Re: 14 bit Luts.
Re: 14 bit Luts.
- Subject: Re: 14 bit Luts.
- From: "tlianza" <email@hidden>
- Date: Fri, 22 Oct 2004 10:35:27 -0400
Hi to all,
I read the following exchange:
A person I asked at Eizo (Photokina booth) where the 14 Bit should be
used didn't had a meaningful answer. I'm still looking for an answer,
but the internal LUT is _not_ 14-Bit, even if it would be 14-Bit I
doubt that interpolating an 8-Bit DVI-Signal to 14 Bits will show a
difference between a 10 Bit LUT.
Best regards
Peter Karp
Very good question, Peter. I don't suppose anyone from Eizo monitoring the
list could clear the confusion?
/***********************************************/
Although I do not work for Eizo, I do have an understanding of the
technology that is being used here, so I thought that I would make some
generic comments. These comments should not be interpreted as any sort of
direct response about any Eizo product, per se, but rather a general
response.
The look up tables inside the LCD play a slightly different, but very
important role, than the traditional tables found in a graphics card lot.
Unlike a CRT, which has a monotonic response to voltage (actually
current...), the LCD response is derived from a drive circuit that is
designed to minimize many of the basic problems with the native LCD response
function, which has a very high native gamma with respect to the traditional
gamma 2.2.-2.8 found in a CRT. Typical drive circuits in these devices
include a series of adjusted sub ranges. In many panels, these sub ranges
are adjusted using resistive voltage dividers and piecewise linear fits
through out the output dynamic range of the display. This will often result
in visual banding when critically viewing grey ramps. High quality panels,
like the Eizo display, derive the drive function through a continuous, table
driven, driver. Eizo, as a company, is acutely aware of quality and in
their high end panels: as I understand from my briefing from them, EVERY
PANEL is individually measured , characterized, and corrected in it's
response on the assembly line. The reason for the need for more bits comes
from the fact that the table is essentially doing two jobs: correcting the
physical "defects" in the native response, and generating the desired
display response. The need for greater than 10 bits comes from the high
slopes required to get the panel to respond in one region and the very low
slopes required to maintain subtle highlight detail.
The look up table inside the LCD device can be thought of a concatenation
of two tables: one for system response and one for physical display
non-linearity correction. The result is a display that will display 8 bit
data with no quantization artifacts and slope errors at points of
discontinuity with below the threshold of vision. Keep in mind that this
process must work for three separate channels and the visual interaction
between these channels is also very complex. Anyone who thinks that they
can realistically do something with these tables using traditional
calibration methods is probably going to be very disappointed. The
capabilities of these new panels place new challenges on the products that
are used to measure them. One thing is certain, the next few years are
going to be interesting in the display and measurement world.....
Tom Lianza
Director of Display and Input Capture Technologies
GretagMacbeth LLC
3 Industrial Drive
Unit 7&8
Windham, NH 03087
603.681.0315 x232 Tel
603.681.0316 Fax
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden