View Full Version : Understanding black level (article)


glennchan
08-17-08, 10:30 PM
Understanding black level (http://www.broadcastreferencemonitors.com/understanding/black-level.htm)

The article looks into the following issues behind black level:
* External reflections and glare
* Inherent black level of the monitor
* Picture content and flaring/halos
* Surround effects
* Black Frame Insertion

It's written from the perspective of a broadcast / production standpoint but the same concepts would apply to home theatre. I hope you find the article useful!
*Unfortunately I didn't put information about plasma in there as I haven't really played around with plasma.

Sweet Humble Pie
08-18-08, 01:18 PM
Couple of issues with the article.

1. Modern black frame insertion does indeed change the backlight during the "black" phase.

2. It is not possible for eCinema to get 15000:1 static/ANSI on an LCD panel without dynamically changing the backlight, either globally or locally. Such LCD technology simply doesn't exist outside Samsung's research lab (and possibly not even there).

3. A 9x9 ANSI checkerboard doesn't have higher average luminance level than the other ANSI checkerboard patterns. In fact it is ~10% lower than a 3x3 ANSI and ~6% lower than a 5x5 ANSI.

glennchan
08-18-08, 03:47 PM
Hi sweet humble pie,

Thanks for the feedback.

1- Which "modern" black frame insertion technique would you be talking about?

1b- I should have clarified that backlight scanning (where the backlight does go off and a black bar is scanned from top to bottom) doesn't have the same effect on black level. I'm not sure how backlight scanning affects black level as I haven't seen information on it.

2- The eCinema DPX has a SEQUENTIAL contrast ratio of 15,000:1. I apologize if that wasn't clear as most of the time, contrast ratio figures are for sequential contrast.

Martin Euredjian (the designer behind the DPX) has stated that the display will actually go lower but he does not have a measurement device that can measure the larger range. In any case, it doesn't make too much sense to measure lower as in practical situations you probably won't get 15,000:1 (e.g. due to flare, glare).

The display definitely doesn't flare like a CRT does.

*I've not measured the display myself. But I don't see any reason not to believe it.
**To my eyes, it's the best blacks in any display I've seen. Slightly better than CRT, better than the field emission prototypes I saw at NAB. Regardless of how you want to measure the display, it definitely has impressive black performance. This doesn't mean it's the ultimate display (e.g. viewing angle ain't perfect, like other LCDs).

2b- The monitor does not use local dimming of the LEDs.

3- A 9x9 ANSI checkerboard doesn't have higher average luminance level than the other ANSI checkerboard patterns. In fact it is ~10% lower than a 3x3 ANSI and ~6% lower than a 5x5 ANSI.
My comment was that the checkerboard pattern has a higher average luminance than real-world scenes. It wasn't in reference to other checkerboard patterns.