than 2) above 40 IRE. :rolleyes:

I'm going to re calibrate my greyscale (just to tighten it up closer to D65), and then use my Lumagen HDQ to adjust the gamma to 2.2 (or as close as I can manage). I will leave the HD350 in a custom gamma (but default settings) of 2.2 and I'll check if then changing just the HD350 gamma to 2.3 gives a flat response, so I have the option of a higher gamma with a flat response.

I can then adjust the SD input on the HDQ to give a gamma of 2.2 from the DVD player too (separate settings available due to resolution). The only issue I have is that DVE doesn't seem to have enough test patterns on the DVD to do the whole gamma curve only 20/40/80/100 IRE windows. Would this be enough or is there another disc or a way to get the 709 disc into PAL DVD format that would have the full 11 steps needed to check greyscale and adjust the gamma?

Obviously I've answered my own question to original post (as Chris suggested too) in that colour gamut doesn't seem to effect the image depth and shadow detail, but gamma does. :o

It's important that you have the same gamma for each step. Software that gives you only a single or a few gamma numbers for your grayscale won't tell you the whole story. If the software you are using doesn't give you a calculated gamma for each step from 10%-90% (you can't calculate gamma for 0% or 100% unless you have a reading below 0% and another reading above 100%), you can download AccuPel's Calibration Calculator from AccuPel's website (on the firmware download page). You enter the xyY measurements for each grayscale step and the calculator will tell you what the gamma is for each step from 10% to 90% and the dE for each step. Obviously, the Lumagen processor should help you zero-in on exactly the right gamma for each step. I find most projectors and panels look right with the gamma somewhere between 2.2 and 2.3 (the better the black level and shadow detail, the better the display can handle the higher end of the range if blacks and shadows are less than ideal, stick to the lower end of the range).

High gammas make it very difficult for displays to maintain differentiation in levels in the shadows since high gammas will "compress" the differences between, say 5% and 6% white while a lower gamma will provide more separation between 5% white and 6% white so you can actually SEE the difference in a pattern (and likely in movie content).

By the way... Lumagen uses the IRE term incorrectly in their processors (and admitted to doing so when I asked). Those steps should really be labeled % white. IRE correlates an analog voltage to some specific luminance level. This term was valid for analog video displays, but in digital video, IRE is "out". You CAN manipulate the Lumagen processor to produce actualy IRE steps if you know what you are doing, but there's really no reason to do so with modern video displays. So when you see "IRE" in the Lumagen menus, you should really be thinking "% white" (for example, for consumer video sources we have 0% white represented by digital 16, but if the patterns represented IRE levels, 7.5 IRE would be represented by digital 16). With digital displays and digital video, there is no need for IRE any more and it's use is really a confusion factor because some of us know what it is and when we see something labeled IRE we have to figure out if it is really IRE or if it is really % white because the terminilogy is being used interchangeably when it really isn't interchangeable (even by experts like Lumagen).

Thanks for your comments Doug. I'm using HFCR so it does allow measurements at all 11 points, but I don't have a DVD dics that allows me to output the full 11 greyscale patterns. I do have the AVS709 disc that I can use to calibrate my BluRay player with though.

I appreciate your comments re the way the Lumagen incorrectly 'measures' IRE, but in this case it was only a relative difference that I was interested in: My DVD player gives a higher value at the low IREs, compared to the BluRay player that was 1 IRE lower on each step. It only served to confirm that there is a difference between the two sources, which is what I'd noticed watching DVDs, then BluRays. As I prefered the result using the DVD player (apart from the lower resolution of course) I need to aim for this gamma (whatever it might actually be). My 'quick and dirty' setting change in the BD player's menu has subjectively improved the result, though I need to do a full calibration and use the Lumagen to alter the gamma curve more accurately.

However I read your posts in another thread about using an Eye one LT with a LCOS projector, so I'm now wondering if it's worth me bothering at all. :(

I set everything back to default last night and started fresh. I figured this would be what an ISF calibrator would do if they came to my house. I also put the HDMI level back to 'Standard' rather than 'Enhanced' (though I had adjusted contrast and brightness to take this into account, I wanted to start from a level playing field).

I spent a long time adjusting the various contrast and brightness controls using my HD350 and the Lumagen to make sure I wasn't raising the black level, but also (and much harder to do) to make sure I could still see the 17IRE bar flashing on the AVS709 test disc (basic patterns section). I backed off the HD350 main contrast by -2 from default as I noticed the above 225 IRE bars looked slightly pink, now looking grey. I double checked these settings using the DVE discs (both my DVD and BluRay copies) to make sure the HDQ was set correctly for both sources (only one notch different as it turned out, but the HDQ 'remembers' the setting based on input res, so why not use this flexibility).

I ran out of time to get the LT out, as I'd promised my Son he could watch the BluRay of 'WallE' so the test discs went away and film night was enjoyed by all. :) I ended up watching two more films as it was looking that good. I think somewhere in my previous 'fiddling' I'd managed to crush the blacks in trying to get the black level as low as possible, but there were so many variables I can't say now where I'd done this, but it's looking great now.

The funny thing is I've now opened the iris slightly rather than using a dimmer setting and the black level doesn't really seem worse (exect maybe comparing a 0 IRE pattern, flicking between two iris settings), but the image has more depth (the very thing I started this thread looking for).

I'm going to measure the gamma curve at some point, more out of academic interest to see if it is flatter than last week's result, plus to compare the different gamma options on the HD350 (I'm using Custom set to 2.2 default values at the moment). I also wondered if I took perhaps 10 greyscale runs, each adjusted as close to D65 as my LT tells me, then I'll average the RGB settings from all 10 runs. This should help in terms of repeatability, but of course I'll still not know if my LT measures (for example) green too low or similar....do you guys think it's worth bothering with this theory?

I've also turned the colour gamut control back on in the HDQ (settings were previously arrived at using the same LT sensor), but it seems subjectively better than 'OFF' and much better than simply turning the HD350 colour control down. My settings are very close to some other HD350/HDQ owners so perhaps colour gamut settings are more 'transferable' anyway?

I agree with those who already posted. Correct brightness, and contrast setting make a large impact on depth and especially gamma.

It's surprising small of an error you can have in brightness but still have a large impact when viewing, since it raises the whole black floor, or can crush shadow detail.

Higher gamma typically adds more image depth, but unless your display is very bright, like 25+ ftL, then it's hard to go with anything above 2.3-2.4 or so. If you are around 12-20 ftL, then you can probably get away with 2.3. The risk is the loss of shadow detail, unless you do a hybrid gamma curve and bump up the shadow detail (10% white and below), but leave the rest of the curve around 2.22-2.3.

Dan

Thanks Dan. :) I think I'm running around 10fL which is below recommended, but I actually find that bright enough (was even less with the iris at 15). I can try a slight tweak to the low IRE results by using the HDQ's parametric greyscale controls: I tried dialing in a setting that worked at 1.5IRE and gave a 1.8IRE output (I could just see it's effect on the 2% bars on a DVE test pattern, so this in theory should improve near black detail without raising the black level itself), I may have this engaged last night, I'll have to check later.

Regarding my averaging of 10 runs theory: Do you think it's worth pursuing or just a waste of time (given I really happy with the image now anyway)?