The Official ChromaPure thread - Page 69

^^^

yea, i've got the right one... i tried that...

thanks for the idea though...

Please make sure that the Lumagen is set for 11-point grayscale adjustment. This is a classic symptom of what happens if it is set to 12 or 21-point.

Tom, it's turkey day quit working

yea tom, quit working...

yea, i checked that, it's set to 11 point... i had put it on 21 point when i first got the unit and was playing around with it, but set it back to 11 for the autocal...

is it possible that if i am using a cms that i previously set to 21 point (and subsequently changed back to 11) may be causing my issue? it's easy enough to restore the radiance to default condition and start again...

This post concerns a comparison of the D3 Pro vs the C5 OEM calibration of my LG 55LH90. This TV is a full LED back lit , ~ 86 zones, with local dimming.

The Duo HDMI video outputs connect to the LG HDMI 1 and HDMI 4 using the expert 1 picture mode in each case. Using CP Pro HDMI 1 is calibrated with the D3 and HDMI 4 with the C5.

Examining the CP reports the values for Gray scale, Gamma and color are very consistent across the board, not identical but very close.
But when playing reference recordings from the Tivo Premiere the LG LED compensated D3 Pro performance is evident. The PQ is noticeably more accurate. Skin tones spot on. Stored recordings of Pro Golf and Nascar races provide what I call reference material recorded outdoors by ESPN with well calibrated HD cameras.
Likewise studio scenes for Fox News and series like CSI NewYork provide additional PQ checks.

Tom's compensated D3 Pro is a winner in this race.

Brief note on my calibration procedure.
Picture reset on the LG, manually calibrate with Duo CMS OFF using the LG 10 point grayscale with the CP Gamma module to bring Gamma to 2.2 from 100% to 10% stim and adjusting RGB to equal values to where DE is at <2. This assures that the Duo CMS contribution will not run out of range.
Engage the Duo CMS and use CP autocal to finalize the calibration and file the reports.

Since the LG has 3 settings for Gamma and white balance the purist should run precal for each settings and pick the setting closest to D65 and Gamma 2.2. More work but worth the effort.
Also look at the settings for Wide Gamut on/off, Real Cinema on/off and True Motion off, low, high. Avoid high, that inserts alternate black frames which, given the D3 (and C5) lack of sync, would de-stabilize the readings.

Note on observer.
After Cataract surgery on both eyes in the past two years, this observer has excellent light transmission and color perception. The implants have essential zero light attenuation, so much so that when the LG backlight is set to 50 or higher the observer has to use sunglasses to watch the TV.

I am wanting to do a calibration 'Chromapure autocal' with my meter facing the Projector, is there a distance the meter should be positioned from the PJ, or meter reading that I should work too or not go above?

Also, I already have offsets for my meter since I had this re-calibrated. Can I still use these when creating my reference & field meter corrections?

Look here:
http://www.avsforum.com/avs-vb/showp...postcount=1871
and here:
http://www.avsforum.com/avs-vb/showp...postcount=1876

Lot's of info on that thread

Ciao Tom,
some questions:

- how does the ChromaPure Standard + Display 3 bundle performs with Kuro plasmas, overall & on low light measurements

- Assuming my laptop has an HDMI out, does Chromapure supports sending patterns to the TV, via the HDMI port? If so, what kind of patterns will Chromapure use? The small ones? I recall about some problems on plasmas, requiring small patterns to prevent power dispersion on large patterns....

Very well. The Kuro plasma is my main test display.


Yes. It will output either full fields or small windows. It is user-selectable.

Thanks for the prompt reply!

More:
- Does ChromaPure ideally follows your famous step-by-step guide for calibration?

- As Kuros don't sport a full featured CMS, does ChromaPure requires special handling with Kuros?

- When dealing with Kuros, do you use a special software such as ControlCal, to access the ISF interface? If so, do you leave the C9 interface untouched, to avoid troubles?

- To your knowledge, is there any trade-in offer for Display 2 owners, with ChromaPure + Display 3?

- Semi-OT: In your opinion, after how many month/years the Display3 will need a factory recalibration?

Thanks again..... on the way to ChromaPure!

The workflow is very similar.


If you want a CMS you'll have to invest in a Radiance or a Duo.


Not generally. The only real advantage is increased light output and my display puts out 119 cd/m2, which is fine.


No.


Unknown. The meter was released only in July.

You will need to buy the ControlCAL and a serial to usb adaptor to access the KURO's 10-Point Gamma and to save a lot of hours from your life accessing all the options very fast. You will open the 2 ISF Modes (Day-Night) with ControlCAL.

KURO has 1D Color Mamagement. Only For Hues, but it's not working very well, it needs care, lol

I own both already

in reference to my earlier contouring issue... success!!!

- reset radiance to defaults...
- manually set 100 ire...
- autocal greyscale... get good greyscale...
- autocal gamut... get decent color...
- autocal gamma... get good gamma, greyscale goes to crap...

at this point, this is still the same results as before...

- run autocal greyscale again...

now greyscale is back to being good, and the gamma only moved a bit (i.e. stayed close to corrected)...

- run autocal gamma...

now greyscale is still good, and gamma is good... and the actual picture is good...

anyone care to take a shot as to why? it is my understanding that the autocal isn't an iterative process?

color me happy, but curious...

Hi Chris,

Don't know if this is relevant (I have a Duo not a Radiance) but when manually correcting a 'way out' grayscale / gamma I have always had more success (in terms of needing to 'revisit' that is) by doing gamma before grayscale.

Usual workflow suggests grayscale then gamma but as you have verified there is not much gamma change when individual changes made to stimulus(s) during grayscale correction.

Just an observation from 'manual' not 'autocal' calibration.

Peter

Yes, order seems to make a difference and you get different results at differetnt times. It reminds me of NTSC... just kidding. I never though I would need more control of autocal.. I have the DUO so I don't have full fields. I let Chromapure do autocal with the DUO patterns and then I adjust with the AVS full field patterns. Actually I do a baseline with the display first, then I do grayscale, gamut, then gamma. I don't really understand what rules autocal uses. It seems pretty unpredictable. Gamma is hard to do on my display by hand and autocal just makes a mess of it most times. I don't necessarily blame autocal.

I would like to be able to autocal from the pattern of my choice.. and just do one things at a time.

For example I would like to use the AVS Blue 75% field pattern and just autocal it. If the current tartget parameter is dE 2.0 for autocal I want to be able to tell ChromaPuer to work it until it hits dE 1.0 or whatever.

Currently autocal it not very transparent and it seems to spend a tremendous amount of time going in the wrong direction where the dEs get higher and higher and higher and highter then it bails out. If it starts at dE 2.5 on blue for example, why should it be going out to 10 and then continuting on to 20. The other night it got so far out it had some weird green color up calibrating blue and starting from about 2.5dE it went all the way inot the high 50s. It seems like the goal seeking logic is not optimal. The end result was okay but was pretty darn close to where it started so all the gyrations were clearly wasted.

Tom, do you see any drawback or limitation in using ChromaPure->Computer->HDMI->TV generated patterns? As it speeds a lot iterating, you know.....

@gtgray
You make some good points (and suggestions) Gary,

I have had good success using autocalibrate as a base for manual fine correction.
This works extremely well with grayscale and gamma but for color the autocalibrate gives me the best compromise I can get between dE accuracy and S & M clipping accuracy.

The autocalibrate zeros all selected headings and always starts from there rather than accepting current value and improving on it.

Perhaps a future version could allow the 'nearly right' value to be finely tuned automatically.
At least then it is more likely to spend time refining rather than 'range finding'.

This is actually similar to what occurs in a manual calibration. Grayscale and gamma are interactive. After doing grayscale and then gamma, you typically need to check grayscale again.

In the next version of auto-cal we will address this.

BTW, on our test display I don't see the extreme interactivity you see, but this will vary among displays.

To be precise, the auto-cal target is 2.0 for grayscale and 1.75 for color. In any case, I guess I would like to know why this is something you would want? Grayscale errors less than 2.0 are generally considered invisible, so why lengthen the auto-cal process (perhaps substantially) in order to achieve something that makes no difference? This seems to run counter to your concerns about blue.


This can happen especially when the color in question is undersaturated to begin with.

You are at the mercy of your graphics card, drivers and settings, OS, media playback software and all of the other variables that come with HTPCs.

I gave up on trying to get my intel laptop to work with my tv because it refused to send anything other than RGB (0-255).

I was suspecting this...very risky..... I wonder why you read articles encouraging this as a boosting factor... probabily those persons aren't professional calibrators..... I used to use my PS3 to calibrate.... I suspect that also this choice is "at risk"....
For a home enthusiast, buying a special purpose device would be too much (IMO)..... any advice?

I too used my PS3 as my main source device but then I noticed when I used those calibrated settings with my Xbox 360, my blacks looked overly red. Turns out PS3 has a large dip in the red at around 30% in the grayscale...

Oh my! There's no other choice than a color generator! However, calibrating the DVD/BR player you will be using should be okay, as the color shift will be compensated with the calibration. What you lose is the possibility to use the same settings on other inputs....
Your thoughts?

Tom, I can understand people wanting to get the best accuracy out of their software if the meter they are using has the ability to measure this accuracy.

Otherwise what is the point of buying that meter and having LUT's that give it that accuracy?

There is a certain satisfaction in reducing the dE as low as possible even if the change is invisible.

Perhaps also one can have some confidence that if dE is seen as tolerance then if dE is at zero there is no skew towards a visible innacuracy caused by the display's variable output.

We like the autocalibrate and are looking forward to the improvements that we know you have in the pipeline.

I have recently got a WDTV Live to use as a cheap portable generator. Its far from perfect, but it is more affordable than any dedicated signal generator on the market.

In your experimentation, did you try one of the usb to hdmi adapters?

Didn't know if it allows you to select 444 or 422 through its software settings.

No and I'm not familar with any such products. I assume it has its own EDID chip? That is what I believe the problem to be. My laptop isn't getting back the required information from my TV's EDID and is locking down the settings.

Windows 7 is an overzealous drag sometimes.

Interested too, that would overcome driver/os colour manipulation... maybe we should move this discussion elsewhere....

You know, there's got to be a way to do this without buying a bunch of electronics that may or may not work.